Re: extensions and translations.

[Marcus (OOo)]

Am 11/02/2012 02:02 PM, schrieb Andre Fischer:

On 02.11.2012 12:09, Marcus (OOo) wrote:

Am 11/01/2012 10:07 PM, schrieb jan iversen:

Can "standard" loosely be defined as an extension:
- is developed by people who have signed ICLA
- uses the apache license header in the source files


It's indeed important but IMHO this shouldn't be part of the decision
to draw the standard as it's about formal and general things.


- is of interest to the general public in different countries


Absolute.


- is willing to let the source be controlled/reviewed by committer.


With the possibility to become a committer later-on.


- accept a vote by the committers to be accepted


If a code grant is necessary depends maybe a bit on the amount of the
extension source code.


If those points are fuillfilled we could add the project to "swext", and
then it would automatically be integrated in the build and l10n process.


I think that is up to the author of an extension to put it in the AOO
source repository.
It is up to the community to decide whether to include it in a release.



Is "swext" only for extension around AOO Writer or general? If for
Writer then it should be located in a different, own directory within
the source code.


Only for Writer.

I know of at least three directories for extensions:

swext/
sdext/
extensions/

I created sdext myself, back in the days.
I think we should join the three directories, with extensions/ being the
natural candidate to remain.


+1

Then there is only one location where to search for extensions. Should 
it make more simple in the future, also for new joiners.


Marcus




By the way, there is no either or for extensions regarding their
presence in the extension repository or in the SVN repository. Many are
present in both.

-Andre




Please help me out here, I am not sure if that is enough for the "apache
way".


I would suggest to define the standard around some factors. Some
thoughts:

- What is the benefit for AOO?
- Is this helful for the general public or only for specific users?
- Does it exchange existing functionality with something own?
- What are the usage numbers / review comments look like?
- How big is the extension (keep in mind we shouldn't blow-up our
software too excessive).
- Don't install the extension by default but let the user decide what
they want, then make 1-3 wizard pages in the installer only for
installing extensions

Of course this can only work if the extension developer is willing to
come into the AOO project with all the things needed (source grant,
signed ICLA, header change, voting for releases, etc.).

Marcus




On 1 November 2012 21:24, Marcus (OOo) wrote:


Am 11/01/2012 01:17 PM, schrieb Rob Weir:

On Thu, Nov 1, 2012 at 3:52 AM, Jürgen Schmidt

wrote:


On 11/1/12 12:39 AM, Marcus (OOo) wrote:


Am 10/27/2012 01:17 AM, schrieb jan iversen:


I see, I have to get used to this license issues (a long time ago I
believed open source was just open source, then I joined an apache
project).

never mind.

Would it be to our advantage if we offered third party developers
(that is
how I see extension developers) the possibility to register a
language
file
and get it translated as part of the language packs ?



Of course it would be to our advantage; or let's say for the
project and
software. A lot of extensions would be available in many languages.

However, I don't know where we should draw the line to set a
limit. When
we select here and there some extensions, then the other
developers will
ask why not their extensions.



It's quite simple I would say, if people want develop extensions
under
ALv2 and want to contribute the code to the project. We can easy
create
a special section in our repo where we can host them.

But this means they have to be handled in the same way as all other
stuff here. Means a new release have to be voted...




+1

I think the important thing is this: We don't just want code. We
want communities. So if an extension author thinks that their
extension is generally useful and he/she wants to join the AOO
community and work on the extension here, and allow others to work on
it as well, then this is good.



Of course, +1.


We can have a set of "standard extensions".




So, we just need to define the standard.

Marcus




And IMHO it's not possible to translate all strings for all extensions.


But maybe others here have a great idea?



we can't probably provide it and I think we have to do enough ;-).
But I
can think of an alternative service hosted somewhere else.

Juergen



Or should we just say extension developers does not concern us (and

help
AOO get more used) so we just look the other way ?

Maybe the right way is somewhere in the middle.



Yeah, maybe. ;-)

Marcus



On 27 October 2012 00:58, Marcus (OOo) wrote:


Am 10/27/2012 12:36 AM, schrieb jan iversen:


While doing an update to the l10n 

Re: extensions and translations.

[Marcus (OOo)]

Am 11/02/2012 01:00 PM, schrieb jan iversen:

+1 to your ideas, much better formulated than mine.

see below for comments.

Jan

On 2 November 2012 12:09, Marcus (OOo)  wrote:


Am 11/01/2012 10:07 PM, schrieb jan iversen:

  Can "standard" loosely be defined as an extension:

- is developed by people who have signed ICLA
- uses the apache license header in the source files



It's indeed important but IMHO this shouldn't be part of the decision to
draw the standard as it's about formal and general things.


  - is of interest to the general public in different countries




Absolute.


  - is willing to let the source be controlled/reviewed by committer.




With the possibility to become a committer later-on.


  - accept a vote by the committers to be accepted




If a code grant is necessary depends maybe a bit on the amount of the
extension source code.


+1, but having the option of a vote is not bad...I did not want to write
"accept that a committer can veto the change".




  If those points are fuillfilled we could add the project to "swext", and

then it would automatically be integrated in the build and l10n process.



Is "swext" only for extension around AOO Writer or general? If for Writer
then it should be located in a different, own directory within the source
code.


At least Wiki publisher attaches only to writer. What do you mean "within
the source code", is main/swext not within ?


It is, but I meant "somewhere else in the code structure".

Maybe it's good to put all extension into an own structure. But I don't 
know the build system and talking maybe nuts. ;-)



  Please help me out here, I am not sure if that is enough for the "apache

way".



I would suggest to define the standard around some factors. Some thoughts:

- What is the benefit for AOO?


This might be a bit problematic, who is to judge it.


The community. When we want some extensions to be put into the AOO 
installer then we have to decide which ones. Of course this should be 
done with objective facts and not with thumbs up/down or something else.


If it's implementing a beautiful sidepane where some elements like 
navigator, stylist, etc. can be placed then it would be candidate #1.


If it's just adding a menu point to change all TOC items into a 
clickable link then this is not enough.


Just to define 2 extremes.


- Is this helful for the general public or only for specific users?


+1


- Does it exchange existing functionality with something own?


+1


- What are the usage numbers / review comments look like?


If I understand it correct, you see the extension first in the usual
extensions place, and then it can "grow" into AOO ?


Right.


Would there not be cases, where it was developed directly within AOO.


Then it would be already within AOO.
There are a few exceptions like the MySQL connector which has still a 
GPL license and therefore has to stay outside.



- How big is the extension (keep in mind we shouldn't blow-up our software
too excessive).


Is that not more a problem of release packaging ?


Hm, not really. I assume that the extension will bring in new code and 
not mostly redundant code. You will come to a point where you cannot get 
more optimazations without to put too much effort into it.


I'm not sure but you should see already a difference in size when 
comiling AOO with the extisting extensions (like the Wiki publisher) and 
then without any.



We could put the extensions in an own installation, like language packs.


Yes, some "Best of AOO extensions" compilations would be a nice idea.


- Don't install the extension by default but let the user decide what they
want, then make 1-3 wizard pages in the installer only for installing
extensions


+1



Of course this can only work if the extension developer is willing to come
into the AOO project with all the things needed (source grant, signed ICLA,
header change, voting for releases, etc.).


+1 that is important, extensions integrated in the source must obey the
same rules as all other source code.


Thanks for your further comments. When we (all of us) can find some more 
and come to an agreement, we can setup a definition how to get more 
extensions into AOO and their developers into our project.


Marcus




  On 1 November 2012 21:24, Marcus (OOo)   wrote:


  Am 11/01/2012 01:17 PM, schrieb Rob Weir:


   On Thu, Nov 1, 2012 at 3:52 AM, Jürgen Schmidt


   wrote:

  On 11/1/12 12:39 AM, Marcus (OOo) wrote:


  Am 10/27/2012 01:17 AM, schrieb jan iversen:


  I see, I have to get used to this license issues (a long time ago I

believed open source was just open source, then I joined an apache
project).

never mind.

Would it be to our advantage if we offered third party developers
(that is
how I see extension developers) the possibility to register a
language
file
and get it translated as 

Re: extensions and translations.

[Marcus (OOo)]

Am 11/01/2012 10:07 PM, schrieb jan iversen:

Can "standard" loosely be defined as an extension:
- is developed by people who have signed ICLA
- uses the apache license header in the source files


It's indeed important but IMHO this shouldn't be part of the decision to 
draw the standard as it's about formal and general things.



- is of interest to the general public in different countries


Absolute.


- is willing to let the source be controlled/reviewed by committer.


With the possibility to become a committer later-on.


- accept a vote by the committers to be accepted


If a code grant is necessary depends maybe a bit on the amount of the 
extension source code.



If those points are fuillfilled we could add the project to "swext", and
then it would automatically be integrated in the build and l10n process.


Is "swext" only for extension around AOO Writer or general? If for 
Writer then it should be located in a different, own directory within 
the source code.



Please help me out here, I am not sure if that is enough for the "apache
way".


I would suggest to define the standard around some factors. Some thoughts:

- What is the benefit for AOO?
- Is this helful for the general public or only for specific users?
- Does it exchange existing functionality with something own?
- What are the usage numbers / review comments look like?
- How big is the extension (keep in mind we shouldn't blow-up our 
software too excessive).
- Don't install the extension by default but let the user decide what 
they want, then make 1-3 wizard pages in the installer only for 
installing extensions


Of course this can only work if the extension developer is willing to 
come into the AOO project with all the things needed (source grant, 
signed ICLA, header change, voting for releases, etc.).


Marcus




On 1 November 2012 21:24, Marcus (OOo)  wrote:


Am 11/01/2012 01:17 PM, schrieb Rob Weir:

  On Thu, Nov 1, 2012 at 3:52 AM, Jürgen Schmidt

  wrote:


On 11/1/12 12:39 AM, Marcus (OOo) wrote:


Am 10/27/2012 01:17 AM, schrieb jan iversen:


I see, I have to get used to this license issues (a long time ago I
believed open source was just open source, then I joined an apache
project).

never mind.

Would it be to our advantage if we offered third party developers
(that is
how I see extension developers) the possibility to register a language
file
and get it translated as part of the language packs ?



Of course it would be to our advantage; or let's say for the project and
software. A lot of extensions would be available in many languages.

However, I don't know where we should draw the line to set a limit. When
we select here and there some extensions, then the other developers will
ask why not their extensions.



It's quite simple I would say, if people want develop extensions under
ALv2 and want to contribute the code to the project. We can easy create
a special section in our repo where we can host them.

But this means they have to be handled in the same way as all other
stuff here. Means a new release have to be voted...




+1

I think the important thing is this:  We don't just want code.  We
want communities.  So if an extension author thinks that their
extension is generally useful and he/she wants to join the AOO
community and work on the extension here, and allow others to work on
it as well, then this is good.



Of course, +1.


  We can have a set of "standard extensions".




So, we just need to define the standard.

Marcus




  And IMHO it's not possible to translate all strings for all extensions.


But maybe others here have a great idea?



we can't probably provide it and I think we have to do enough ;-). But I
can think of an alternative service hosted somewhere else.

Juergen



  Or should we just say extension developers does not concern us (and

help
AOO get more used) so we just look the other way ?

Maybe the right way is somewhere in the middle.



Yeah, maybe. ;-)

Marcus



  On 27 October 2012 00:58, Marcus (OOo)wrote:


  Am 10/27/2012 12:36 AM, schrieb jan iversen:


While doing an update to the l10n workflow I think I found a slight


problem.

Extensions offers the capability to integrate/extend our UI.

Assuming somebody writes an extension, and publishes it on
http://www.openoffice.org/extensions/<http://www.openoffice.org/**extensions/>
<http://www.**openoffice.org/extensions/<http://www.openoffice.org/extensions/>

how

does that get integrated into the
translation process ?



Simply, not at all.


As far as I can see the sources are not integrated into our "build
--all


--with-lang".



Right.


If I am right that they are not part of the general translation,
then is


that per design so or should it be different ?



Yes, this is by design.

Extensions are offered to extent your AOO install at any point of
time.
These

Re: extensions and translations.

[Marcus (OOo)]

Am 11/01/2012 01:17 PM, schrieb Rob Weir:

On Thu, Nov 1, 2012 at 3:52 AM, Jürgen Schmidt  wrote:

On 11/1/12 12:39 AM, Marcus (OOo) wrote:

Am 10/27/2012 01:17 AM, schrieb jan iversen:

I see, I have to get used to this license issues (a long time ago I
believed open source was just open source, then I joined an apache
project).

never mind.

Would it be to our advantage if we offered third party developers
(that is
how I see extension developers) the possibility to register a language
file
and get it translated as part of the language packs ?


Of course it would be to our advantage; or let's say for the project and
software. A lot of extensions would be available in many languages.

However, I don't know where we should draw the line to set a limit. When
we select here and there some extensions, then the other developers will
ask why not their extensions.


It's quite simple I would say, if people want develop extensions under
ALv2 and want to contribute the code to the project. We can easy create
a special section in our repo where we can host them.

But this means they have to be handled in the same way as all other
stuff here. Means a new release have to be voted...




+1

I think the important thing is this:  We don't just want code.  We
want communities.  So if an extension author thinks that their
extension is generally useful and he/she wants to join the AOO
community and work on the extension here, and allow others to work on
it as well, then this is good.


Of course, +1.


We can have a set of "standard extensions".


So, we just need to define the standard.

Marcus




And IMHO it's not possible to translate all strings for all extensions.

But maybe others here have a great idea?


we can't probably provide it and I think we have to do enough ;-). But I
can think of an alternative service hosted somewhere else.

Juergen




Or should we just say extension developers does not concern us (and help
AOO get more used) so we just look the other way ?

Maybe the right way is somewhere in the middle.


Yeah, maybe. ;-)

Marcus




On 27 October 2012 00:58, Marcus (OOo)   wrote:


Am 10/27/2012 12:36 AM, schrieb jan iversen:

   While doing an update to the l10n workflow I think I found a slight

problem.

Extensions offers the capability to integrate/extend our UI.

Assuming somebody writes an extension, and publishes it on
http://www.openoffice.org/**extensions/<http://www.openoffice.org/extensions/>how
does that get integrated into the
translation process ?



Simply, not at all.


   As far as I can see the sources are not integrated into our "build
--all

--with-lang".



Right.


   If I am right that they are not part of the general translation,
then is

that per design so or should it be different ?



Yes, this is by design.

Extensions are offered to extent your AOO install at any point of time.
These are developed by people that do not have to belong to our project
(when we put aside some exceptions). They can act independently. And
therefore they are allowed to (or have to ;-) ) do all on their own;
incl.
translation.

That applies for all extensions and templates available on:

-
http://extensions.services.**openoffice.org<http://extensions.services.openoffice.org>

-
http://templates.services.**openoffice.org<http://templates.services.openoffice.org>



   I might be following a wrong track here, but please forgive me for
trying

to make the l10n process as complete as I can.



Don't panic. That's a great goal and everybody is thankful to you for
doing this task.

Marcus

Re: extensions and translations.

[Marcus (OOo)]

Am 10/27/2012 01:17 AM, schrieb jan iversen:

I see, I have to get used to this license issues (a long time ago I
believed open source was just open source, then I joined an apache project).

never mind.

Would it be to our advantage if we offered third party developers (that is
how I see extension developers) the possibility to register a language file
and get it translated as part of the language packs ?


Of course it would be to our advantage; or let's say for the project and 
software. A lot of extensions would be available in many languages.


However, I don't know where we should draw the line to set a limit. When 
we select here and there some extensions, then the other developers will 
ask why not their extensions.


And IMHO it's not possible to translate all strings for all extensions.

But maybe others here have a great idea?


Or should we just say extension developers does not concern us (and help
AOO get more used) so we just look the other way ?

Maybe the right way is somewhere in the middle.


Yeah, maybe. ;-)

Marcus




On 27 October 2012 00:58, Marcus (OOo)  wrote:


Am 10/27/2012 12:36 AM, schrieb jan iversen:

  While doing an update to the l10n workflow I think I found a slight

problem.

Extensions offers the capability to integrate/extend our UI.

Assuming somebody writes an extension, and publishes it on
http://www.openoffice.org/**extensions/<http://www.openoffice.org/extensions/>how
 does that get integrated into the
translation process ?



Simply, not at all.


  As far as I can see the sources are not integrated into our "build --all

--with-lang".



Right.


  If I am right that they are not part of the general translation, then is

that per design so or should it be different ?



Yes, this is by design.

Extensions are offered to extent your AOO install at any point of time.
These are developed by people that do not have to belong to our project
(when we put aside some exceptions). They can act independently. And
therefore they are allowed to (or have to ;-) ) do all on their own; incl.
translation.

That applies for all extensions and templates available on:

- 
http://extensions.services.**openoffice.org<http://extensions.services.openoffice.org>
- 
http://templates.services.**openoffice.org<http://templates.services.openoffice.org>


  I might be following a wrong track here, but please forgive me for trying

to make the l10n process as complete as I can.



Don't panic. That's a great goal and everybody is thankful to you for
doing this task.

Marcus

Re: extensions and translations.

[Marcus (OOo)]

Am 10/27/2012 03:53 AM, schrieb Ariel Constenla-Haile:

On Sat, Oct 27, 2012 at 01:17:33AM +0200, jan iversen wrote:

I see, I have to get used to this license issues (a long time ago I
believed open source was just open source, then I joined an apache project).


It has nothing to do with licensing. Even if the extension code and all
its dependencies are under the ALv2, why should OpenOffice include
extensions by default in the install set? This goes against the concept
of an extension.

The fact that now there are three supported extensions is just
a question of old Sun/Oracle decisions to release these as extension and
not integrated as part of the application.



never mind.

Would it be to our advantage if we offered third party developers (that is
how I see extension developers) the possibility to register a language file
and get it translated as part of the language packs ?


This will break several concepts and things. Mainly extension developers
have complete freedom about when to release updates, how to integrate
translation in their extensions (use the configuration API and XCU
files, use the resource API and Java-property-like files, etc.), most
important what license to choose, etc.


Right. Then they are no longer independent from their own release plan 
as they may would like to be.


Furthermore, I don't know how many extensions Sourceforge is hosting but 
it could hundreads if not thousands. Lets assume 10% of them will say 
"yes AOO, please translate the strings for me" then we have a lot more 
to do. And who should verify the translations? If we, then we have to 
install these hundreads/thousands extensions to see how it's working. If 
the developer, then do we get feedback in time? What about errors, are 
we able to fix them in time?


I think you can see now that the extension development has it own small 
but fine ecosystem. It doesn't fit into the AOO release process and IMHO 
this was never the idea.


However, thanks for making these thoughts. :-)

Marcus




In short, you will have to implement a new framework and force
extensions developers to use it. Besides several concerns, legal
concerns among them.



Or should we just say extension developers does not concern us (and help
AOO get more used) so we just look the other way ?


Programmability and extensibility has always been a priority in
OpenOffice, just read the Developer's Guide and other parts of the wiki.

I tend to agree that it will be useful for an extension developer a way
to submit a set of resource strings and get them translated, as long as
the extension developer is not forced with release/legal/other concerns.

Re: [DISCUSS] Cleanup installation files, make them more modular

[Marcus (OOo)]

Am 10/27/2012 12:58 AM, schrieb Ariel Constenla-Haile:

On Sat, Oct 27, 2012 at 12:47:56AM +0200, Marcus (OOo) wrote:

I don't see the context to bad inet connections. Maybe you can tell
with some other words?

When we do the restructure then we will have less content on the
mirrors. But for the user there is no change as they still have to
download all files that are necessary to do the installation; maybe
less when they explicitely disables some applications and content
that are currently be installed by default.


If I understood this approach, the tiny installer downloads all the rest
at install time, that means you need a good internet connection at
install time.


Not really, the installer can first download everything that is 
necessary and then start the installation. It should be technically also 
possible to continue a download where it stopped due to an interrupted 
inet connection.


So, at the end it's up to us how to design the tiny installer.

Marcus

Re: extensions and translations.

[Marcus (OOo)]

Am 10/27/2012 12:36 AM, schrieb jan iversen:

While doing an update to the l10n workflow I think I found a slight problem.

Extensions offers the capability to integrate/extend our UI.

Assuming somebody writes an extension, and publishes it on
http://www.openoffice.org/extensions/ how does that get integrated into the
translation process ?


Simply, not at all.


As far as I can see the sources are not integrated into our "build --all
--with-lang".


Right.


If I am right that they are not part of the general translation, then is
that per design so or should it be different ?


Yes, this is by design.

Extensions are offered to extent your AOO install at any point of time. 
These are developed by people that do not have to belong to our project 
(when we put aside some exceptions). They can act independently. And 
therefore they are allowed to (or have to ;-) ) do all on their own; 
incl. translation.


That applies for all extensions and templates available on:

- http://extensions.services.openoffice.org
- http://templates.services.openoffice.org


I might be following a wrong track here, but please forgive me for trying
to make the l10n process as complete as I can.


Don't panic. That's a great goal and everybody is thankful to you for 
doing this task.


Marcus

Re: [DISCUSS] Cleanup installation files, make them more modular

[Marcus (OOo)]

Am 10/27/2012 12:33 AM, schrieb Ariel Constenla-Haile:

On Fri, Oct 26, 2012 at 11:06:45PM +0200, Marcus (OOo) wrote:

I've modified the subject as I think this topic deserves its own,
new thread.

Am 10/26/2012 07:28 PM, schrieb Ariel Constenla-Haile:

On Wed, Oct 24, 2012 at 05:27:41PM +0200, Jürgen Schmidt wrote:

Once thing to pay attention for the next release is the increasing size:
more than 14 Gb for Linux packages only. This is going to be even more,
as more languages are added. INFRA has already complained after the
first release (can't find the message right now) about the size of our
dist/ folder, so we must think about a solution, before they complain
once the next release is uploaded.


IMHO you can think and try whatever you want. At the end there is
only one solution:

Cleanup the packaging, delete redundat files, rearrange how the
install files will be packed, think new how the installation on the
users-side could be done.

Example:
For every platform, we have exactly the same files in every full
install, except for the language resource files. So, when we can
make it that only the core (languages-independent) files are once on
the mirrors and then the language resource files besides, then it
would be possible to do the installation process completely new -
with the following rough steps:

1. Create a new basis installer: little, tiny and already localized.
2. The user can choose what he wants: applications, languages,
templates, extensions.
3. The basis installer downloads this file set from the mirrors.


I've read this approach the other times this was discussed. While this
might be the current mainstream market trend, handy for those who send
their e-mails from their i-Phones and their i-Pads, it won't be suitable
for users from less-developed countries with bad internet connection.

Does OpenOffice user base come from this kind of countries? These
numbers don't seem to tell so:
http://www.openoffice.org/stats/countries.html But I've also read "why
would someone use OpenOffice if he/she can pay for MS Office?", meaning
that OpenOffice user base is made of people who can't afford MS Office,
we could also asume they can't afford a good internet connection, even
on developed countries like the top 7 in the list (this argument missed
the point than someone may want to use OpenOffice just because it's free
software, even if can pay for MS Office - or get an ilegal copy).


I don't see the context to bad inet connections. Maybe you can tell with 
some other words?


When we do the restructure then we will have less content on the 
mirrors. But for the user there is no change as they still have to 
download all files that are necessary to do the installation; maybe less 
when they explicitely disables some applications and content that are 
currently be installed by default.


Marcus

Re: [RELEASE] Releasing new languages for 3.4.1

[Marcus (OOo)]

Am 10/26/2012 11:46 PM, schrieb jan iversen:

On 26 October 2012 23:38, Marcus (OOo)  wrote:


Am 10/26/2012 11:20 PM, schrieb jan iversen:

  On 26 October 2012 23:06, Marcus (OOo)   wrote:


  Am 10/26/2012 07:43 PM, schrieb Andrea Pescetti:


   Rob Weir wrote:



  1) release new languages via lang packs only for now

2) release full installs, but for only these new languages



I don't see a big difference between a langpack and a full install in
this case, so I'd go for full installs, unless releasing langpacks helps
in communicating that these are "late" additions and that full installs
will come with the next release.

   Can we really skip the release process? PO files == source, right?





Yes, not exactly but quite (PO files are not taken verbatim into source,
but they are imported and influence resource files which are in the
source tree).

   Maybe a question for legal-discuss if we're not certain.





If in the end we have consensus on releasing new languages for 3.4.1
instead of making a new release, indeed we will ask.

   How do we want to handle this on an ongoing basis? New point release


for every new language? Every 5 new languages? It is certainly good
for volunteers to get the encouragement of a fast turnaround for their
work. But this is the same for a C++ programmer.



There are big differences here, that are also the reason for me to
consider releasing these new languages as soon as possible:
- A translation is often done by a team; if we can publish it
immediately, the team can the be involved in other activities like
revamping the N-L website, local promotion and so on; if we wait too
much, we risk to have no volunteers for the following release.



Really? I'm not that convinced that this would happen. When we
communicate
from the beginning when new loalizations will be released then everybody
should be able to understand and handle this.


   - Releasing a new language is totally risk-free: a new language can't


break functionality in OpenOffice, while any feature could have bugs and
needs more qualified testing.



Besides the comment from Jan I remember a case from the old OOo project.
There were some translations for the names of Calc functions that got the
same name but had to get (slightly) different names. The result was that
there were 2-3 sum, 2-3 average, etc. functions. This was also - more or
less the only - reason for another respin for a OOo RC; 3.2.1, 3.3.0, I
don't remember anymore.

So, the risk of new languages may not be high but I wouldn't say it's
totally risk-free.


   In the end, I wonder whether the best solution is to get into a steady


release cycle of quarterly releases (every 3 or 4 months)?




  +1


IMHO a regular release schedule is a very good idea. Then everybody can
cope with this, can see when the next version will come and we can plan
with a regular release plan (when to branch, freeze, localize, etc.).

Of course the timeframe will need some discussions to find the right one.

Previously it was tried to release every 6 months a new major release and
every 6 months a point release. So, with overlapping there was a new
release every 3 month. Maybe a good timeframe to continue?



+1 to a relatively fixed time frame for new releases. Not only developers
benefit from that but also end-users !



Right


  However do we have the logistic in place to handle ideas/request/bug fixes

with these short intervals. It would mean (in my opinion) that we have an



OK, maybe the following fitts better to our current situation. Every 6
months a new major release and a point release on demand - enough new
languages, urgent/severe bugfixes; that means outside a regular release
plan.


+1




  open catalog (new development) for 2-3 releases and have to prioritize

within a limited timeframe what goes where ? We should also consider to
apply a field in bugzilla, "targeted for version".



That's already existing. Just look for the "Target Milestone" field.


I think it is not really used (I might be wrong) but with frequent releases
we should use it intensively, because today those who submit a bug must be
pretty disappointed, I looked at a bug the other day, dated 2007 which are
still a bug.


Ah, now I get it.

This problem comes from the migration from the old OOo's Issuezilla to 
Apache's Bugzilla. Here the old field disappeared and the information 
was not shown in Bugzilla.


The field you see now was added afterwards and has to be filled newly.

Marcus




  I really like the idea, but it has a tendency of killing long term

developments, because they are hard to put into this framework, so we need
something in the middle.



When we plan which new and planned feature goes into what release should
work.


I think I did not express it correctly, resources tend to be used for short
term targets (next release, high motivation, lets make it folks, and after
that 

Re: [RELEASE] Releasing new languages for 3.4.1

[Marcus (OOo)]

Am 10/26/2012 11:35 PM, schrieb Rob Weir:

On Fri, Oct 26, 2012 at 5:06 PM, Marcus (OOo)  wrote:

Am 10/26/2012 07:43 PM, schrieb Andrea Pescetti:


Rob Weir wrote:


1) release new languages via lang packs only for now
2) release full installs, but for only these new languages



I don't see a big difference between a langpack and a full install in
this case, so I'd go for full installs, unless releasing langpacks helps
in communicating that these are "late" additions and that full installs
will come with the next release.


Can we really skip the release process? PO files == source, right?



Yes, not exactly but quite (PO files are not taken verbatim into source,
but they are imported and influence resource files which are in the
source tree).


Maybe a question for legal-discuss if we're not certain.



If in the end we have consensus on releasing new languages for 3.4.1
instead of making a new release, indeed we will ask.


How do we want to handle this on an ongoing basis? New point release
for every new language? Every 5 new languages? It is certainly good
for volunteers to get the encouragement of a fast turnaround for their
work. But this is the same for a C++ programmer.



There are big differences here, that are also the reason for me to
consider releasing these new languages as soon as possible:
- A translation is often done by a team; if we can publish it
immediately, the team can the be involved in other activities like
revamping the N-L website, local promotion and so on; if we wait too
much, we risk to have no volunteers for the following release.



Really? I'm not that convinced that this would happen. When we communicate
from the beginning when new loalizations will be released then everybody
should be able to understand and handle this.



- Releasing a new language is totally risk-free: a new language can't
break functionality in OpenOffice, while any feature could have bugs and
needs more qualified testing.



Besides the comment from Jan I remember a case from the old OOo project.
There were some translations for the names of Calc functions that got the
same name but had to get (slightly) different names. The result was that
there were 2-3 sum, 2-3 average, etc. functions. This was also - more or
less the only - reason for another respin for a OOo RC; 3.2.1, 3.3.0, I
don't remember anymore.

So, the risk of new languages may not be high but I wouldn't say it's
totally risk-free.



Certainly the risk is reduced.  But there are two areas:

1) Risk of defects caused by interaction between the core product and
the translated strings

2) Risk of a "bad build", for whatever reason, say due to change in a
system library, leading to an undetected new defect.




In the end, I wonder whether the best solution is to get into a steady
release cycle of quarterly releases (every 3 or 4 months)?



+1

IMHO a regular release schedule is a very good idea. Then everybody can cope
with this, can see when the next version will come and we can plan with a
regular release plan (when to branch, freeze, localize, etc.).

Of course the timeframe will need some discussions to find the right one.

Previously it was tried to release every 6 months a new major release and
every 6 months a point release. So, with overlapping there was a new release
every 3 month. Maybe a good timeframe to continue?



Did you do betas for all releases?  Or only major ones?  Or was this a
case-by-case decision?


Always for big major releases, here in a really public way (with 
announcements, etc.) and full blown install files. But also here and 
there also for chosen releases with en-US full install + langpacks. I 
think everybody remembers the last one - OOo 3.4.0. ;-)



We have the ability to do betas if we want.  From an Apache
perspective they would still be releases, but we could set the right
expectations with users.  For example, we wouldn't send update
notifications for beta releases.


Right.

And looking into my crystal ball, I predict that AOO 4.0 will start with 
a beta release.


Marcus




This could be a solution too. In this case we would have the problem of
choosing what to translate (3.4 or 3.5? probably we would ask new
volunteers to focus on strings that will be in the next release, even
though they aren't frozen yet).



In any case we should continue to release new languages; regardless if major
or point versions.

Marcus

Re: The Impossible Question

[Marcus (OOo)]

Am 10/26/2012 11:38 PM, schrieb Kay Schenk:

On Fri, Oct 26, 2012 at 2:30 PM, Louis Suárez-Pottswrote:



On 12-10-26, at 17:25 , Rob Weir  wrote:


Of course, the real question, Louis, is what do *you* want to do?


Be cynical, or I mean, more cynical.

Seriously, I have the same objections and observations as you do but I've
also spent 11 years dealing with people who don't. No one thing will work.
That's why I suggest multiple redundancies, and to place the instances
alerting users of opportunities in key areas—the download, for instance
(where we also used to locate contribution options), but also the Help,
About, etc. But what's usually best is to follow what others have done, and
to figure that the path has been torched, it's the one that will be
followed, so let's use it.



Louis--

I see "Support" on the home page as it's own major link; "Support" as a tab
available from all pages; and "Support" (probably not best placed) on the
download page in Documentation. We could certainly move this to "Additional
Information" or its own category placement.


Hm, or an own colored button below a specific button, so not to put it 
at the bottom of the webpage. Then link it to 
"http://www.openoffice.org/support";.


Marcus





If you feel something additional might help, it would be best to state some
specific navigation areas or 



As I don't use MSFT Office and the last time I did found myself deep in
the sea of thick frustration, I'm the last person to ask about this.

louis

Re: The Impossible Question

[Marcus (OOo)]

Am 10/26/2012 11:25 PM, schrieb jan iversen:

Just an idea, which once helped me.

Typically users dont think things will go wrong so they dont pay attention
to whatever we write, but one apache project had a quite clever solution


And when they remember that there was a banner talking about "How to get 
help in case of problems", hm, how to get this information back; as the 
user has of course not wrote it up.



(if I remember right it was Axis, but dont depend on my memory), when the
installation failed it did not just tell you failed, it came up with the
link to FAQ, right when the user needed it, simple but very effective.


That is good with problems while installing. However, when the install 
was fine but AOO is crashing every 5 seconds after every start is 
different. Even a link to the support webpage in the "Help" menu is not 
really helping the user.


OK, maybe an extrem case. ;-)

Marcus




On 26 October 2012 23:06, Louis Suárez-Potts  wrote:


Hi
Every now and then a user finds the experience of downloading, installing,
using AOO disappointing and frankly frustrating if not worse. They will
usually go to the user forums, but sometimes they will contact the Apache
Foundation directly. Okay, but this does not really help them.

What we did with OpenOffice was set up a Support page, which has since
been moved to here, http://www.openoffice.org/support/. It's pretty much
an improved version of the old but of course the "ecosystem" needs further
fleshing out—it suffers from a lack of substantial existence.

I'm also not persuaded that the route to it from either the application
download page or homepage or wherever is redundantly clear enough for the
befuddled enduser who installs AOO to replace his or her whatever suite and
doesn't really know where to go…..

So, my query is the usual impossible question: What can we do to make it
clearer to the puzzled and frustrated how to get help? Sure, we can have a
knowledge base (kb), FAQ, etc., and also enthusiastic community members.

But what would you suggest as a path, or paths for the user? I personally
would include something in the installation sets that point to the support
page above; but also banners, say, or tags, stickers—glaringly obvious neon
coloured blinking lights?—to relay users to useful pages.

Ideas?

Thanks
Louis

Re: [RELEASE] Releasing new languages for 3.4.1

[Marcus (OOo)]

Am 10/26/2012 11:20 PM, schrieb jan iversen:

On 26 October 2012 23:06, Marcus (OOo)  wrote:


Am 10/26/2012 07:43 PM, schrieb Andrea Pescetti:

  Rob Weir wrote:



1) release new languages via lang packs only for now
2) release full installs, but for only these new languages



I don't see a big difference between a langpack and a full install in
this case, so I'd go for full installs, unless releasing langpacks helps
in communicating that these are "late" additions and that full installs
will come with the next release.

  Can we really skip the release process? PO files == source, right?




Yes, not exactly but quite (PO files are not taken verbatim into source,
but they are imported and influence resource files which are in the
source tree).

  Maybe a question for legal-discuss if we're not certain.




If in the end we have consensus on releasing new languages for 3.4.1
instead of making a new release, indeed we will ask.

  How do we want to handle this on an ongoing basis? New point release

for every new language? Every 5 new languages? It is certainly good
for volunteers to get the encouragement of a fast turnaround for their
work. But this is the same for a C++ programmer.



There are big differences here, that are also the reason for me to
consider releasing these new languages as soon as possible:
- A translation is often done by a team; if we can publish it
immediately, the team can the be involved in other activities like
revamping the N-L website, local promotion and so on; if we wait too
much, we risk to have no volunteers for the following release.



Really? I'm not that convinced that this would happen. When we communicate
from the beginning when new loalizations will be released then everybody
should be able to understand and handle this.


  - Releasing a new language is totally risk-free: a new language can't

break functionality in OpenOffice, while any feature could have bugs and
needs more qualified testing.



Besides the comment from Jan I remember a case from the old OOo project.
There were some translations for the names of Calc functions that got the
same name but had to get (slightly) different names. The result was that
there were 2-3 sum, 2-3 average, etc. functions. This was also - more or
less the only - reason for another respin for a OOo RC; 3.2.1, 3.3.0, I
don't remember anymore.

So, the risk of new languages may not be high but I wouldn't say it's
totally risk-free.


  In the end, I wonder whether the best solution is to get into a steady

release cycle of quarterly releases (every 3 or 4 months)?




+1

IMHO a regular release schedule is a very good idea. Then everybody can
cope with this, can see when the next version will come and we can plan
with a regular release plan (when to branch, freeze, localize, etc.).

Of course the timeframe will need some discussions to find the right one.

Previously it was tried to release every 6 months a new major release and
every 6 months a point release. So, with overlapping there was a new
release every 3 month. Maybe a good timeframe to continue?


+1 to a relatively fixed time frame for new releases. Not only developers
benefit from that but also end-users !


Right


However do we have the logistic in place to handle ideas/request/bug fixes
with these short intervals. It would mean (in my opinion) that we have an


OK, maybe the following fitts better to our current situation. Every 6 
months a new major release and a point release on demand - enough new 
languages, urgent/severe bugfixes; that means outside a regular release 
plan.



open catalog (new development) for 2-3 releases and have to prioritize
within a limited timeframe what goes where ? We should also consider to
apply a field in bugzilla, "targeted for version".


That's already existing. Just look for the "Target Milestone" field.


I really like the idea, but it has a tendency of killing long term
developments, because they are hard to put into this framework, so we need
something in the middle.


When we plan which new and planned feature goes into what release should 
work.


Marcus




  This could be a solution too. In this case we would have the problem of

choosing what to translate (3.4 or 3.5? probably we would ask new
volunteers to focus on strings that will be in the next release, even
though they aren't frozen yet).



In any case we should continue to release new languages; regardless if
major or point versions.

Marcus

[DISCUSS] Cleanup installation files, make them more modular

[Marcus (OOo)]

I've modified the subject as I think this topic deserves its own, new 
thread.


Am 10/26/2012 07:28 PM, schrieb Ariel Constenla-Haile:

On Wed, Oct 24, 2012 at 05:27:41PM +0200, Jürgen Schmidt wrote:

Once thing to pay attention for the next release is the increasing size:
more than 14 Gb for Linux packages only. This is going to be even more,
as more languages are added. INFRA has already complained after the
first release (can't find the message right now) about the size of our
dist/ folder, so we must think about a solution, before they complain
once the next release is uploaded.


IMHO you can think and try whatever you want. At the end there is only 
one solution:


Cleanup the packaging, delete redundat files, rearrange how the install 
files will be packed, think new how the installation on the users-side 
could be done.


Example:
For every platform, we have exactly the same files in every full 
install, except for the language resource files. So, when we can make it 
that only the core (languages-independent) files are once on the mirrors 
and then the language resource files besides, then it would be possible 
to do the installation process completely new - with the following rough 
steps:


1. Create a new basis installer: little, tiny and already localized.
2. The user can choose what he wants: applications, languages, 
templates, extensions.

3. The basis installer downloads this file set from the mirrors.
4. It does the installation on the PC.
5. Finally AOO is ready to use.

And as goody maybe it's possible to create an install file, so that the 
user can re-install his AOO version whenever he wants, without to 
download everything again from the mirrors.


Marcus

Re: [RELEASE] Releasing new languages for 3.4.1

[Marcus (OOo)]

Am 10/26/2012 07:43 PM, schrieb Andrea Pescetti:

Rob Weir wrote:

1) release new languages via lang packs only for now
2) release full installs, but for only these new languages


I don't see a big difference between a langpack and a full install in
this case, so I'd go for full installs, unless releasing langpacks helps
in communicating that these are "late" additions and that full installs
will come with the next release.


Can we really skip the release process? PO files == source, right?


Yes, not exactly but quite (PO files are not taken verbatim into source,
but they are imported and influence resource files which are in the
source tree).


Maybe a question for legal-discuss if we're not certain.


If in the end we have consensus on releasing new languages for 3.4.1
instead of making a new release, indeed we will ask.


How do we want to handle this on an ongoing basis? New point release
for every new language? Every 5 new languages? It is certainly good
for volunteers to get the encouragement of a fast turnaround for their
work. But this is the same for a C++ programmer.


There are big differences here, that are also the reason for me to
consider releasing these new languages as soon as possible:
- A translation is often done by a team; if we can publish it
immediately, the team can the be involved in other activities like
revamping the N-L website, local promotion and so on; if we wait too
much, we risk to have no volunteers for the following release.


Really? I'm not that convinced that this would happen. When we 
communicate from the beginning when new loalizations will be released 
then everybody should be able to understand and handle this.



- Releasing a new language is totally risk-free: a new language can't
break functionality in OpenOffice, while any feature could have bugs and
needs more qualified testing.


Besides the comment from Jan I remember a case from the old OOo project. 
There were some translations for the names of Calc functions that got 
the same name but had to get (slightly) different names. The result was 
that there were 2-3 sum, 2-3 average, etc. functions. This was also - 
more or less the only - reason for another respin for a OOo RC; 3.2.1, 
3.3.0, I don't remember anymore.


So, the risk of new languages may not be high but I wouldn't say it's 
totally risk-free.



In the end, I wonder whether the best solution is to get into a steady
release cycle of quarterly releases (every 3 or 4 months)?


+1

IMHO a regular release schedule is a very good idea. Then everybody can 
cope with this, can see when the next version will come and we can plan 
with a regular release plan (when to branch, freeze, localize, etc.).


Of course the timeframe will need some discussions to find the right one.

Previously it was tried to release every 6 months a new major release 
and every 6 months a point release. So, with overlapping there was a new 
release every 3 month. Maybe a good timeframe to continue?



This could be a solution too. In this case we would have the problem of
choosing what to translate (3.4 or 3.5? probably we would ask new
volunteers to focus on strings that will be in the next release, even
though they aren't frozen yet).


In any case we should continue to release new languages; regardless if 
major or point versions.


Marcus

Re: [PROPOSAL] "difficulty" field for Bugzilla

[Marcus (OOo)]

Am 10/24/2012 09:08 PM, schrieb Rob Weir:

As you have probably noticed, I'm engaged in a variety of initiatives
to grow the community, bring in more volunteers, etc.  One additional
piece that I think would be useful is to add a new field to Bugzilla
to indicate the difficulty level of the bug.  Of course, this will
often not be known.  But in some cases, we do know, and where we do
know we can indicate this.

What this allows us to do is then have search filters that return only
open easy bugs.  These are ideal for new developer volunteers on the
project who are looking for items that match their lesser familiarity
with the code.  It also allows a developer to step up to more
challenging bugs over time.

A similar approach, which they called "easy hacks", was successfully
used by LibreOffice.

If there are no objections, I'll add a new field to Bugzilla called
"cf_difficulty_level", and which a drop down UI with the following
choices:

UNKNOWN (default)
TRIVIAL
EASY
MODERATE
HARD
WIZARD

(I'm certainly open to variations on the names)

I'd then rely on other developers to help "seed" the database with
some TRIVIAL and EASY bugs, so new volunteers will have something to
work with as they familiarize themselves with the project.

I'll wait 72 hours, etc.


Even if it was not really 3 days ;-) and I'm a bit late, I just wanted 
to tell that this seems good idea to attract more new volunteers to get 
an entry point into our project.


Marcus

Re: Updated Stats page

[Marcus (OOo)]

Am 10/23/2012 08:48 PM, schrieb Rob Weir:

On Tue, Oct 23, 2012 at 12:10 PM, jan iversen  wrote:

A report would serve the same purpose, and posted with regular intervals
(like 1 month), with figures based on e.g. last month, last half year.



OK.  I posted a snapshot of the downloads since AOO 3.4.0 was released
back in May:

http://www.openoffice.org/stats/countries.html


Great.

Is also possible to have a row with hte total sum on the top of the 
table? Then we would have a (more or less) up-to-date number in a 
prominent place.



One option would be to repeat this in 3 months or whatever, and add a
new column and % difference for each country.


Sounds good.

Marcus




On 23 October 2012 18:00, Rob Weir  wrote:


On Tue, Oct 23, 2012 at 11:23 AM, jan iversen
wrote:

+1, fine page. Monthly should be enough.

Would it be worth to consider to include the country discussion (e.g.

list

top 5 countries, and the total number) ?



I'll take a look to see if there is anything interesting here.  But my
guess is the top 5 countries will be static over time, and would not
be an interesting chart or a time series.  But maybe we could
periodically post a table of these numbers?   I have a pythons script
that gathers these numbers and generates a CSV report.  It would be
easy to have it write out an HTML page instead.

-Rob


Jan.

On 23 October 2012 17:20, Rob Weir  wrote:


I've moved the download stats to its own page:
http://www.openoffice.org/stats/downloads.html

That allowed me to give a fuller description of what the stats are,
how they were gathered, etc.  I think we should aim for this level of
detail and transparency in any claims we make.

This move then allowed me to clean up the Stats home page a little,
and include links to other charts we have, as well as add a section
(with caveats) on 3rd party stats:   http://www.openoffice.org/stats/

If anyone has ideas for other relevant stats that might be interested,
I'd be interested in adding more.  I can help on the data wrangling
and charting side.I'd love to have a regular chart on wiki and
forums traffic or edits or posts or whatever.  This doesn't need to be
totally automated.  For example, it could be something where someone
volunteers to run a monthly report and posts that new stat once a
month.


Regards,

-Rob







--

Ciao

Marcus

Re: [WWW]: shared ideas and looking for feedback

[Marcus (OOo)]

Am 10/19/2012 10:26 AM, schrieb jan iversen:

It would be a good idea to have the same structure and then one directory
with country special parts...as you say it makes it easier to maintain, and
with the extra directory nobody is limited.


I support this idea.

When we extent this also for the translated release notes like:

.../rn/release_notea_aoo341.html
.../de//rn/release_notea_aoo341.html
.../it//rn/release_notea_aoo341.html

then we can change the already existing link on the download website, 
from the now English only release notes to the language-related notes.


Would be a nice additonal service for our users.


On 19 October 2012 10:22, Jürgen Schmidt  wrote:


Hi,

yesterday I had problems to find a good place for the German translation
of the graduation press release. And I thought that it is probably a
good idea to cleanup the whole page with a clear and well defined
structure. I know that there is work ongoing and that we move already in
this direction. But nevertheless I would like to share the things I have
in mind to check if it is aligned with the already ongoing work or if it
makes sense at all.

1. a clear structure for the English content as well as the translated
pages.

.../index.hmtl
.../de/index.html
.../it/index.html
...
.../press/msg_20121019.html
.../de/press/msg_20121019.html
.../it/press/msg_20121019.html
...

Means we have for all pages a translated version in the related sub
directory. Same path and same name only the content is translated. This
makes it easy to find the related translation for any files.

We can also use Pootle to do the translation of the web content in the
future.

2. we have special news areas where local communities can spread further
news relevant to their local activities, e.g. local conferences, events.
But in general we have the same content on all pages. Other local
community relevant content should be moved in the wiki. The main idea is
to have a smaller but cleaner and well structured and organized user
portal www.openoffice.org. Community internal things should be move on
openoffice.apache.org or the wiki.

I know it is not really new and it is probably more to remind myself but
I am interested to hear others opinion.


Simply +1, with the addition from above.

Marcus

Re: Estimating contributors by looking at wiki accounts?

[Marcus (OOo)]

Am 10/19/2012 05:28 PM, schrieb Rob Weir:

I recently saw another open source project claim that they had over
3000 contributors.  They derived this estimate by looking at the
number of user accounts they had in their wiki.

That is quite clever, I thought.  Since we use the same wiki software,
I thought I'd check this metric for us.  Our wiki says we have over
58,000 user accounts.

I know we're doing well, but would it really make sense to claim that
we have over 58,000 contributors?  I don't think so.

I suppose we could look only at accounts where the person has actually
contributed edits, or even recent edits. (MediaWiki is a well-known
target of registration spam).  Although the other project did not seem
to filter out inactive or unused accounts, I think the metrics are
meaningless unless we do that.

What do you think?  Or do we even care?


Yes, maybe a good chance to tell others some numbers from our project.

However, the wording of the number is (for some people) the more 
important part. So, this should be double-checked.


That means it doesn't make sense to say "hey, we have 58,000+ 
contributors" but more like "... in the last 12 months we got 
contributions from ~3,000 active people (incl. accounts from SVN, BZ, 
Wiki, MLs, etc.).


Marcus

Re: [EXT] MySQL SDBC Connector

[Marcus (OOo)]

Am 10/15/2012 09:13 PM, schrieb Andrea Pescetti:

On 12/10/2012 Ariel Constenla-Haile wrote:

On Sun, Sep 09, 2012 at 07:52:03PM +0200, Roberto Galoppini wrote:

Hi Ariel, would you be interested in managing the AOO Extension page
dedicated to the MySQL connector? At the present stage is
unmaintained, I
believe you'd be the 'natural' candidate for that.


I finally ended up setting a new project page:
http://extensions.openoffice.org/en/project/mysql-driver
It wouldn't be fair to take the current page


Actually, I feel it would be fair to take it! If the extension is
unmaintained (meaning that the user who was maintaining it does no
longer have an account on the Extensions site) then it is confusing to
users to have the unmaintained version and your patched one available.


Sorry that I haven't reported this earlier.

I remember that I've maintained this webpage. Of course not as the 
developer but as (kind of) liaison officer between OOo community and 
Sun/Oracle. And yes, I don't have access to it anymore, so I cannot 
change it myself.


I agree with Andrea that it's better to continue the already (and wrt 
the stats very popular!) webpage than to have two with the same purpose.



It occurred already a couple times that I had to explicitly point users
to your patched version and that they solved their problems with it. So
it surely deserves better exposure.


Week: 795 | Month: 3,850 | Year: 45,050


If you don't host your version on extensions.openoffice.org, it will
always be shadowed by this (buggy) one, because downloads won't be tracked.

So, seeing this as a user, the best solutions I can imagine are:

1) Take over http://extensions.openoffice.org/en/project/mysql_connector
; old releases are archived, available and credited appropriately;
nothing new will happen there since no users can update the page.


+1


2) Keep both projects, but state clearly on the older one that it is
abandoned and that bugfixes and 3.4.x compatibility can be obtained in
your version, and link to it. In this case, it would help if you hosted
the packages on extensions.openoffice.org so that it can be shown in the
"most popular" listing and gradually shadow the old one.

But at the very least we should agree on a replacement text for the 3.4
compatibility notice at
http://extensions.openoffice.org/en/project/mysql_connector and get it
replaced.


Marcus

Re: Language Pack Pt

[Marcus (OOo)]

Am 10/15/2012 02:44 PM, schrieb Jean Milot:

I search some informations about the pt language pack.


You can find all files for the AOO 3.4.1 release here:

http://www.openoffice.org/download/other.html

A language pack for European Portuguese (pt-PT) is not yet available but 
if you don't mind you can use the language pack for Brazilian Portuguese 
(pt-BR).


If you really need an office installation with pt-PT please have a look 
for the legacy OOo release files (OOo 3.3.0 and older) here:


http://www.openoffice.org/download/legacy/other.html

HTH

Marcus

Re: Support questions to ooo-private?

[Marcus (OOo)]

Am 10/15/2012 02:23 PM, schrieb Rob Weir:

I know there was a mystery why we were getting a few questions/week to
ooo-private.   I think I found a clue.

Start from the bug tracking page, our BZ instance:

https://issues.apache.org/ooo/

Note the language at the bottom "In case of problems with the
functioning of Apache OOo Bugzilla...".  Average user doesn't know
Bugzilla from their elbow.  They see "problem" and "Apache" and send a
note.

The get an auto-response that says:

"Thank you for contacting The Apache Software Foundation OpenOffice Bugzilla
Administrators.

- If your concern is a bug in the OpenOffice software, please re-send your
   question to the ooo-us...@incubator.apache.org mailing list, or look
   for an answer on the user forums at http://user.services.openoffice.org/.

  - For confidential matters ONLY, you may follow-up to the privately-archived
   ooo-priv...@incubator.apache.org  mailing list.  Normal bug reports are not
   confidential and should be directed to ooo-us...@incubator.apache.org."


IMHO the "confidential", "only", "private" etc. wording should be 
deleted resp. replaced with a "ooo-dev@" fitting formulation.



Does the user want private/confidential help?  Of course they do.
Identity theft, spam, etc.  Who wouldn't want this?

IMHO, there is no good reason to mention ooo-private in the context of
support.  This will only draw a percentage of users to send support
questions to the private list.  In the very rare case where there is a
security-related issue, where confidentiality is warranted,  the user
would be clueful enough to find the security page on the website and
submit the report that way.


When there is a case that private mails are necessary then we should 
treat the user individually. That's much better than to point all users 
generlly to our private ML.


So, +1 for changing this.

Marcus

Re: [PROPOSAL][WWW] style and content changes to home page

[Marcus (OOo)]

Am 10/15/2012 12:29 AM, schrieb Rob Weir:

On Sun, Oct 14, 2012 at 6:15 PM, Marcus (OOo)  wrote:

Am 10/14/2012 05:56 PM, schrieb Rob Weir:


On Sun, Oct 14, 2012 at 11:42 AM, Marcus (OOo)
wrote:


Am 10/14/2012 05:17 PM, schrieb Rob Weir:


On Sun, Oct 14, 2012 at 10:45 AM, Marcus (OOo)
wrote:



Am 10/14/2012 04:10 PM, schrieb Rob Weir:


On Sun, Oct 14, 2012 at 8:52 AM, Marcus (OOo)
wrote:




Am 10/10/2012 09:08 AM, schrieb Andrea Pescetti:


On 09/10/2012 Kay Schenk wrote:





http://www.openoffice.org/test/ ...
I am invoking *lazy consensus* on these changes and put this in
place
sometime on Sat, PDT -- say 15:30, unless there are objections.






It's nice indeed. I only see the "Valid XHTML" icon positioned a bit
too
high maybe... Is it wanted?
http://people.apache.org/~pescetti/tmp/ooo-www-test.png

And, by the way, clicking on it reveals that there are a couple of
markup fixes to apply, but I don't know if those are due to the CMS
or
to specific markup of the page.






Currenty it's 1 warning and 1 error. The warning comes because the
validator
uses a new HTML 5 checker which is still in Beta status. IMHO it's
irrelevant.

The error is due to the "PUBLISHER" tag in the link reference (line
8).

Due to the following webpage "PUBLISHER" is no valid HTML style.
However
I
wouldn't change it as it seems to be used for Google index
referencing:



If you make it lower case "publisher" it should be OK.

-Rob





http://www.thoughtsfromgeeks.com/resources/2793-Rel-publisher-standard-HTML-markup-or.aspx

Marcus





I've made the change but this doesn't make a difference, see:

http://www.w3.org/TR/html401/types.html#type-links



Look at the detailed error message here:


http://validator.w3.org/check?uri=http%3a%2f%2fwww.openoffice.org%2ftest%2f

It looks like the W3C Validator looks at more than the values in the
HTML specification.  They also look at the Microformats Wiki:



http://microformats.org/wiki/existing-rel-values#HTML5_link_type_extensions

"publisher" is listed there.

Of course, that is what the error message says.  I have no idea if the
Validator actually works that way ;-)




For me the Wiki says "do not use 'publisher', it's no longer valid HTML
4.x
style":




Maybe you are not seeing what I am seeing.

The W3C Validator says:

"Syntax of link type valid for:
  A whitespace-separated list of link types listed as allowed on
   in the HTML specification or listed as an allowed on   on
the Microformats wiki without duplicate keywords in the list. You can
register link types on the Microformats wiki yourself."

It links to this Microformats wiki page:


http://microformats.org/wiki/existing-rel-values#HTML5_link_type_extensions

It says there:

" HTML5 link type extensions

The following values are registered as link type extensions per the
requirements in the WHATWG HTML spec and the requirements in the W3C
HTML5 spec. "

And in that table "publisher" is defined.



Yes, but as dropped. And for HTML 5 just proposed and not yet accepted.



No.  Look carefully.  There are two entries.  The one in the "dropped"
tabled is for an older meaning of "publisher".  But look again at the
first table.  "publisher" is still there and references the Google
definition.  So they dropped the old definition and added the a new


yes, but it's just "proposed". That means it's not yet valid.


one.  Net result is the error goes away if we just change the
attribute value to all lowercase.


As you can see with the staged version of the index page the error is 
still there. ;-)



rel value | summary | defining specification | why dropped
---
publisher | identifies a hypertext link to a publisher | HTML4dropped |
unknown

However, it could come back in HTML 5 as it's already proposed:

Keyword | Effect on link | Effect on a, area | Brief description | Link
to
specification | Synonyms | Status
---
publisher | External Resource | Contextual External Resource |
indicate[s]
that the destination of that hyperlink is a metadata profile (e.g. a
social
/ real name profile like Google+) for the current page or portion
thereof. |
rel-publisher | proposed

And IMHO the validator recognizes this already.

But when deleting it from our webpage I can imagine what would happen.
;-),
so we should leave all as it is for the moment.



The question is whether we want to declare the page as HTML4, XHTML4
or HTML5.  Right now we don't declare anything specific.  So the



The page is already declared, as "XHTML 1.0 Strict", see the first line in
the source file:

http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd";>

But it seems to be deleted when it's staged and publi

Re: [PROPOSAL][WWW] style and content changes to home page

[Marcus (OOo)]

Am 10/14/2012 05:56 PM, schrieb Rob Weir:

On Sun, Oct 14, 2012 at 11:42 AM, Marcus (OOo)  wrote:

Am 10/14/2012 05:17 PM, schrieb Rob Weir:


On Sun, Oct 14, 2012 at 10:45 AM, Marcus (OOo)
wrote:


Am 10/14/2012 04:10 PM, schrieb Rob Weir:


On Sun, Oct 14, 2012 at 8:52 AM, Marcus (OOo)
wrote:



Am 10/10/2012 09:08 AM, schrieb Andrea Pescetti:


On 09/10/2012 Kay Schenk wrote:




http://www.openoffice.org/test/ ...
I am invoking *lazy consensus* on these changes and put this in place
sometime on Sat, PDT -- say 15:30, unless there are objections.





It's nice indeed. I only see the "Valid XHTML" icon positioned a bit
too
high maybe... Is it wanted?
http://people.apache.org/~pescetti/tmp/ooo-www-test.png

And, by the way, clicking on it reveals that there are a couple of
markup fixes to apply, but I don't know if those are due to the CMS or
to specific markup of the page.





Currenty it's 1 warning and 1 error. The warning comes because the
validator
uses a new HTML 5 checker which is still in Beta status. IMHO it's
irrelevant.

The error is due to the "PUBLISHER" tag in the link reference (line 8).

Due to the following webpage "PUBLISHER" is no valid HTML style.
However
I
wouldn't change it as it seems to be used for Google index referencing:



If you make it lower case "publisher" it should be OK.

-Rob




http://www.thoughtsfromgeeks.com/resources/2793-Rel-publisher-standard-HTML-markup-or.aspx

Marcus




I've made the change but this doesn't make a difference, see:

http://www.w3.org/TR/html401/types.html#type-links



Look at the detailed error message here:

http://validator.w3.org/check?uri=http%3a%2f%2fwww.openoffice.org%2ftest%2f

It looks like the W3C Validator looks at more than the values in the
HTML specification.  They also look at the Microformats Wiki:


http://microformats.org/wiki/existing-rel-values#HTML5_link_type_extensions

"publisher" is listed there.

Of course, that is what the error message says.  I have no idea if the
Validator actually works that way ;-)



For me the Wiki says "do not use 'publisher', it's no longer valid HTML 4.x
style":



Maybe you are not seeing what I am seeing.

The W3C Validator says:

"Syntax of link type valid for:
 A whitespace-separated list of link types listed as allowed on
  in the HTML specification or listed as an allowed on  on
the Microformats wiki without duplicate keywords in the list. You can
register link types on the Microformats wiki yourself."

It links to this Microformats wiki page:

http://microformats.org/wiki/existing-rel-values#HTML5_link_type_extensions

It says there:

" HTML5 link type extensions

The following values are registered as link type extensions per the
requirements in the WHATWG HTML spec and the requirements in the W3C
HTML5 spec. "

And in that table "publisher" is defined.


Yes, but as dropped. And for HTML 5 just proposed and not yet accepted.


rel value | summary | defining specification | why dropped
---
publisher | identifies a hypertext link to a publisher | HTML4dropped |
unknown

However, it could come back in HTML 5 as it's already proposed:

Keyword | Effect on link | Effect on a, area | Brief description | Link to
specification | Synonyms | Status
---
publisher | External Resource | Contextual External Resource | indicate[s]
that the destination of that hyperlink is a metadata profile (e.g. a social
/ real name profile like Google+) for the current page or portion thereof. |
rel-publisher | proposed

And IMHO the validator recognizes this already.

But when deleting it from our webpage I can imagine what would happen. ;-),
so we should leave all as it is for the moment.



The question is whether we want to declare the page as HTML4, XHTML4
or HTML5.  Right now we don't declare anything specific.  So the


The page is already declared, as "XHTML 1.0 Strict", see the first line 
in the source file:


"http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd";>


But it seems to be deleted when it's staged and published, it's just the 
following:





Validator assumes we're HTML5 and uses those rules.  If we want to be
validated as HTML 4.01 Transitional then we should declare that
doctype.


Or investigate and fix whats going wrong in staging and publishing. ;-)


But honestly, the website is all over the place, with a mix of
markups.  I don't know if it really makes sense to have the W3C Valid
HTML on the home page, since we cannot claim this even for that single
page.  Maybe we should just remove it?


Deleting because we cannot fix it? Hm.

Marcus

Re: [PROPOSAL][WWW] style and content changes to home page

[Marcus (OOo)]

Am 10/14/2012 05:58 PM, schrieb Kay Schenk:

On Sun, Oct 14, 2012 at 5:53 AM, Marcus (OOo)  wrote:


Am 10/10/2012 05:29 PM, schrieb Kay Schenk:


On Wed, Oct 10, 2012 at 12:08 AM, Andrea Pescetti**
wrote:

  On 09/10/2012 Kay Schenk wrote:


  http://www.openoffice.org/test/<http://www.openoffice.org/**test/>

<http://www.openoffice.**org/test/<http://www.openoffice.org/test/>>
  ...


I am  invoking *lazy consensus* on these changes and put this in place
sometime on Sat, PDT -- say 15:30, unless there are objections.



It's nice indeed. I only see the "Valid XHTML" icon positioned a bit too
high maybe... Is it wanted?
http://people.apache.org/~pescetti/tmp/ooo-www-test.png<http://people.apache.org/~**pescetti/tmp/ooo-www-test.png>
<**http://people.apache.org/~**pescetti/tmp/ooo-www-test.png<http://people.apache.org/~pescetti/tmp/ooo-www-test.png>





And, by the way, clicking on it reveals that there are a couple of markup
fixes to apply, but I don't know if those are due to the CMS or to
specific
markup of the page.



hmmm...thanks for the feedback. Yes, the positioning can be changed and I
will look at the fixes it tags.



I've changed the positioning of the W3C validator logo to show it left
aligned. IMHO it's looking better than stuck to the rightmost. I hope you
don't mind.

Marcus



Well I kind of like it off to the right better, but this isn't a big issue
for me.

As for the validator element itself...I would be in favor of removing it
entirely. It's only on the home page so what's the point unless we want to
add it to the footer area, so it would display on ALL pages.  Ah! I just


Good point. Moving it into the footer would be great.

Marcus




saw Rob's response! I'm all for removing this little graphic entirely as
well.

Yes, I too saw the "publisher" error, but decided to just leave it alone.

Re: [PROPOSAL][WWW] style and content changes to home page

[Marcus (OOo)]

Am 10/14/2012 05:17 PM, schrieb Rob Weir:

On Sun, Oct 14, 2012 at 10:45 AM, Marcus (OOo)  wrote:

Am 10/14/2012 04:10 PM, schrieb Rob Weir:


On Sun, Oct 14, 2012 at 8:52 AM, Marcus (OOo)
wrote:


Am 10/10/2012 09:08 AM, schrieb Andrea Pescetti:


On 09/10/2012 Kay Schenk wrote:



http://www.openoffice.org/test/ ...
I am invoking *lazy consensus* on these changes and put this in place
sometime on Sat, PDT -- say 15:30, unless there are objections.




It's nice indeed. I only see the "Valid XHTML" icon positioned a bit too
high maybe... Is it wanted?
http://people.apache.org/~pescetti/tmp/ooo-www-test.png

And, by the way, clicking on it reveals that there are a couple of
markup fixes to apply, but I don't know if those are due to the CMS or
to specific markup of the page.




Currenty it's 1 warning and 1 error. The warning comes because the
validator
uses a new HTML 5 checker which is still in Beta status. IMHO it's
irrelevant.

The error is due to the "PUBLISHER" tag in the link reference (line 8).

Due to the following webpage "PUBLISHER" is no valid HTML style. However
I
wouldn't change it as it seems to be used for Google index referencing:



If you make it lower case "publisher" it should be OK.

-Rob



http://www.thoughtsfromgeeks.com/resources/2793-Rel-publisher-standard-HTML-markup-or.aspx

Marcus



I've made the change but this doesn't make a difference, see:

http://www.w3.org/TR/html401/types.html#type-links



Look at the detailed error message here:
http://validator.w3.org/check?uri=http%3a%2f%2fwww.openoffice.org%2ftest%2f

It looks like the W3C Validator looks at more than the values in the
HTML specification.  They also look at the Microformats Wiki:

http://microformats.org/wiki/existing-rel-values#HTML5_link_type_extensions

"publisher" is listed there.

Of course, that is what the error message says.  I have no idea if the
Validator actually works that way ;-)


For me the Wiki says "do not use 'publisher', it's no longer valid HTML 
4.x style":


rel value | summary | defining specification | why dropped
---
publisher | identifies a hypertext link to a publisher | HTML4dropped | 
unknown


However, it could come back in HTML 5 as it's already proposed:

Keyword | Effect on link | Effect on a, area | Brief description | Link 
to specification | Synonyms | Status

---
publisher | External Resource | Contextual External Resource | 
indicate[s] that the destination of that hyperlink is a metadata profile 
(e.g. a social / real name profile like Google+) for the current page or 
portion thereof. | rel-publisher | proposed


And IMHO the validator recognizes this already.

But when deleting it from our webpage I can imagine what would happen. 
;-), so we should leave all as it is for the moment.


Marcus

Re: [PROPOSAL][WWW] style and content changes to home page

[Marcus (OOo)]

Am 10/14/2012 04:10 PM, schrieb Rob Weir:

On Sun, Oct 14, 2012 at 8:52 AM, Marcus (OOo)  wrote:

Am 10/10/2012 09:08 AM, schrieb Andrea Pescetti:


On 09/10/2012 Kay Schenk wrote:


http://www.openoffice.org/test/ ...
I am invoking *lazy consensus* on these changes and put this in place
sometime on Sat, PDT -- say 15:30, unless there are objections.



It's nice indeed. I only see the "Valid XHTML" icon positioned a bit too
high maybe... Is it wanted?
http://people.apache.org/~pescetti/tmp/ooo-www-test.png

And, by the way, clicking on it reveals that there are a couple of
markup fixes to apply, but I don't know if those are due to the CMS or
to specific markup of the page.



Currenty it's 1 warning and 1 error. The warning comes because the validator
uses a new HTML 5 checker which is still in Beta status. IMHO it's
irrelevant.

The error is due to the "PUBLISHER" tag in the link reference (line 8).

Due to the following webpage "PUBLISHER" is no valid HTML style. However I
wouldn't change it as it seems to be used for Google index referencing:



If you make it lower case "publisher" it should be OK.

-Rob


http://www.thoughtsfromgeeks.com/resources/2793-Rel-publisher-standard-HTML-markup-or.aspx

Marcus


I've made the change but this doesn't make a difference, see:

http://www.w3.org/TR/html401/types.html#type-links

Marcus

Re: [PROPOSAL][WWW] style and content changes to home page

[Marcus (OOo)]

Am 10/10/2012 05:29 PM, schrieb Kay Schenk:

On Wed, Oct 10, 2012 at 12:08 AM, Andrea Pescettiwrote:


On 09/10/2012 Kay Schenk wrote:


http://www.openoffice.org/**test/  ...

I am  invoking *lazy consensus* on these changes and put this in place
sometime on Sat, PDT -- say 15:30, unless there are objections.



It's nice indeed. I only see the "Valid XHTML" icon positioned a bit too
high maybe... Is it wanted?
http://people.apache.org/~**pescetti/tmp/ooo-www-test.png

And, by the way, clicking on it reveals that there are a couple of markup
fixes to apply, but I don't know if those are due to the CMS or to specific
markup of the page.



hmmm...thanks for the feedback. Yes, the positioning can be changed and I
will look at the fixes it tags.


I've changed the positioning of the W3C validator logo to show it left 
aligned. IMHO it's looking better than stuck to the rightmost. I hope 
you don't mind.


Marcus

Re: [PROPOSAL][WWW] style and content changes to home page

[Marcus (OOo)]

Am 10/10/2012 09:08 AM, schrieb Andrea Pescetti:

On 09/10/2012 Kay Schenk wrote:

http://www.openoffice.org/test/ ...
I am invoking *lazy consensus* on these changes and put this in place
sometime on Sat, PDT -- say 15:30, unless there are objections.


It's nice indeed. I only see the "Valid XHTML" icon positioned a bit too
high maybe... Is it wanted?
http://people.apache.org/~pescetti/tmp/ooo-www-test.png

And, by the way, clicking on it reveals that there are a couple of
markup fixes to apply, but I don't know if those are due to the CMS or
to specific markup of the page.


Currenty it's 1 warning and 1 error. The warning comes because the 
validator uses a new HTML 5 checker which is still in Beta status. IMHO 
it's irrelevant.


The error is due to the "PUBLISHER" tag in the link reference (line 8).

Due to the following webpage "PUBLISHER" is no valid HTML style. However 
I wouldn't change it as it seems to be used for Google index referencing:


http://www.thoughtsfromgeeks.com/resources/2793-Rel-publisher-standard-HTML-markup-or.aspx

Marcus

Re: Consultants Directory: Update and Help Needed

[Marcus (OOo)]

Am 10/11/2012 12:11 AM, schrieb Rob Weir:

So here's where we are with the consultants listing:

1) XLST script, data file, strings for en and it are checked in under
ooo-site/content/bizdev/consultants

2) Announcement to ooo-dev/ooo-users/forums/twitter/google+/facebook
(but not ooo-announce since that is almost all end-users)

3) Blog post

4) I also went through the legacy consultant listings and reviewed the
websites of the English-speaking entities.  Where one appeared to
still be relevant I sent them an email and pointed them to the new
listing instructions and invited them to submit a new listing.  This
was a judgement call, but generally if the website was still up and I
found an explicit mention of OpenOffice services then I considered it
relevant.  But if it was just a generic IT services company with no
mention of OpenOffice then I did not send an invite. Of course,
nothing prevents them from submitting a listing, and if they do we
review it like any other.

5) What we have so far is here:
http://www.openoffice.org/bizdev/consultants.html

Note, there are more coming.  One was posted to the list today that I
need to check in.  And there are a few more were I was emailed that
they were interested and would submit a listing.

So areas where I could use help:

1) Reviewing the non-English legacy consultants for relevant ones who
should be invited.  German, French, Italian and Spanish volunteers are
especially needed.  If you can help I can send you a spreadsheet of
the websites and an English version of my invitation note.  Maybe 30
or so per language.

2 (less urgent) At some point we might enable the automatic generation
of the HTML from the XML + XSLT.  Since I don't speak Perl I'm going
to need someone else to tackle this.

Thanks!

-Rob


I've committet a localized version for DE.

Marcus

Re: [UX]: potential new default color for application background

[Marcus (OOo)]

Am 10/12/2012 10:34 AM, schrieb Jürgen Schmidt:

Hi,

I am seeking for some small changes to give our product a little fresh
look. Changes that are more low cost and don't require huge
implementation efforts.

For example all the grey colors are not so nice and I tried a new color
for the application background. I would like to propose that we change
the default color to a new fresh one. If our users don't like and prefer
the old grey one they can easy revert it via the option dialog.

The question is about the start center because the changed application
background has no impact on the start center yet. But I think this is a bug.


Or maybe not. The dialog option talked about the *Application 
background*. As the start center is not part of any application module, 
it has to be seen separated.


But I'm with you, it should be configurable via this dialog option, too.

Marcus




Examples:
http://people.apache.org/~jsc/test/aoo_new_defautl_background.png
http://people.apache.org/~jsc/test/aoo_new_defautl_background2.png

I know that probably many have an opinion about the color. But more
important for now is the question if we want such a change in general.
In the end the color should fit to other branding items as well.

UX has probably some comments and feedback here as well?

Juergen

Re: [RELEASE] milestone build (Was: [RELEASE] 3.5, 4.0, fixpack, milestone build...)

[Marcus (OOo)]

Am 10/09/2012 03:58 AM, schrieb Shenfeng Liu:

2012/10/9 Ariel Constenla-Haile


Hi Jürgen, *

On Mon, Oct 08, 2012 at 04:58:03PM +0200, Jürgen Schmidt wrote:

The build bots are still not build the same as we do for the binary
releases (please correct me if I am wrong). Means as long as we don't
have build bots which are building with the same configuration we should
provide the builds manually in the same way we did it for the release.

@Ariel, would that be ok for you fro now until we have a better solution?


Yes, I will apply the set up described in
https://issues.apache.org/ooo/show_bug.cgi?id=119385 , that is,
decreasing Linux system requirements to glibc 2.5

Any one is welcome to take any of the two architectures (building on
Linux is multiplied by 4: rpm/deb, 32 and 64 bits; this counts on
building time and uploading the packages); if not, I will take care of
both.



I will take care of Windows and MacOS.



  (2) How many language support can we get for this milestone build? Not
necessary to be 100% translated, but can be a base for volunteers to

verify

the translation.


We should include the languages that we have released and add all
languages where we notice active volunteers who help us to support these
further languages (eg. Polish, Danish, Scots Gaelic, ...)



  (3) The current development snapshot naming [a] is a little confusing

to

me. I wonder if we can change the naming to reflect the date of the

build?


I am not sure if understand you correct. The revision is a unique
identifier and makes it clear what went in the snapshot. We probably
upload the builds not all on the same day. Means I am not sure how a
date can help here.


I guess that besides the revision, milestone builds can be identified by
their milestone number, which should be increased in every milestone
build: AOO350m1 AOO350m2 etc

just like in OOo times there was DEV300m105 DEV300m106 etc
http://www.openoffice.org/development/releases/DEV300m106_snapshot.html
it could start now from DEV350m1



OK, I understand the revision now, and let's forget the "date".
And I agree with Ariel that a milestone number like AOO350m1 will be better
when we promote it.
I personally do not think we need to use mirror. But a download page that
Marcus suggested will be good.


Sure, the download page can point to the builds on the mirror system or 
the ASF people's directories (when the paths are unified then automatism 
is much easier).


But when using the mirrors we could:
- stear the timeframe how long a milestone should be online,
- when to release the next dev build,
- a simple point of downloadable dev builds,
- and of course we can see how often which file was downloaded. To see 
if it's worth the efforts at all.


So, I think we should try to distribute the dev builds via the mirror 
system. If we laster think that it doesn't make sense anymore then we 
can stop it.


Marcus

Re: [CHECKSUMS MAC OS] Need help to check and fix the instructions for the ASC and KEYS hashes on Mac OS

[Marcus (OOo)]

Am 10/09/2012 10:25 AM, schrieb Jürgen Schmidt:

On 10/8/12 8:59 PM, Marcus (OOo) wrote:

Am 10/08/2012 11:05 AM, schrieb Jürgen Schmidt:

On 10/5/12 9:23 PM, Marcus (OOo) wrote:

Am 10/05/2012 09:57 AM, schrieb Jürgen Schmidt:

On 10/4/12 10:33 PM, Marcus (OOo) wrote:

Hi Mac fans,

as I've no Mac at hand please can someone help me with verifing and
fixing the checksum instructions:

http://www.openoffice.org/download/checksums/3.4.1_checksums.html#howto


"This is how you verify with ASC and KEYS hashes on Mac OS"

A user has reported a problem with the following line:

KEYID="0x`...

(see IZ 121159 for reference)

Thanks in advance


why so complicate? I am no gpg guru but I thought it should be
enough to


I simply collected what I've found in the Internet.


1. import the official KEYS as you have described
2. gpg --verify
Apache_OpenOffice_incubating_3.4.1_MacOS_x86_install_en-GB.dmg.asc
Apache_OpenOffice_incubating_3.4.1_MacOS_x86_install_en-GB.dmg

Result if it's ok:

gpg: Signature made Mon Aug 13 15:47:11 2012 CEST using RSA key ID
ABABABAB
gpg: Good signature from "Juergen Schmidt"
gpg: aka "Juergen Schmidt"
gpg: aka "Juergen Schmidt"

Result if it's a bad signature:

gpg: Signature made Mon Aug 13 15:47:11 2012 CEST using RSA key ID
ABABABAB
gpg: BAD signature from "Juergen Schmidt"


Interesting, at least on Linux I've to do all the steps listed in the
Linux section. So, you don't have to do this on Mac OS? A simple "gpg
--verify KEYS" is enough to get a "Good ..." or "Bad ..." result?


I am not sure if I understand you here, I verified the *.asc file with
the original file.


So, the question is, how? Please be as detailed possible, I've no chance
to reproduce. ;-)


I would expect that this works on all platforms more
or less in the same way.


OK, to make it simple. The following commands do not work on Linux:

$ wget http://www.apache.org/dist/incubator/ooo/KEYS
$ gpg --import KEYS
$ gpg --verify install_binary.tar.gz.asc install_binary.tar.gz

Does this work on MacOS X? If not please help me to get it working.



the problem is that the file
http://www.apache.org/dist/incubator/ooo/KEYS contains only 2 keys,
Rob's and my key.


oh right, this cannot work on Linux as these builds are not from you or 
Rob but from Ariel.



Please try it with https://people.apache.org/keys/group/ooo.asc

The ooo.asc file will updated automatically via id.apache.org if people
maintain their data there as far as I know


Great. Now I get the "Good signature ..." result also on Linux.

Thanks a lot Jürgen. :-)

Marcus

Re: [RESULT][VOTE][PMC] PMC Chair

[Marcus (OOo)]

Am 10/09/2012 12:27 AM, schrieb Andrew Rist:

The Vote started on 2012-10-04 has ended.

Results:
Andrea Pescetti (pescetti) - 28 (24 PPMC + 2 Committers + 2 Community)
Drew Jensen (atjensen) - 3 (3 PPMC)

There were no other votes cast.

The PPMC has approved Andrea Pescetti for PMC Chair Apache OpenOffice PMC
This can now be added to the TLP resolution for AOO.


congratulations to you, Andrea. :-)

Marcus

Re: [RELEASE] milestone build (Was: [RELEASE] 3.5, 4.0, fixpack, milestone build...)

[Marcus (OOo)]

Am 10/08/2012 04:58 PM, schrieb Jürgen Schmidt:

On 10/8/12 3:36 PM, Shenfeng Liu wrote:

Hi, all,
   I'm just back from China Mid-Autumn/National-Day holidays. So I'd like to
pick up the milestone build topic again.


thanks for picking it up again, we should definitely start with snapshot
builds asap


   We already get the support from QE team for the test plan. So I think we
need to do the following:

1. build out the milestone build. I saw from the buildbot that we have the
Windows and Linux-64 build successfully today. So it can be the candidate
of our first milestone build. While my question are:
  (1) Can we add more platforms (e.g. Mac, Linux-32, OS2...)? They need to
be built manually.


The build bots are still not build the same as we do for the binary
releases (please correct me if I am wrong). Means as long as we don't
have build bots which are building with the same configuration we should
provide the builds manually in the same way we did it for the release.

@Ariel, would that be ok for you fro now until we have a better solution?

I will take care of Windows and MacOS.



  (2) How many language support can we get for this milestone build? Not
necessary to be 100% translated, but can be a base for volunteers to verify
the translation.


We should include the languages that we have released and add all
languages where we notice active volunteers who help us to support these
further languages (eg. Polish, Danish, Scots Gaelic, ...)



  (3) The current development snapshot naming [a] is a little confusing to
me. I wonder if we can change the naming to reflect the date of the build?


I am not sure if understand you correct. The revision is a unique
identifier and makes it clear what went in the snapshot. We probably
upload the builds not all on the same day. Means I am not sure how a
date can help here.


I also wouldn't rely on a date value, even if it's looking right on the 
fisrt view. The revision number is what referenced everywhere (SVN, BZ). 
So, whenthe developer asks you "In which build have you seen this or 
that?" then you can tell her/him just the rev number.



2. upload the build and update the development snapshot wiki [a].

3. Run the testing against the build.

4. prepare related documents.
  (1) update the release notes wiki [b] for new values in this milestone
build.
  (2) a wiki page to record the testing result to this milestone build.


1 + 2 are good ideas to keep the info up-to-date and to reduce the work
later on short in front of the release.


  (2) a web page to announce this milestone build.


mmh, when we want to promote the dev builds more we should consider the
use of the mirrors. But this means we have to do more "release"
preparation. I am not sure if we can and want manage this additional
overhead at the moment. But I am open and we should check what's
necessary to achieve this.


For the normal testing IMHO we can rely on an en-US full build and 
langpacks only for all other languages.


For special issues (e.g., fixes for the installer and system 
integration) we could build special install files on demand.


BTW:
I think it's clear that - if we all agree that we want it - I can 
support this process with updating the download webpages to offer the 
dev builds here instead (or additionally) of the Wiki. ;-)


Marcus




   I can contribute to #4.

   Any thing else to add? Or any suggestion/comments?


- Simon


[a]
https://cwiki.apache.org/confluence/display/OOOUSERS/Development+Snapshot+Builds
[b]
https://cwiki.apache.org/confluence/display/OOOUSERS/AOO+3.5+Release+Notes



2012/9/27 Shenfeng Liu


Xue Fei,
   I think BVT + fidelity automation will be good.
   And since we are still facing the buildbots limitation, we can start
from only some of the platforms and expand to more gradually.
   Thanks!

- Simon



2012/9/27 Xue Fei Duan


For milestone build, I think BVT + fidelity automation(load/save some
samples) running is ok. we needn't have extra test plan on it, how do you
think about it?
-Xue Fei

On Wed, Sep 26, 2012 at 6:50 PM, Shenfeng Liu  wrote:


2012/9/26 Ji Yan


Simon,

   IMO, milestone test plan should base on milestone schedule and

feature

plan, what feature goes in and which defects are fixed in this

milestone.

Based on this scope we can define our testing plan. Otherwise we are
running into target unclear testing, defects will be missed.



Ji,
   Thanks for your complete thought!
   While IMO, the milestone build is still a development snapshot. We

don't

need to ensure a kind of GA quality on it. And we just need to ensure

this

build:

1. Will not cause data lost (e.g. damage the file in editing).
2. Will not lead to operation system crash.
3. No severe defects that blocks user to try out the new features in

this

build.

   Of course we need to test the new features, but I think it should

fall to

another category of our planned testing. And for milestone build

testing, I

think an acceptance test should be able to cove

Re: [CHECKSUMS MAC OS] Need help to check and fix the instructions for the ASC and KEYS hashes on Mac OS

[Marcus (OOo)]

Am 10/08/2012 11:05 AM, schrieb Jürgen Schmidt:

On 10/5/12 9:23 PM, Marcus (OOo) wrote:

Am 10/05/2012 09:57 AM, schrieb Jürgen Schmidt:

On 10/4/12 10:33 PM, Marcus (OOo) wrote:

Hi Mac fans,

as I've no Mac at hand please can someone help me with verifing and
fixing the checksum instructions:

http://www.openoffice.org/download/checksums/3.4.1_checksums.html#howto

"This is how you verify with ASC and KEYS hashes on Mac OS"

A user has reported a problem with the following line:

KEYID="0x`...

(see IZ 121159 for reference)

Thanks in advance


why so complicate? I am no gpg guru but I thought it should be enough to


I simply collected what I've found in the Internet.


1. import the official KEYS as you have described
2. gpg --verify
Apache_OpenOffice_incubating_3.4.1_MacOS_x86_install_en-GB.dmg.asc
Apache_OpenOffice_incubating_3.4.1_MacOS_x86_install_en-GB.dmg

Result if it's ok:

gpg: Signature made Mon Aug 13 15:47:11 2012 CEST using RSA key ID
ABABABAB
gpg: Good signature from "Juergen Schmidt"
gpg: aka "Juergen Schmidt"
gpg: aka "Juergen Schmidt"

Result if it's a bad signature:

gpg: Signature made Mon Aug 13 15:47:11 2012 CEST using RSA key ID
ABABABAB
gpg: BAD signature from "Juergen Schmidt"


Interesting, at least on Linux I've to do all the steps listed in the
Linux section. So, you don't have to do this on Mac OS? A simple "gpg
--verify KEYS" is enough to get a "Good ..." or "Bad ..." result?


I am not sure if I understand you here, I verified the *.asc file with

> the original file.

So, the question is, how? Please be as detailed possible, I've no chance 
to reproduce. ;-)



I would expect that this works on all platforms more
or less in the same way.


OK, to make it simple. The following commands do not work on Linux:

$ wget http://www.apache.org/dist/incubator/ooo/KEYS
$ gpg --import KEYS
$ gpg --verify install_binary.tar.gz.asc install_binary.tar.gz

Does this work on MacOS X? If not please help me to get it working.

Thanks

Marcus

Re: [RESULT][VOTE][PMC] Starting Membership for Apache OpenOffice PMC

[Marcus (OOo)]

Am 10/06/2012 05:48 PM, schrieb Kay Schenk:

On Fri, Oct 5, 2012 at 3:30 PM, Andrew Rist  wrote:


The Vote started on 2012-10-01 has ended.

There were 28 +1 votes, and 1 abstention.

There were no other votes cast.

The PPMC has approved the Starting Membership for Apache OpenOffice PMC
This list can now be added to the TLP resolution for AOO.



The list has been added to:


https://cwiki.apache.org/confluence/display/OOOUSERS/Graduation+Resolution+%28draft%29

Please verify...


Thanks for updating the page, looks good.

Marcus




Voting Details:

+1 - In Favor
Dennis Hamilton
Albino Neto
Dave Fisher
TJ - Thomas J. Frazier
imacat (Yang Shih-Ching)
Carl Marcum
Graham Lauder
Donald Harbison
Pedro Giffuni
Andre Fischer
Armin Le Grand
Kazunari Hirano
Andrea Pescetti
Renizinger Zoltan
Oliver-Rainer Wittmann
Olaf Felka
Ian Lynch
RGB.ES - Ricardo
Peter Junge
Roberto Galoppini
Jürgen Schmidt
Yong Lin Ma
Regina Henschel
Kay Schenk
Marcus Lange
Andrew Rist
Dave Barton
Wolf Halton


0 - Abstain
Ariel Constenla-Haile

Re: [VOTE][PMC] PMC Chair

[Marcus (OOo)]

[X] Andrea Pescetti (pescetti)

Marcus



Am 10/05/2012 01:41 AM, schrieb Andrew Rist:

This is a call for vote on selecting the PMC Chair the Apache OpenOffice
PMC.
Per the IPMC's "Guide to Successful Graduation" [1] the Apache
OpenOffice podling needs to select a PMC to be listed in the TLP
resolution.


Everyone in the community is encouraged to vote. Votes from PPMC members
and Mentors are binding. This vote will run 72-hours.

The balloting will be until UTC midnight Sunday, 7 October:
2012-10-07T24:00Z.


[ ] Andrea Pescetti (pescetti)
[ ] Drew Jensen (atjensen)


The [DISCUSS] for this vote can be found at [2] and [3]. Note: Andrea
has accepted the nomination - Drew has not responded.


[1] http://incubator.apache.org/guides/graduation.html#toplevel
[2]http://mail-archives.apache.org/mod_mbox/incubator-ooo-dev/201210.mbox/%3c506a1e72.20...@oracle.com%3E

[3]http://mail-archives.apache.org/mod_mbox/incubator-ooo-dev/201209.mbox/%3c5064c859.9090...@oracle.com%3E

Re: [CHECKSUMS MAC OS] Need help to check and fix the instructions for the ASC and KEYS hashes on Mac OS

[Marcus (OOo)]

Am 10/05/2012 09:57 AM, schrieb Jürgen Schmidt:

On 10/4/12 10:33 PM, Marcus (OOo) wrote:

Hi Mac fans,

as I've no Mac at hand please can someone help me with verifing and
fixing the checksum instructions:

http://www.openoffice.org/download/checksums/3.4.1_checksums.html#howto

"This is how you verify with ASC and KEYS hashes on Mac OS"

A user has reported a problem with the following line:

KEYID="0x`...

(see IZ 121159 for reference)

Thanks in advance


why so complicate? I am no gpg guru but I thought it should be enough to


I simply collected what I've found in the Internet.


1. import the official KEYS as you have described
2. gpg --verify
Apache_OpenOffice_incubating_3.4.1_MacOS_x86_install_en-GB.dmg.asc
Apache_OpenOffice_incubating_3.4.1_MacOS_x86_install_en-GB.dmg

Result if it's ok:

gpg: Signature made Mon Aug 13 15:47:11 2012 CEST using RSA key ID ABABABAB
gpg: Good signature from "Juergen Schmidt"
gpg: aka "Juergen Schmidt"
gpg: aka "Juergen Schmidt"

Result if it's a bad signature:

gpg: Signature made Mon Aug 13 15:47:11 2012 CEST using RSA key ID ABABABAB
gpg: BAD signature from "Juergen Schmidt"


Interesting, at least on Linux I've to do all the steps listed in the 
Linux section. So, you don't have to do this on Mac OS? A simple "gpg 
--verify KEYS" is enough to get a "Good ..." or "Bad ..." result?


Thanks

Marcus

[CHECKSUMS MAC OS] Need help to check and fix the instructions for the ASC and KEYS hashes on Mac OS

[Marcus (OOo)]

Hi Mac fans,

as I've no Mac at hand please can someone help me with verifing and 
fixing the checksum instructions:


http://www.openoffice.org/download/checksums/3.4.1_checksums.html#howto

"This is how you verify with ASC and KEYS hashes on Mac OS"

A user has reported a problem with the following line:

KEYID="0x`...

(see IZ 121159 for reference)

Thanks in advance

Marcus

Re: [WWW] Do we need a new site logo (for graduation) -- call for volunteers

[Marcus (OOo)]

Am 10/04/2012 08:05 PM, schrieb Andrea Pescetti:

RGB ES wrote:

2012/10/4 Kay Schenk

We all love the orb with gulls, but maybe it's time for a slight
update to
our logo? Or the header line on our websites (project and user portal).

I do not like the idea of another logo change: IMO, the last one is too
recent. On the other hand, a reorganization of the header line sounds
interesting.


Same for me. It's probably early to update the logo, and the new one
does a good job at balancing continuity and innovation.

But it is a very good idea to rethink the site header, that is currently
missing some very common functionality, like a language selector to
access Native-Lang sites.


I fully support Andrea's opinion.

Marcus

Re: [DISCUSS] [PMC] Proposed PMC Chair

[Marcus (OOo)]

Also a +1 from me for Andrea as chair.

Marcus



Am 10/01/2012 11:11 AM, schrieb Oliver-Rainer Wittmann:

Hi,

On 27.09.2012 23:54, Andrew Rist wrote:

I would like to nominate Andrea Pescetti.
Andrea is one of the most active and well respected members of the
project, and
I believe he would make a great PMC Chair.



I am supporting the nomination of Andrea. Thus, +1 from my side.

Best regards, Oliver.





On 9/27/2012 2:42 PM, Andrew Rist wrote:

Moving right along...

I'd like to restart the discussion over choosing a PMC Chair. The
previous
proposed process was discussed in:
http://markmail.org/message/mrgnjtiuum5bovjd
I'd like to take up where that left off.

Here is the process (with minor modifications) from the original:

You can read about the duties of a PMC Chair here:
http://www.apache.org/dev/pmc.html#chair

1) Nominations would be open for 72 hours. Anyone can nominate
someone for the role. Self-nominations are fine. And of course
nominations can be declined.

2) If there is only one nomination, then we are done, provided there
are no sustained objections.

3) If there is more than one nomination we discuss on the list for
another 72 hours. Discussion would primarily be on ooo-dev, but some
subjects might be directed to ooo-private.

4) If after 72-hours discussion there are still two or more nominees
then we vote. Everyone would be welcome to vote, but binding votes
would be from PPMC members. If there are more than 2 candidates,
there will be a run-off vote between the top two nominees if none of
the nominees receive an outright majority.


(note: the last item was changed from the original - nothing more
complicated
than a two way run-off is needed, as it is unlikely we have more than
two
nominees.)
I think this process is straight forward enough that we should just
begin the
process now.

Andrew

Re: [VOTE] [PMC] Starting Membership for Apache OpenOffice PMC

[Marcus (OOo)]

+1 for the entire list.

Marcus



Am 10/02/2012 12:38 AM, schrieb Andrew Rist:

This is a call for vote on selecting the following list as the starting
membership for the Apache OpenOffice PMC, to be listed in the TLP
resolution. The voting is for the entire slate as listed.

Apache OpenOffice PMC Starting Membership:
Andre Fischer (af)
Andrea Pescetti (pescetti)
Andrew Rist (arist)
Ariel Constenla-Haile (arielch)
Armin Le Grand (alg)
Dave Fisher (wave)
Donald Harbison (dpharbison)
Drew Jensen (atjensen)
Ian Lynch (ingotian)
Jürgen Schmidt (jsc)
Kay Schenk (kschenk)
Kazunari Hirano (khirano)
Louis Suarez-Potts (louis)
Marcus Lange (marcus)
Oliver-Rainer Wittmann (orw)
Pedro Giffuni (pfg)
Peter Junge (pj)
Raphael Bircher (rbircher)
Regina Henschel (regina)
RGB.ES (rgb-es)
Roberto Galoppini (galoppini)
Yang Shih-Ching (imacat)
Yong Lin Ma (mayongl)


The balloting will be until UTC midnight Thursday,
4 October: 2012-10-04T24:00Z.

Approval requires a majority of +1 over -1 votes cast by members of the
PPMC.

[ ] +1 approve
[ ] 0 abstain
[ ] -1 disapprove, for the following reasons:


The [DISCUSS] for this vote was enthusiastically in favor. There were no
concerns expressed other than issues with the timeframe of discussions,
which were suitably extended. (note: All members of this list, except
for Drew and Raphael, accepted their nomination to this list. I have
left Drew and Raphael on the list as neither declined, and they still
have the ability to decline later)

Re: Did we ever reach consensus on support for Windows 2000

[Marcus (OOo)]

Am 10/02/2012 12:33 AM, schrieb Rob Weir:

On Mon, Oct 1, 2012 at 4:46 PM, Marcus (OOo)  wrote:

Am 10/01/2012 01:32 PM, schrieb Rob Weir:


A response to this thread in general.

Why has no action been taken on this yet?

I'm not trying to blame or accuse anyone, but I would like to figure
out why nothing has happened here yet.  This is one example of several
where issues seem to be discussed at length, but nothing ever happens.

1) Does someone want to fix this but does not feel empowered to do so?
   For example, is someone uncertain whether there is consensus, or is
concerned that if they do act that someone won't like it?

2) Does someone want to fix this but does not know how to do this from
a technical perspective, i.e., how to use the Apache CMS?

3) Does someone want to fix this but does not think that they have
sufficient permissions to do this?



4) Is there a (urgent) need to do the changes now or do we want to wait for
the next release to give a little sentence in the announcement?



Nothing is urgent right now, so it is an ideal time to do this.  When
we are ready for the next release there will be many urgent items to
attend to, making it easy to miss this.

In any case, I just made the changes.  It took 4 minutes to edit,
check-in, verify on staging and publish the HTML files.  I think it
took longer to write this note.


Sure, also this is a possibility. JFDI, right? ;-)

Marcus




Hm, I thought we wanted to wait. But I could be wrong.

Marcus





If someone wants to fix the website, but feels blocked by one of the
above, please speak up, either on the list, or privately.  IMHO none
of the above should be an issue.  We have techniques like "lazy
consensus" that can be used to see whether there is objection to a
given way forward.  And the Apache CMS is open for everyone to use,
whether or not you are a committer.

Regards,

-Rob

On Sun, Sep 16, 2012 at 4:26 PM, Keith N. McKenna
   wrote:


Greetings All;

I was going through FAQ's and other pages on the AOO (incubating) site
and
noticed that many still are showing that we support Windows 2000 as a
baseline operating system. I though I remembered some discussions a while
back on this list around that subject and thought we had decided that we
would no longer do that due to lack of testing resources.

I went back through the archives and did find a number of threads but
they
never seemed to reach a definite conclusion. I we are going to continue
to
support it all well and good, but if we cannot then all FAQ's and other
documentation on the site should change to reflect that.

Regards
Keith

Re: Did we ever reach consensus on support for Windows 2000

[Marcus (OOo)]

Am 10/01/2012 11:53 PM, schrieb Kay Schenk:



On 10/01/2012 01:46 PM, Marcus (OOo) wrote:

Am 10/01/2012 01:32 PM, schrieb Rob Weir:

A response to this thread in general.

Why has no action been taken on this yet?

I'm not trying to blame or accuse anyone, but I would like to figure
out why nothing has happened here yet. This is one example of several
where issues seem to be discussed at length, but nothing ever happens.

1) Does someone want to fix this but does not feel empowered to do so?
For example, is someone uncertain whether there is consensus, or is
concerned that if they do act that someone won't like it?

2) Does someone want to fix this but does not know how to do this from
a technical perspective, i.e., how to use the Apache CMS?

3) Does someone want to fix this but does not think that they have
sufficient permissions to do this?


4) Is there a (urgent) need to do the changes now or do we want to wait
for the next release to give a little sentence in the announcement?

Hm, I thought we wanted to wait. But I could be wrong.

Marcus


ditto Marcus...we just need to make sure this doesn't slip through the
cracks as it were.

I'll tag this for a future "to do" unless someone else makes these
changes to the System Requirements and Installation areas.


Good. I've also put this as kind of todo on the Wiki page for the AOO 
3.5 release planning:


https://cwiki.apache.org/confluence/display/OOOUSERS/AOO+3.5+Release+Planning


A few last words on "support". Despite the fact that Win 2000 accounts
for such a small portion of our downloads, this will not stop folks
using Windows 2000 to ask for support I would imagine. So, yes, we need
to change the System Requirements and Installation instructions, but
keep in mind that it's likely these folks will still be around on older
OSes needing "support".


I'm sure it will stop over the time. E.g., I haven't seen any 
requests/questions since, hm I don't know, about "how to use OpenOffice 
on Win 98 / Win NT / Win ME?".


So, let's see :-)

Marcus




If someone wants to fix the website, but feels blocked by one of the
above, please speak up, either on the list, or privately. IMHO none
of the above should be an issue. We have techniques like "lazy
consensus" that can be used to see whether there is objection to a
given way forward. And the Apache CMS is open for everyone to use,
whether or not you are a committer.

Regards,

-Rob

On Sun, Sep 16, 2012 at 4:26 PM, Keith N. McKenna
 wrote:

Greetings All;

I was going through FAQ's and other pages on the AOO (incubating)
site and
noticed that many still are showing that we support Windows 2000 as a
baseline operating system. I though I remembered some discussions a
while
back on this list around that subject and thought we had decided
that we
would no longer do that due to lack of testing resources.

I went back through the archives and did find a number of threads but
they
never seemed to reach a definite conclusion. I we are going to
continue to
support it all well and good, but if we cannot then all FAQ's and other
documentation on the site should change to reflect that.

Regards
Keith

Re: Did we ever reach consensus on support for Windows 2000

[Marcus (OOo)]

Am 10/01/2012 01:32 PM, schrieb Rob Weir:

A response to this thread in general.

Why has no action been taken on this yet?

I'm not trying to blame or accuse anyone, but I would like to figure
out why nothing has happened here yet.  This is one example of several
where issues seem to be discussed at length, but nothing ever happens.

1) Does someone want to fix this but does not feel empowered to do so?
  For example, is someone uncertain whether there is consensus, or is
concerned that if they do act that someone won't like it?

2) Does someone want to fix this but does not know how to do this from
a technical perspective, i.e., how to use the Apache CMS?

3) Does someone want to fix this but does not think that they have
sufficient permissions to do this?


4) Is there a (urgent) need to do the changes now or do we want to wait 
for the next release to give a little sentence in the announcement?


Hm, I thought we wanted to wait. But I could be wrong.

Marcus




If someone wants to fix the website, but feels blocked by one of the
above, please speak up, either on the list, or privately.  IMHO none
of the above should be an issue.  We have techniques like "lazy
consensus" that can be used to see whether there is objection to a
given way forward.  And the Apache CMS is open for everyone to use,
whether or not you are a committer.

Regards,

-Rob

On Sun, Sep 16, 2012 at 4:26 PM, Keith N. McKenna
  wrote:

Greetings All;

I was going through FAQ's and other pages on the AOO (incubating) site and
noticed that many still are showing that we support Windows 2000 as a
baseline operating system. I though I remembered some discussions a while
back on this list around that subject and thought we had decided that we
would no longer do that due to lack of testing resources.

I went back through the archives and did find a number of threads but they
never seemed to reach a definite conclusion. I we are going to continue to
support it all well and good, but if we cannot then all FAQ's and other
documentation on the site should change to reflect that.

Regards
Keith

Re: Did we ever reach consensus on support for Windows 2000

[Marcus (OOo)]

Am 09/29/2012 10:27 PM, schrieb Wolf Halton:

On Thu, Sep 27, 2012 at 5:37 PM, Marcus (OOo)  wrote:

Am 09/27/2012 05:03 AM, schrieb Wolf Halton:


I think it more feasible to edit the website than to test on win2k in any
meaningful way.



I understand it in this way, that you suggest to delete the Windows 2000
support from the webpages and therefore cancel any support somewhat
silently. Is that right?


We don't have to make it a secret.  We can announce on the support
page that we are not supporting 12-year-old operating systems that are
not supported by those operating-systems' own manufacturer.  It seems
to me that the developers' time is better served focusing on newer
features that work with more modern operating systems.  I do
appreciate that there is some population still using Win2k.
Wikipedia's web-server stats say in August 2012, 1.45% of their
visitors used Windows other than XP, Vista or 7.  They are not giving
an authoritative count of those users, but 1.45% of Wikipedia users is
probably a large number.  I have seen in my own web stats that the
number of pre-2001 Windows systems that announce themselves to the
web-server are less than 1%. For my own web-stats that number is
between 20 and 50 individual users.  All are precious, but the cost of
supporting small populations is high, even in a time-only pro bono
system like Apache OpenOffice.


Rob has posted the best arguments for this dicussion: numbers. ;-)

Due to the (not really suprising) low numbers I support the suggestion 
from Armin: drop everything below 1%.


Marcus




Are we claiming to support win98se? or winME?


No, IMHO this topic is already done and in the past.

Marcus





On Sep 24, 2012 7:06 AM, "Stuart
Swales"
wrote:


On 23/09/2012 23:51, Kay Schenk wrote:




On 09/16/2012 09:48 PM, Rob Weir wrote:


On Sep 16, 2012, at 11:38 PM, "Keith N. McKenna"
   wrote:


Rob Weir wrote:


On Sun, Sep 16, 2012 at 4:26 PM, Keith N. McKenna
   wrote:


Greetings All;

I was going through FAQ's and other pages on the AOO (incubating)
site and
noticed that many still are showing that we support Windows 2000 as
a
baseline operating system. I though I remembered some discussions a
while
back on this list around that subject and thought we had decided
that we
would no longer do that due to lack of testing resources.



IMHO, "support" is determined by what we do, not by what we say.  If
no one is testing with Windows 2000, then it is hard to say we
support
it.  And if Microsoft does not make Windows 2000 CD's available to
developers for testing, due to a lawsuit, then it is rather difficult
for anyone who wants to test.  Not impossible, but they would need to
get access to CD's or ISO images through unofficial means.



The major disagreement I have with this Rob is that we publish FAQ's
and installation documents on our official web site that lead people
to believe that Windows 2000 is supported.



Actually I don't think we disagree on this. At one point in time
(OpenOffice.org 3.3?) Windows 2000 was presumably tested and that is
why it is on the supported list. The fact that it remains on that list
is purely due to a kind of inertia: documentation in rest stays at
rest unless acted on by an outside force.

So I agree that the website is out of synch with reality here and that
this is suboptimal. Two easy ways to fix: someone volunteers to do
some minimal testing with Windows 2000 to confirm basic operations, or
we remove it from the supported list.

Of course even if removed it could come back once tested.


What does it say for us as a responsible project when we tell people
that despite what we clearly show as a minimum requirement to use our
software is really not what we meant. All that does is leave a bad
taste in the consumers mouth that they most likely will tell there
friends about. That to me is NOT the image we should project.



If you feel strongly about this then you could propose to change the
website and if their are no objections after 72 hours assume lazy
consensus and go ahead and make the changes.


Of course, we could have a dozen people say we *should* support
Windows 2000.  But should does not mean anything.  We really need to
find even a single person who says they *will* test with Windows 2000
and fix any problems that arise.  Until that happens we don't really
support Windows 2000 in any meaningful way.



That is all well and good Rob, but again that needs to be clear to
people and not come as a surprise. I personally do not care one way
or the other if 2000 is supported or not. My concern is with the
image that we project to our user base. I am not a software engineer
or coder so therefore not qualified to judge what is or is not
supportable withing the code. That is why I brought this to the
attention of the people that are qualified to get better information
to present to our users.


I went back through the archives and did

Re: Did we ever reach consensus on support for Windows 2000

[Marcus (OOo)]

Am 09/27/2012 05:11 AM, schrieb Fernando Cassia:

On Mon, Sep 24, 2012 at 8:06 AM, Stuart Swales
  wrote:

Also, moving to Visual Studio 2010 will likely kill off running on
Windows 2000 (and Windows XP prior to SP2). The Visual C++ run-time
library now uses the EncodePointer function which was introduced in XP SP2.


Oracle switched to VS2010 for some part of Java6, that´s why the
latest Java6 JRE updates won´t install on Win2K.

However, I managed to hammer the latest Java6u35+ or thereabouts just
by copying files over from a WinXP install over the Win2K previous
JRE6 install and guess what? it works it seems it´s part of the
installer that needs MSVC runtimes that can´t be run in Win2K...


Sure, there will be the one or other way to hack a Java version into an 
old Windows release. However, the most important part is:


Is that suitable and doable for an average user?

Especially for what you wrote abut I guess it's not. ;-(

Of course, IMHO.

Marcus

Re: Did we ever reach consensus on support for Windows 2000

[Marcus (OOo)]

Am 09/27/2012 05:03 AM, schrieb Wolf Halton:

I think it more feasible to edit the website than to test on win2k in any
meaningful way.


I understand it in this way, that you suggest to delete the Windows 2000 
support from the webpages and therefore cancel any support somewhat 
silently. Is that right?


> Are we claiming to support win98se? or winME?

No, IMHO this topic is already done and in the past.

Marcus




On Sep 24, 2012 7:06 AM, "Stuart Swales"
wrote:


On 23/09/2012 23:51, Kay Schenk wrote:



On 09/16/2012 09:48 PM, Rob Weir wrote:

On Sep 16, 2012, at 11:38 PM, "Keith N. McKenna"
  wrote:


Rob Weir wrote:

On Sun, Sep 16, 2012 at 4:26 PM, Keith N. McKenna
  wrote:

Greetings All;

I was going through FAQ's and other pages on the AOO (incubating)
site and
noticed that many still are showing that we support Windows 2000 as a
baseline operating system. I though I remembered some discussions a
while
back on this list around that subject and thought we had decided
that we
would no longer do that due to lack of testing resources.



IMHO, "support" is determined by what we do, not by what we say.  If
no one is testing with Windows 2000, then it is hard to say we support
it.  And if Microsoft does not make Windows 2000 CD's available to
developers for testing, due to a lawsuit, then it is rather difficult
for anyone who wants to test.  Not impossible, but they would need to
get access to CD's or ISO images through unofficial means.


The major disagreement I have with this Rob is that we publish FAQ's
and installation documents on our official web site that lead people
to believe that Windows 2000 is supported.


Actually I don't think we disagree on this. At one point in time
(OpenOffice.org 3.3?) Windows 2000 was presumably tested and that is
why it is on the supported list. The fact that it remains on that list
is purely due to a kind of inertia: documentation in rest stays at
rest unless acted on by an outside force.

So I agree that the website is out of synch with reality here and that
this is suboptimal. Two easy ways to fix: someone volunteers to do
some minimal testing with Windows 2000 to confirm basic operations, or
we remove it from the supported list.

Of course even if removed it could come back once tested.


What does it say for us as a responsible project when we tell people
that despite what we clearly show as a minimum requirement to use our
software is really not what we meant. All that does is leave a bad
taste in the consumers mouth that they most likely will tell there
friends about. That to me is NOT the image we should project.



If you feel strongly about this then you could propose to change the
website and if their are no objections after 72 hours assume lazy
consensus and go ahead and make the changes.


Of course, we could have a dozen people say we *should* support
Windows 2000.  But should does not mean anything.  We really need to
find even a single person who says they *will* test with Windows 2000
and fix any problems that arise.  Until that happens we don't really
support Windows 2000 in any meaningful way.


That is all well and good Rob, but again that needs to be clear to
people and not come as a surprise. I personally do not care one way
or the other if 2000 is supported or not. My concern is with the
image that we project to our user base. I am not a software engineer
or coder so therefore not qualified to judge what is or is not
supportable withing the code. That is why I brought this to the
attention of the people that are qualified to get better information
to present to our users.


I went back through the archives and did find a number of threads
but they

never seemed to reach a definite conclusion. I we are going to
continue to
support it all well and good, but if we cannot then all FAQ's and
other
documentation on the site should change to reflect that.



Support is not determined by consensus wishes.  It is determined by
someone actually doing it.


Again Rob that is all well and good, but why are we publishing to the
world that Windows 2000 is the minimum Windows OS environment that
our product can run in?


Do we have any evidence that users have successfully installed and
used AOO 3.4.x on Windows 2000?  If it works, we might just list it
"not a tested configuration, but some users report success.".  In
other words, between "tested and supported" and "known to be broken"
is a middle territory where it is "use at your own risk".


I really do not know if we do our not Rob. What I do know is that we
are telling users on our official web site that Windows 2000 is the
minimum Revision of the OS that our product will run on.

Regards
Keith


-Rob


Regards
Keith








At least the following web pages need some attention:

* http://www.openoffice.org/dev_docs/source/sys_reqs.html
(not sure of navigation to this one)
* http://www.openoffice.org/dev_docs/source/sys_reqs_aoo34.html
(linked from download)
* http://www.openoffice.org/download/common

Re: [proposed PMC] call for feedback: Are you willing to serve as a PMC member?

[Marcus (OOo)]

Am 09/26/2012 11:37 AM, schrieb Oliver-Rainer Wittmann:

Hi,

this is a call for feedback to the people who are (currently) on the
proposed PMC list.
Please provide feedback, if you are willing to serve as a PMC member.

The current (copied from initial post of thread [1]) proposed PMC list is:


Proposed Working List:

Andre Fischer (af)
Andrea Pescetti (pescetti)
Andrew Rist (arist)
Ariel Constenla-Haile (arielch)
Armin Le Grand (alg)
Dave Fisher (wave)
Donald Harbison (dpharbison)
Drew Jensen (atjensen)
Ian Lynch (ingotian)
Jürgen Schmidt (jsc)
Kay Schenk (kschenk)
Kazunari Hirano (khirano)
Louis Suarez-Potts (louis)
Marcus Lange (marcus)


I'm willing to go on and spent some of my spare time for this project to 
bring it forward.


Marcus




Oliver-Rainer Wittmann (orw)
Pedro Giffuni (pfg)
Peter Junge (pj)
Raphael Bircher (rbircher)
Regina Henschel (regina)
RGB.ES (rgb-es)
Roberto Galoppini (galoppini)
Yang Shih-Ching (imacat)
Yong Lin Ma (mayongl)



Jürgen already expressed his willingness in thread [1].

[1] http://markmail.org/message/erjdi3wbe4hsjhqx


Best regards, Oliver.

Re: [PMC] Proposed Initial PMC List and process

[Marcus (OOo)]

Am 09/25/2012 08:35 PM, schrieb Andrew Rist:

I would like to propose the following list as the initial PMC for Apache
OpenOffice, with a procedure for consolidating this list into a final list.

[...]


Thanks a lot for your effort to process the listed names. Looks like the 
diversity is given and a big step forward in direction of graduation.


Marcus

Re: OpenOffice 3.4.1

[Marcus (OOo)]

Am 09/25/2012 02:07 AM, schrieb Errol Raymond:

I am new to OpenOffice, and want to install the new version. I
currently have a imac-g5 ppc, and it is not supported on my
architecture. Is there anyway to get around this issue?


Currently we don't offer any builds for MacOS X on the PPC platform as 
it was the case with OOo 3.3.0 and older releases. And to be honest, I 
don't think that we will do it with newer releases. This platform is - 
from the today's point of view - too old and we have nobody here in the 
project that would support it. To support it means taking the source 
code, building the installable files, do detailed QA tests to verify its 
stability.



I would like to effectively install, and contribute to OpenOffice.


Finally, I can just offer you the most recent OOo legacy release (3.3.0 
and older) for the PPC platform and hope it's available for your 
favorite language:


http://www.openoffice.org/download/legacy/other.html

HTH

Marcus

Re: [RELEASE] 3.5, 4.0, fixpack, milestone build...

[Marcus (OOo)]

Am 09/24/2012 08:01 AM, schrieb Andrea Pescetti:

On 23/09/2012 Shenfeng Liu wrote:

3. Deliver milestone builds to harvest our development fruits. A
milestone
build is:
(a) a development snapshot that contains the features/enhancements
that implemented till now;
(b) passed regression test to ensure no severe defects;
(c) announced on a development wiki;
(d) with documents on the wiki for the list of features and bug fixes
in this milestone build (like a release notes).


Milestone builds are a good idea: they would not be the same as the
daily snapshots (which by the way are not yet available for all
platforms). I would add:

(e) available in all official supported languages, at least as langpack;

and make the builds available on a monthly basis, so they can be easily
identified and we can engage QA testers more effectively. This, of
course, unless people who provide the builds find it too time-consuming
to upload a build per month.


A clear +1 from me.

Marcus

Re: [RELEASE] 3.5, 4.0, fixpack, milestone build...

[Marcus (OOo)]

Am 09/24/2012 03:13 PM, schrieb Ariel Constenla-Haile:

On Mon, Sep 24, 2012 at 07:56:26AM -0400, Rob Weir wrote:

Maybe also start a "release notes" page on the wiki.  Whenever a new
feature or important bug fix is added to the trunk also add something
to the release notes.   If something can be show with a "before and
after" screen shot, include that.  This might be easier than waiting
until the end to prepare the release notes.


We should come back to the old days, where there was a clear separation
between RC and Dev. Snapshot, see ooo-site/content/development/releases/
for example
http://www.openoffice.org/development/releases/DEV300m106_snapshot.html

I also think we should bring the Dev. Snapshot page back home, I'll try
to make a design with a left side bar, à la
http://www.openoffice.org/product/ and http://www.openoffice.org/why/
so that the download page (the forth, orange item) points to the same
site.


Thanks for bringing up this first. ;-)

When we could agree on to deliver dev builds on a regular basis (e.g., 
every 4 months) then we should offer them also from the main download 
page with clear wording. Like it was done it the previous project.


Marcus

Re: [RELEASE] 3.5, 4.0, fixpack, milestone build...

[Marcus (OOo)]

Am 09/24/2012 08:01 AM, schrieb Andrea Pescetti:

On 23/09/2012 Shenfeng Liu wrote:

3. Deliver milestone builds to harvest our development fruits. A
milestone
build is:
(a) a development snapshot that contains the features/enhancements
that implemented till now;
(b) passed regression test to ensure no severe defects;
(c) announced on a development wiki;
(d) with documents on the wiki for the list of features and bug fixes
in this milestone build (like a release notes).


Milestone builds are a good idea: they would not be the same as the
daily snapshots (which by the way are not yet available for all
platforms). I would add:

(e) available in all official supported languages, at least as langpack;

and make the builds available on a monthly basis, so they can be easily
identified and we can engage QA testers more effectively. This, of
course, unless people who provide the builds find it too time-consuming
to upload a build per month.


A clear +1 from me.

Marcus

Re: LICENSE text in the user interface

[Marcus (OOo)]

Am 09/24/2012 08:17 PM, schrieb Rob Weir:

On Mon, Sep 24, 2012 at 1:10 PM, Ariel Constenla-Haile
  wrote:

Hi *,

The About dialog (menu Tools - Help - About OOo) on trunk has now
a button to display the README, LICENSE and NOTICE; you can try with the
following binaries:
Win: http://people.apache.org/~arielch/packages/r1389156/win/
Linux 64 bits:  http://ci.apache.org/projects/openoffice/



So a single button that launches all three in one window?


According to http://markmail.org/message/67kwkxcs5fubpua3 the LICENSE
has to fulfill some legal requirements and include the other licenses,
besides the ALv2. This results in a very large LICENSE file that cannot
be completely displayed in the UI inside a MultilineEdit control due to
a limitation in the tools String class, see STRING_MAXLEN in
tools/inc/tools/string.hxx

Possible solutions:

a) don't display the license at all, only the README and the NOTICE

b) display a stripped down version containing only the ALv2 and a last
paragraph telling to read the licenses from other components



The about box already has this statement:

"Copyright © 2012 Apache Software Foundation. All rights reserved."

It would probably be fine to say, "Applicable licenses and notices for
included 3rd party components can be found..." and then either have
the path, URL or a button to bring up the details.


c) simply add at the end of the text a note telling that the full text
can be found in the office installation

d) kill the tools String class, and use rtl::OUString everywhere
(someone will have to sit down and write the code for this)


Hm, depends on the effort and if it's really worth it.
However, only a core coder can realisitically estimate this. ;-)


e) ...<-- another options are welcome



e1) Launch them in OpenOffice, as (read-only) documents

e2) Launch via the default web browser


+1

In our Internet times nowadays this is IMHO the best solution. A browser 
window should be always opened.


Marcus




e3) Launch via default desktop handler for *.txt file extension





Regards
--
Ariel Constenla-Haile
La Plata, Argentina

Re: download page path

[Marcus (OOo)]

Am 09/21/2012 08:31 PM, schrieb Albino B Neto:

Hi.

On Fri, Sep 21, 2012 at 11:47 AM, Donald Whytock  wrote:

Is there a path to the download page that's based on the apache.org
domain?  My company's firewall system has decided openoffice.org is


http://ooo-site.staging.apache.org/download/index.html

As it's not in the test area or something else but the one that is 
publicly known, it should be OK to use it as alternative.


HTH


"Freeware and Shareware".

And yes, they say that like it's a bad thing.


There will always be people that think in black and white pattern and do 
not want to see whats in between. ;-(



Yes.

The university I study also blocked the access openoffice.org

I asked why blocked, and not answer. :(


Really? A university? I cannot believe it.

Marcus

Re: Apache OpenOffice Awarded Two InfoWorld Bossie (Best of Open Source Software) Awards

[Marcus (OOo)]

Am 09/19/2012 05:23 PM, schrieb Rob Weir:

On Wed, Sep 19, 2012 at 3:35 AM, Shenfeng Liu  wrote:

What a great news!
Congratulations!



I'm especially proud of the recognition the community received for the
OpenOffice Templates website.   This almost did not happen.  It
required a lot of effort, especially by SourceForge and Apache Infra,
to make this a reality.

Remember back when OpenOffice first came to Apache.  We had the OOo
code, and a bunch of legacy services hosted on Oracle (and other 3rd
party) servers:  website, wiki, forums, templates, extensions,
bugzilla, etc.   We could have just abandoned the services and focused
100% on the code.  We could have just kissed the legacy web services
good bye and let them die.  That would have been an easier choice.
This option was mentioned at the time.  It would have been less
effort.  It would have allowed us to make more progress on the code,
and faster.  Certainly that was the choice that could have been made,
and which was made by another related project.  IMHO, it is akin to an
army leaving its wounded behind in order to make a faster advance.

Personally, I'm glad we did not abandon these services.  Instead we
invested in them.  Yes, it took time.  Yes, it took energy.  Yes, it
diverted resources that could have been used to work on product
features.  If you recall there was much wailing and gnashing of teeth
in some circles, saying that we had delayed our OpenOffice release so
long that it would no longer be relevant, that we would never catch
up, that we would inevitably fail.

Well, guess what happened?  We released 3.4.0 and had phenomenal
download numbers.  And then we released 3.4.1 and the download numbers
are even better.And now we receive awards both for the product,
but also for one of the legacy web services.  Hopefully this puts the
doubts to rest.  The answer to an "either/or" question is sometimes
"both".


great summary, well done!

I think especially the Sourceforge team around Roberto can be proud of 
have been gone the way to keep the support for the extension & templates 
and finally to got were they are now.


Marcus




2012/9/19 Rob Weir


InfoWorld is out with their 2012 "Bossie" awards (Best of Open Source).

As far as I can see we are the only open source project that won two
awards!

In the "desktop applications" category we received an award for the
OpenOffice application:


http://www.infoworld.com/slideshow/64914/bossie-awards-2012-the-best-open-source-desktop-applications-202288#slide4

And in the same category we received another award for our Template
Repository:


http://www.infoworld.com/slideshow/64914/bossie-awards-2012-the-best-open-source-desktop-applications-202288#slide5

The full article is here:


http://www.infoworld.com/d/open-source-software/bossies-2012-the-best-of-open-source-software-awards-202465

So congratulations to all for the hard work that led to these honors.
This was a broad effort, extending out from the core project to
partners like SourceForge and all those power users whose creative
contributions bring value to the template repository.  These are truly
awards for the whole community!

Regards,

-Rob

Re: [PMC] Proposed PMC List

[Marcus (OOo)]

It's really hard to choose just 10 people if you want to see more people 
on a list. But OK, here is my list:


Andrea Pescetti (pescetti)
Ariel Constenla-Haile (arielch)
Armin Le Grand (alg)
Donald P. Harbison (dpharbison)
Drew Jensen (atjensen)
Oliver-Rainer Wittmann (orw)
Pedro Giffuni (pfg)
RGB.ES (rgb-es)
Roberto Galoppini (galoppini)
Yang Shih-Ching (imacat)

Marcus

Re: [DL][Website] Second draft of the new DL scripting

[Marcus (OOo)]

Am 09/17/2012 11:40 PM, schrieb Kay Schenk:



On 09/17/2012 05:54 AM, Marcus (OOo) wrote:

Hi all,

I was busy another weekend to start to implement the new structure for
the DL scripting. The new webpages are accessible only here (as long as
they are not published):

http://ooo-site.staging.apache.org/download/test/new/index.html
http://ooo-site.staging.apache.org/download/test/new/other.html

Some notes before you start with checking out:

1.
The latest consensus about the new directory structure on mirrors shows
(IMHO) the following that I've used:

...///source/
...///binaries//
...///binaries/SDK/

2.
Only the binaries are working for now. The checksum files are still
coming form the old structure.

3.
But they are no real binaries, just dummy test files with meaningful
content. ;-)


ummm...I'm getting a Not Found for my "binary"

http://people.apache.org/~marcus/aoo/3.4.1/binaries/en-US/Apache_OpenOffice_incubating_3.4.1_Linux_x86_install-rpm_en-US.tar.gz


Is this what's supposed to happen

OK, more info...the en-GB link seems to do something but not en-US...so
maybe just a permission thing with this one file.


Have you already looked at the "analyze.html" webpage? Is there more 
shown wrong?



Generally, so far, so good.

This is definitely moving along well! :)


Thanks for testing.

Marcus




Now happy testing.

And don't forget to use the following webpage to support your error
reports:

http://ooo-site.staging.apache.org/download/test/new/analyze.html

Thanks in advance for your comments.

Marcus

Re: [DL][Website] Second draft of the new DL scripting

[Marcus (OOo)]

Am 09/17/2012 06:45 PM, schrieb Andrea Pescetti:

Marcus (OOo) wrote:

I was busy another weekend to start to implement the new structure for
the DL scripting. The new webpages are accessible only here (as long as
they are not published):
http://ooo-site.staging.apache.org/download/test/new/index.html
http://ooo-site.staging.apache.org/download/test/new/other.html


Thanks! Detection worked fine for me, but apparently the filenames in
http://people.apache.org/~marcus/aoo/3.4.1/binaries/it/
contain a trailing carriage return (!), so I get a "File not Found"
error. However, the link assembled by the download page is right, the
error is just in the dummy files used for testing.


Right, that was an issue when copying some scripts to my Apache account. 
I've corrected it now.


Thanks for the hint.

Marcus

[DL][Website] Second draft of the new DL scripting

[Marcus (OOo)]

Hi all,

I was busy another weekend to start to implement the new structure for 
the DL scripting. The new webpages are accessible only here (as long as 
they are not published):


http://ooo-site.staging.apache.org/download/test/new/index.html
http://ooo-site.staging.apache.org/download/test/new/other.html

Some notes before you start with checking out:

1.
The latest consensus about the new directory structure on mirrors shows 
(IMHO) the following that I've used:


...///source/
...///binaries//
...///binaries/SDK/

2.
Only the binaries are working for now. The checksum files are still 
coming form the old structure.


3.
But they are no real binaries, just dummy test files with meaningful 
content. ;-)


Now happy testing.

And don't forget to use the following webpage to support your error reports:

http://ooo-site.staging.apache.org/download/test/new/analyze.html

Thanks in advance for your comments.

Marcus

Re: Did we ever reach consensus on support for Windows 2000

[Marcus (OOo)]

Am 09/17/2012 01:34 AM, schrieb Fernando Cassia:

On Sun, Sep 16, 2012 at 8:14 PM, Marcus (OOo)  wrote:


I've still a VM with this. But what all should be tested to truthfully
write "tested with Windows 2000"? ;-)



I´ve got a Win2K vm too, haven´t attempted a AOO install. However one point
to keep in mind is that for Win2K to be usable with semi-recent software,
it´s needed to install at least Win2K SP4 + Update Rollup 2 (SR2).


Of course. I, too, have SP4 installed.

Marcus

Re: Record one-day download numbers

[Marcus (OOo)]

Am 09/16/2012 11:17 PM, schrieb Joost Andrae:

Hi Rob,


We set new records for 1-day downloads on the following two days as well:

Wed Aug 29: 195305
Thurs Aug 30: 197479

So four days in a row of new records. Good momentum, I think,
especially for the traditionally slow vacation season.

It looks like around 1/3 of these are upgrade downloads, i.e.,
downloads originating from upgrade notifications in OpenOffice.

As of yesterday the total download count for 3.4.x installs is 16.4
million.



thumbs-up for this! This is really good news - especially for the press...

Kind regards, Joost


Yes, thanks Rob for doing the numbers work.

Marcus

Re: Did we ever reach consensus on support for Windows 2000

[Marcus (OOo)]

Am 09/16/2012 11:35 PM, schrieb Rob Weir:

On Sun, Sep 16, 2012 at 4:26 PM, Keith N. McKenna
  wrote:

Greetings All;

I was going through FAQ's and other pages on the AOO (incubating) site and
noticed that many still are showing that we support Windows 2000 as a
baseline operating system. I though I remembered some discussions a while
back on this list around that subject and thought we had decided that we
would no longer do that due to lack of testing resources.



IMHO, "support" is determined by what we do, not by what we say.  If
no one is testing with Windows 2000, then it is hard to say we support
it.  And if Microsoft does not make Windows 2000 CD's available to
developers for testing, due to a lawsuit, then it is rather difficult
for anyone who wants to test.  Not impossible, but they would need to
get access to CD's or ISO images through unofficial means.

Of course, we could have a dozen people say we *should* support
Windows 2000.  But should does not mean anything.  We really need to
find even a single person who says they *will* test with Windows 2000
and fix any problems that arise.  Until that happens we don't really
support Windows 2000 in any meaningful way.


I went back through the archives and did find a number of threads but they
never seemed to reach a definite conclusion. I we are going to continue to
support it all well and good, but if we cannot then all FAQ's and other
documentation on the site should change to reflect that.



Support is not determined by consensus wishes.  It is determined by
someone actually doing it.


That's right.


Do we have any evidence that users have successfully installed and
used AOO 3.4.x on Windows 2000?  If it works, we might just list it


I've still a VM with this. But what all should be tested to truthfully 
write "tested with Windows 2000"? ;-)



"not a tested configuration, but some users report success.".  In
other words, between "tested and supported" and "known to be broken"
is a middle territory where it is "use at your own risk".


I guess that this will be needed for the next version we will release.

Marcus

Re: Moderating ooo-private

[Marcus (OOo)]

Am 09/15/2012 01:20 PM, schrieb Ariel Constenla-Haile:

Hi Ross,

On Sat, Sep 15, 2012 at 08:42:13AM +0100, Ross Gardler wrote:

Many user enquiries come to ooo-private and are moderated through. They
shouldn't be.  That list should have almost zero traffic.

Users are finding the list somehow, seems the documentation needs fixing.


Grep'ing the main site, it seems they come from somewhere else:

ooo-site]$ git grep --line-number "ooo-priv...@incubator.apache.org"
content/distribution/index.html:69:contact themailto:ooo-priv...@incubator.apache.org";>Apache
content/distribution/index.html:84:Please contact themailto:ooo-priv...@incubator.apache.org";>Apache

Only two matches, and in this page, not very accessible.


Thanks for your quick search.

I doubt that this could be the reason for the many requests on private@. 
Also they are typical enduser/support questions and contain no hint or 
pointer to something distribution-specifc.


However, it could be changed to point to ooo-dev@.

Hm, but this wouldn't help for support questions as they should go to 
ooo-users@. ;-)


Marcus

Re: [DL][Website] First draft of the new DL scripting

[Marcus (OOo)]

Am 09/07/2012 12:03 AM, schrieb Kay Schenk:



On 09/06/2012 02:52 PM, Marcus (OOo) wrote:

Am 09/06/2012 01:33 AM, schrieb Kay Schenk:



On 09/02/2012 03:18 PM, Kay Schenk wrote:



On 09/02/2012 08:02 AM, Marcus (OOo) wrote:

Hi all,

I was busy at the weekend to clean-up the DL scripting.

The new webpages are accessible here (as long as they are not
published):

http://ooo-site.staging.apache.org/download/test/new/index.html
http://ooo-site.staging.apache.org/download/test/new/other.html

Some notes before you start with checking out:

1.
The current focus is on reducing complexity, better simplification and
more automatism. Not the design and layout. This is nearly the same.

2.
It's normal that not all links will work as it's only a part of the
download area. It depends on the link itself if it's not yet
working or
really broken. ;-)

3.
When you find errors and want to report them please make sure to
browse
to the following webpage to help me with some debug data:

http://ooo-site.staging.apache.org/download/test/new/analyze.html

Next big step is to implement the new directory structure on the
mirrors.

Thanks in advance for your comments.

Marcus


VERY good! I am looking at changes in svn and I can see MUCH
improvement. I will try to help with suggestions in the coming week.





...more

You are making very good progress and I hesitate to just make changes
without discussion.

What do you suggest for this?? Additional sections on the wiki page??

https://cwiki.apache.org/confluence/display/OOOUSERS/Improvements+for+the+download+webpages




Hm, I've already created a new table in the Wiki to document the
separating the general improvements from the one towards the new mirror
structure.

But I think this is now enough. Otherwise the overview is really gone.


attachment of pertinent files to the wiki page with comments?


When the things are discussed and done then these data normally remains
- at least this is my impression from other Wiki pages. Then the is also
gone. Besides this I think it becomes very fast a real clutch.

Let's see if we can make some progress here on the ML. When there is
something valuable, then we can take it over to the Wiki page. So, I
would suggest to create new threads with appropriate tags like "[DL
scripting][SCHEMA] ..."

What do you think?


yes, that's a good idea...after re-thinking, maybe my ideas were too
complicated for what you've already successfully simplified.

As things go on, we can revive this if needed.

Thanks again for all you've done, really. Your changes have already
improved the logic a lot. I'll do some more testing.


On the weekend I'll have more spare time and I'm planning to do the next 
updates. So, if you (or anybody else !) have some ideas while testing, 
just tell me.


Marcus

Re: [DL][Website] First draft of the new DL scripting

[Marcus (OOo)]

Am 09/10/2012 11:53 AM, schrieb Jürgen Schmidt:

On 9/6/12 11:52 PM, Marcus (OOo) wrote:

Am 09/06/2012 01:33 AM, schrieb Kay Schenk:



On 09/02/2012 03:18 PM, Kay Schenk wrote:



On 09/02/2012 08:02 AM, Marcus (OOo) wrote:

Hi all,

I was busy at the weekend to clean-up the DL scripting.

The new webpages are accessible here (as long as they are not
published):

http://ooo-site.staging.apache.org/download/test/new/index.html
http://ooo-site.staging.apache.org/download/test/new/other.html

Some notes before you start with checking out:

1.
The current focus is on reducing complexity, better simplification and
more automatism. Not the design and layout. This is nearly the same.

2.
It's normal that not all links will work as it's only a part of the
download area. It depends on the link itself if it's not yet working or
really broken. ;-)

3.
When you find errors and want to report them please make sure to browse
to the following webpage to help me with some debug data:

http://ooo-site.staging.apache.org/download/test/new/analyze.html

Next big step is to implement the new directory structure on the
mirrors.

Thanks in advance for your comments.

Marcus


VERY good! I am looking at changes in svn and I can see MUCH
improvement. I will try to help with suggestions in the coming week.





...more

You are making very good progress and I hesitate to just make changes
without discussion.

What do you suggest for this?? Additional sections on the wiki page??

https://cwiki.apache.org/confluence/display/OOOUSERS/Improvements+for+the+download+webpages



Hm, I've already created a new table in the Wiki to document the
separating the general improvements from the one towards the new mirror
structure.

But I think this is now enough. Otherwise the overview is really gone.


attachment of pertinent files to the wiki page with comments?


When the things are discussed and done then these data normally remains
- at least this is my impression from other Wiki pages. Then the is also
gone. Besides this I think it becomes very fast a real clutch.

Let's see if we can make some progress here on the ML. When there is
something valuable, then we can take it over to the Wiki page. So, I
would suggest to create new threads with appropriate tags like "[DL
scripting][SCHEMA] ..."

What do you think?

Marcus



I think we can finalize the future directory structure #part 2 on the
wiki page.

Current proposal:
.../ooo//source/
.../ooo//bin//
.../ooo//bin/SDK/

When I remember there were some concerns regarding "bin". And I would


Hm, I don't know if it makes really a difference in case of names on 
mirrors. ;-) But OK, let's call it "binaries".



keep the project name open or flexible assuming a change in favor of "aoo".

...///source/
...///binaries//
...///binaries/SDK/


Where PROJECT_NAME is something like "ooo" or "aoo" or "openoffice". I
would say we need no voting for this and need further discussion only if
somebody see some problems with the new proposed structure.


So, waiting for the final definition. For the moment I'll use "aoo" for 
my testing.


Marcus

Re: Volunteers needed to pickup some tasks

[Marcus (OOo)]

Am 09/08/2012 07:50 AM, schrieb Rob Weir:

I'm raising my focus a little in this project to look more into some
larger ecosystem opportunities. As part of this change in focus I'll
have less time to deal with day-to-day operational aspects of the
project. So I'd like to shed some responsibilities and I hope there
are volunteers able or willing to learn to pick up the following:

1. List moderation. I intend to stop moderating dev, users, marketing,
qa and private.

2. Bugzilla Admin


See my other answer to TJ.


3. Taking the lead on the AOO Security team, tracking vulnerability
reports, writing disclosure bulletins, coordinating with security
analysts and related open source projects.

4. Generally making sure things don't fall through the cracks. For
example, look at posts from a week ago and see which ones have not
been responded to.  Look at trademark, press, movie property requests,
etc. and make sure necessary workflow and approvals are secured.
Because of the fast "scroll rate" of our posts, if someone is not
proactive about such things they will be overlooked.

5. I have a couple blog posts in progress. I'll finish those. But we
really could use some more content. This is something anyone can do.
Maybe even turn your ApacheCon abstracts into a blog post with a link
to you full presentation?

6. Tracking project metrics for downloads, committers, etc.


I could takeover the metrics for downloads. Maybe Kay wants to help, 
too? ;-)



Perhaps a few other things as well, but that is a start.

Regards,

Rob


Marcus

Re: Volunteers needed to pickup some tasks

[Marcus (OOo)]

Am 09/08/2012 11:06 PM, schrieb TJ Frazier:

On 9/8/2012 16:50, tj wrote:

On 9/8/2012 12:01, David McKay wrote:


On 08/09/12 06:50, Rob Weir wrote:

I'm raising my focus a little in this project to look more into some
larger ecosystem opportunities. As part of this change in focus I'll
have less time to deal with day-to-day operational aspects of the
project. So I'd like to shed some responsibilities and I hope there
are volunteers able or willing to learn to pick up the following:

1. List moderation. I intend to stop moderating dev, users, marketing,
qa and private.

2. Bugzilla Admin

I can probably help out on this item, if someone can send me some
details of what is involved/expected (I have been a Bugzilla admin as
part of a paid role in a previous job).

Dave.



Hi, Dave,

I remember your previous posts on this, and your help will be very
welcome. I am totally ignorant of the underlying code, so as a BZ admin
I stick to the little day-to-day stuff: adding privileges or deleting
too-personal attachments. Others are of course welcome to do those, too;
if I don't "know" somebody from the list, I generally wait three days
(lazy consensus) and add privileges if nobody objects.

Some of the larger things which have been done by others (including Rob)
involve adding fields or groups, and adding values to field lists (e.g.
"3.4.1" to the version options).

If you are not already an admin, let me know your BZ id.


Never mind, found it. Welcome to BZ admin. --/tj/


In the past I requested this and that in BZ. Maybe time to do it myself? ;-)

So, if you still need help please add me as BZ admin. My spare time is 
limited but little BZ things to maintain should still fit into this.


Thanks

Marcus

Re: [DL][Website] First draft of the new DL scripting

[Marcus (OOo)]

Am 09/06/2012 01:33 AM, schrieb Kay Schenk:



On 09/02/2012 03:18 PM, Kay Schenk wrote:



On 09/02/2012 08:02 AM, Marcus (OOo) wrote:

Hi all,

I was busy at the weekend to clean-up the DL scripting.

The new webpages are accessible here (as long as they are not
published):

http://ooo-site.staging.apache.org/download/test/new/index.html
http://ooo-site.staging.apache.org/download/test/new/other.html

Some notes before you start with checking out:

1.
The current focus is on reducing complexity, better simplification and
more automatism. Not the design and layout. This is nearly the same.

2.
It's normal that not all links will work as it's only a part of the
download area. It depends on the link itself if it's not yet working or
really broken. ;-)

3.
When you find errors and want to report them please make sure to browse
to the following webpage to help me with some debug data:

http://ooo-site.staging.apache.org/download/test/new/analyze.html

Next big step is to implement the new directory structure on the
mirrors.

Thanks in advance for your comments.

Marcus


VERY good! I am looking at changes in svn and I can see MUCH
improvement. I will try to help with suggestions in the coming week.





...more

You are making very good progress and I hesitate to just make changes
without discussion.

What do you suggest for this?? Additional sections on the wiki page??

https://cwiki.apache.org/confluence/display/OOOUSERS/Improvements+for+the+download+webpages


Hm, I've already created a new table in the Wiki to document the 
separating the general improvements from the one towards the new mirror 
structure.


But I think this is now enough. Otherwise the overview is really gone.


attachment of pertinent files to the wiki page with comments?


When the things are discussed and done then these data normally remains 
- at least this is my impression from other Wiki pages. Then the is also 
gone. Besides this I think it becomes very fast a real clutch.


Let's see if we can make some progress here on the ML. When there is 
something valuable, then we can take it over to the Wiki page. So, I 
would suggest to create new threads with appropriate tags like "[DL 
scripting][SCHEMA] ..."


What do you think?

Marcus

Re: [PERSONAL] My absence

[Marcus (OOo)]

Don, I'm sorry to hear that. I wish your wife and you all the best for 
her recovery which will take likely much energy.


We would be happy to welcome you back. Take your time off as long as 
it's needed.


Marcus



Am 09/06/2012 10:28 PM, schrieb Donald Harbison:

I feel it necessary to explain why I have been absent from discussions here
and on ooo-private recently.

My wife was stricken with a serious medical emergency Wednesday, August
29th. She is recovering after surgery, and is in an intensive care unit
here in one of the top hospitals in Boston. Needless to say, my attention
is not on Apache OpenOffice as much as I love this project and community.
The past week has been personally very stressful and exhausting. Next week
looks a whole lot better as her condition has improved dramatically and she
is now on the path to recovery.

Graduation is a very important goal, I will be doing what I can to help
with this. I've also been actively leading the OpenOffice track development
for ApacheCon EU. Oliver-Rainer Wittmann has stepped in to take over this
effort over the past week. Thank you Oliver.

Thanks for your understanding. The medical team is optimistic that my wife
will have a full recovery, but it will be a lengthy process.  In the
meantime, I do plan to crank up my work engine as her care and situation
have now become much more stabilized.

Re: board report for Sept. 20, 2012

[Marcus (OOo)]

Sorry, when I'm a bit late with this.

Not that it makes a real difference. But in the March and June report it 
was named "OpenOffice (was OpenOffice.org)". Now it's again 
"OpenOffice.org".


Furthermore, I don't like the formulation

"We expect to graduate in the next quarter."

Maybe it's because I'm not a native speaker but for me it sounds like 
too much demanding. A sentence like "We plan and work to fulfill 
graduation in the next quarter." fits much better.


Of course, my 2 ct.

Marcus



Am 09/05/2012 11:01 PM, schrieb Kay Schenk:



On 09/04/2012 09:04 PM, Dennis E. Hamilton wrote:

Minor correction:

The proposal is at 
down the list of reports in their alphabetical order.


- Dennis


yes, thanks for the correction Dennis and for forging ahead with this Dave.




-Original Message-
From: Dave Fisher [mailto:dave2w...@comcast.net]
Sent: Tuesday, September 04, 2012 20:50
To: ooo-dev@incubator.apache.org
Cc: orc...@apache.org
Subject: Re: board report for Sept. 20, 2012

Hi,


Thanks. I've essentially followed your suggestions. I hope that
others will review the wiki at
https://cwiki.apache.org/confluence/display/OOOUSERS/2012+Sept before
it is moved to http://wiki.apache.org/incubator/June2012 on Wednesday.


The report has been moved to the incubator wiki. Please let us know if
you think anything should be edited.

Regards,
Dave

[ ... ]

Re: [BZ] Please add 3.4.1 as version to all components in BugZilla

[Marcus (OOo)]

Am 08/28/2012 10:19 PM, schrieb Marcus (OOo):

Ping :-)


Thanks to the unknown helper. :-)

Marcus




Am 08/25/2012 12:08 AM, schrieb Marcus (OOo):

Hi BZ admins,

as AOO 3.4.1 is now available, please add the 3.4.1 version into the BZ
instance of AOO.

Re: [DL Website] Prototype of an automatic table with download links

[Marcus (OOo)]

Am 08/27/2012 11:25 PM, schrieb Kay Schenk:



On 08/27/2012 02:07 PM, Marcus (OOo) wrote:

Am 08/27/2012 12:38 AM, schrieb Marcus (OOo):

Am 08/27/2012 12:12 AM, schrieb Dave Fisher:


On Aug 26, 2012, at 2:59 PM, Marcus (OOo) wrote:


Am 08/26/2012 11:37 PM, schrieb Marcus (OOo):

Am 08/26/2012 09:38 PM, schrieb Dave Fisher:


On Aug 26, 2012, at 11:05 AM, Marcus (OOo) wrote:


Hi all,

as promized on Friday I've adjusted the existing automatism to the
lastest AOO 3.4.1 parameters.

The most recent webpage is here:
http://ooo-site.staging.apache.org/download/test/other_print.html


There is one change needed with the source downloads. Please use the
Apache Mirrors and not dist directly for the packages. See [1].


Only for the source files, not for the hash files, right? Done.


Should the language packs be mixed with the full installation
sets? If
so then maybe the language fields should span two rows?


To have all in a single table it's easier to point the user to resp.
the
user has only to remember this.

I'll think about the span thing.


As "fast solution" I've deleted every second language strings and
expanded the row highlighting.

Dave, what do you think?


Much better.

Maybe the go back to the top break in the language group should be
more frequent than every 10?


Ah, right. With the combined table (full install, langpacks, hashes)
it's still after 10 languages but actually a lot of big rows more which
is indeed to much. I'll change it tomorrow.


I've changed it in a way that max. 4 languages are printed and than a
new sub-table header.

With the most high screen resolutions nowadays this should fit. On my
screen the visible part begins with "Start of page" and it ends with
this.

Marcus




This seems fine to me. Nice job!


Thanks :-)

I was busy the last hours and have started the re-work.

The reduction is already quite nice. Only 1 "download.js" is left. The 
new "download_tables.js" is for the new "other.html" file only. Also 
"globalvars.js" is smaller. I've also adjusted a bit the "languages.js".


And - at least for me on Linux - all is still working. :-)

PS:
I don't write you this to push you to something. It's just that I want 
to let you know that it was indeed time to do this. The reduction is 
really great.


CU

Marcus




[1] http://incubator.apache.org/openofficeorg/downloads.html


I'm open for opinions. Otherwise I can put it easily to the
production website as the new "other.html".

[DL][Website] First draft of the new DL scripting

[Marcus (OOo)]

Hi all,

I was busy at the weekend to clean-up the DL scripting.

The new webpages are accessible here (as long as they are not published):

http://ooo-site.staging.apache.org/download/test/new/index.html
http://ooo-site.staging.apache.org/download/test/new/other.html

Some notes before you start with checking out:

1.
The current focus is on reducing complexity, better simplification and 
more automatism. Not the design and layout. This is nearly the same.


2.
It's normal that not all links will work as it's only a part of the 
download area. It depends on the link itself if it's not yet working or 
really broken. ;-)


3.
When you find errors and want to report them please make sure to browse 
to the following webpage to help me with some debug data:


http://ooo-site.staging.apache.org/download/test/new/analyze.html

Next big step is to implement the new directory structure on the mirrors.

Thanks in advance for your comments.

Marcus

Re: [ooo-site]

[Marcus (OOo)]

Am 09/01/2012 09:00 PM, schrieb Yotam Dikstein:

Hi
All the mirrors to the full pack of open office 3.4.1 in English are
broken. The download begins, but stops after a few MBs.

YT,
yotam


On the "other.html" webpage I've started to download the following file:

http://sourceforge.net/projects/openofficeorg.mirror/files/localized/de/3.4.1/Apache_OpenOffice_incubating_3.4.1_Win_x86_install_de.exe/download

I was redirected to:

http://heanet.dl.sourceforge.net/project/openofficeorg.mirror/localized/de/3.4.1/Apache_OpenOffice_incubating_3.4.1_Win_x86_install_de.exe

At the end I got the complete file and verified this with the MD5 checksum.

Maybe you want to try this specific mirror 
("http://heanet.dl.sourceforge.net/project/openofficeorg.mirror/";) to 
try yourself.


HTH

Marcus

Re: svn commit: r1377138 - /incubator/ooo/ooo-site/trunk/content/development/releases/3.4.1.html

[Marcus (OOo)]

Am 08/31/2012 10:48 PM, schrieb Andrea Pescetti:

Marcus (OOo) wrote:

[...]
Apache OpenOffice 3.4.0 and 3.4.1 support Java 7, which is the
recommended configuration; but (especially on 64-bit Windows) you might
receive warnings about the Java version being defective. In that case,
download and install [the most current JRE 6 version | link]. Make sure
you get the "Windows x86 Offline (32-bit)" file. Then configure
OpenOffice to use it at "Tools - Options - OpenOffice.org - Java". See
[this forum topic | link] for more information.
[...]
With this everybody should be happy, right?


Yes, the only reason for me to suggest a specific version was based on
the previous analysis by Andreas; after his new analysis with the latest
versions of OpenOffice and Java, this is the best recommendation we can
make.


OK, I've adjusted the text.

Marcus

Re: svn commit: r1377138 - /incubator/ooo/ooo-site/trunk/content/development/releases/3.4.1.html

[Marcus (OOo)]

Am 08/31/2012 10:55 PM, schrieb Michal Hriň:

Dňa Fri, 31 Aug 2012 22:39:13 +0200 Marcus (OOo) 
napísal:


Am 08/31/2012 11:58 AM, schrieb Andreas Säger:

Am 31.08.2012 00:01, Marcus (OOo) wrote:

Am 08/30/2012 09:09 AM, schrieb Andrea Pescetti:

On 29/08/2012 Andreas Säger wrote:

Then extract some old version of Java6 (<= revision 23), point the
office to it and try again. It's many times faster.


So, to recap: the best version of Java we could mention as a
workaround
(after saying that Java 7 is supported and more secure, of course) is
"jre-6u23-windows-i586.exe" as found on
http://www.oracle.com/technetwork/java/javasebusiness/downloads/java-archive-downloads-javase6-419409.html#jre-6u23-oth-JPR




right?


But then we should write a bit more text with another paragraph to
point
explicitely to this specific version when the user needs it. Then the
users can choose themselves:

a) The most recent JRE 6 if they look for security.
b) JRE 6u23 if they need the AOO Base compatibility.


That one is older than the versions that both I (6.32) and Marcus (6.x
current) had put on that page, but since people will go to that page
expecting to find a workaround I'd prefer to mention a version that
can
be a workaround for multiple issues (Java not found, Java operations
slow in Base and so on). Andreas in
http://markmail.org/message/w6x6e4dlzgwtfyd7 mentions that 6.22 works,
but if 6.23 is fine too I'd list the newer one.


+1

So, when this is our consensus I will update the webpage on the
weekend.

Marcus



Hi,

Please bear with me. I created some Oracle account to access their
binary attic and went throug a testing parcour on my 32 bit system where
I found out that _something_ must have changed between OOo 3.3 and
AOO 3.4:

JRE version Linux i586 (Ubuntu 10.4, AOO 3.4)
6.0.22 good /which is what I used to use with OOo 3.3
6.0.23 good /don't know about OOo 3.3
6.0.24 good /used to be bad with OOo 3.3
6.0.24(open-jdk) good /used to be bad with OOo 3.3
6.0.35(latest) good /used to be bad with OOo 3.3
7.0.4 bad


This is the result for my 64 bit Linux machine:
JRE version Linux x64 (Ubuntu 10.4, AOO 3.4.1)
6.0.24(open-jdk) good
6.0.35 good
7.x bad

So the issue has a clear red line between JRE6 and JRE7.


Thanks a lot for you testing effort. With this information I will
adjust the text like the follwing:

[...]
Apache OpenOffice 3.4.0 and 3.4.1 support Java 7, which is the
recommended configuration; but (especially on 64-bit Windows) you
might receive warnings about the Java version being defective. In that
case, download and install [the most current JRE 6 version | link].
Make sure you get the "Windows x86 Offline (32-bit)" file. Then
configure OpenOffice to use it at "Tools - Options - OpenOffice.org -
Java". See [this forum topic | link] for more information.
[...]

With this everybody should be happy, right?

Marcus


I'm only comment, for me AOO 3.4.1 on Windows 64bit doesn't work at all
with Orace Java 7 (32 or 64).

The only option that works for me is to use to Orace/Sun Java 6 32bit on
Windows 64bit.
On Ubuntu 12.04 64bit I'm using 64bit openjdk6-jre 64 bit and everything
works (i didn't try jdk7).


Thanks for your comment. Exactly this should be the message in the text 
I want to update.


Marcus

Re: AOO341 Solaris SPARC

[Marcus (OOo)]

Am 08/31/2012 01:15 PM, schrieb Jean-Louis 'Hans' Fuchs:

We have compiled Apache OpenOffice 3.4.1 for Solaris SPARC

http://adfinis-sygroup.ch/aoo-solaris-sparc

We tested it on:

- OpenSolaris 2009.06 SPARC
- Oracle Solaris 10 8/11 s10s_u10wos_17b SPARC

it does NOT run on:

- Oracle Solaris 11 11/11 SPARC

We are investigating the problem.


Great to see progress also for Solaris. I've adjusted a bit the links on 
the porting webpage. These are pointing now directly to the Sparc and 
x86 website.


Marcus

Re: svn commit: r1377138 - /incubator/ooo/ooo-site/trunk/content/development/releases/3.4.1.html

[Marcus (OOo)]

Am 08/31/2012 11:58 AM, schrieb Andreas Säger:

Am 31.08.2012 00:01, Marcus (OOo) wrote:

Am 08/30/2012 09:09 AM, schrieb Andrea Pescetti:

On 29/08/2012 Andreas Säger wrote:

Then extract some old version of Java6 (<= revision 23), point the
office to it and try again. It's many times faster.


So, to recap: the best version of Java we could mention as a workaround
(after saying that Java 7 is supported and more secure, of course) is
"jre-6u23-windows-i586.exe" as found on
http://www.oracle.com/technetwork/java/javasebusiness/downloads/java-archive-downloads-javase6-419409.html#jre-6u23-oth-JPR



right?


But then we should write a bit more text with another paragraph to point
explicitely to this specific version when the user needs it. Then the
users can choose themselves:

a) The most recent JRE 6 if they look for security.
b) JRE 6u23 if they need the AOO Base compatibility.


That one is older than the versions that both I (6.32) and Marcus (6.x
current) had put on that page, but since people will go to that page
expecting to find a workaround I'd prefer to mention a version that can
be a workaround for multiple issues (Java not found, Java operations
slow in Base and so on). Andreas in
http://markmail.org/message/w6x6e4dlzgwtfyd7 mentions that 6.22 works,
but if 6.23 is fine too I'd list the newer one.


+1

So, when this is our consensus I will update the webpage on the weekend.

Marcus



Hi,

Please bear with me. I created some Oracle account to access their
binary attic and went throug a testing parcour on my 32 bit system where
I found out that _something_ must have changed between OOo 3.3 and AOO 3.4:

JRE version Linux i586 (Ubuntu 10.4, AOO 3.4)
6.0.22 good /which is what I used to use with OOo 3.3
6.0.23 good /don't know about OOo 3.3
6.0.24 good /used to be bad with OOo 3.3
6.0.24(open-jdk) good /used to be bad with OOo 3.3
6.0.35(latest) good /used to be bad with OOo 3.3
7.0.4 bad


This is the result for my 64 bit Linux machine:
JRE version Linux x64 (Ubuntu 10.4, AOO 3.4.1)
6.0.24(open-jdk) good
6.0.35 good
7.x bad

So the issue has a clear red line between JRE6 and JRE7.


Thanks a lot for you testing effort. With this information I will adjust 
the text like the follwing:


[...]
Apache OpenOffice 3.4.0 and 3.4.1 support Java 7, which is the 
recommended configuration; but (especially on 64-bit Windows) you might 
receive warnings about the Java version being defective. In that case, 
download and install [the most current JRE 6 version | link]. Make sure 
you get the "Windows x86 Offline (32-bit)" file. Then configure 
OpenOffice to use it at "Tools - Options - OpenOffice.org - Java". See 
[this forum topic | link] for more information.

[...]

With this everybody should be happy, right?

Marcus

Re: svn commit: r1377138 - /incubator/ooo/ooo-site/trunk/content/development/releases/3.4.1.html

[Marcus (OOo)]

Am 08/30/2012 09:09 AM, schrieb Andrea Pescetti:

On 29/08/2012 Andreas Säger wrote:

Then extract some old version of Java6 (<= revision 23), point the
office to it and try again. It's many times faster.


So, to recap: the best version of Java we could mention as a workaround
(after saying that Java 7 is supported and more secure, of course) is
"jre-6u23-windows-i586.exe" as found on
http://www.oracle.com/technetwork/java/javasebusiness/downloads/java-archive-downloads-javase6-419409.html#jre-6u23-oth-JPR

right?


But then we should write a bit more text with another paragraph to point 
explicitely to this specific version when the user needs it. Then the 
users can choose themselves:


a) The most recent JRE 6 if they look for security.
b) JRE 6u23 if they need the AOO Base compatibility.


That one is older than the versions that both I (6.32) and Marcus (6.x
current) had put on that page, but since people will go to that page
expecting to find a workaround I'd prefer to mention a version that can
be a workaround for multiple issues (Java not found, Java operations
slow in Base and so on). Andreas in
http://markmail.org/message/w6x6e4dlzgwtfyd7 mentions that 6.22 works,
but if 6.23 is fine too I'd list the newer one.


+1

So, when this is our consensus I will update the webpage on the weekend.

Marcus

Re: svn commit: r1377138 - /incubator/ooo/ooo-site/trunk/content/development/releases/3.4.1.html

[Marcus (OOo)]

Am 08/29/2012 06:48 PM, schrieb Ariel Constenla-Haile:

Hi Andrea, *

On Wed, Aug 29, 2012 at 06:38:48PM +0200, Andrea Pescetti wrote:

I found this commit from Marcus, from a few days ago:


Author: marcus
Date: Fri Aug 24 22:17:10 2012
New Revision: 1377138
URL: http://svn.apache.org/viewvc?rev=1377138&view=rev
Log:
#i120681# Updated the link to JRE 6 to point to the lastest build instead of a 
fixed version.


Now, the idea of updating the link is correct, but if I recall
correctly the safest solution is to download the file we explicitly
mentioned (see the diff at the URL above for more details):
downloading a certain JRE version will still leave the doubt on
whether the user should get 32-bit or 64-bit files, like the ones
listed in
http://www.oracle.com/technetwork/java/javase/downloads/jre-6u32-downloads-1594646.html


I've changed it because I thought to point to a specific version doesn't 
make any sense.


Maybe I haven't followed the Windows <--> JRE 6/7 discussions not 
closely enough. :-/



Unless we know that all work (and from what I heard not all of them
do) it would be good to at least mention they need to use the
"Windows x86 Offline (32-bit)" file, but not having tried I won't
change the page again. Did anybody try?

Regards,
   Andrea.


According to Andreas http://markmail.org/message/w6x6e4dlzgwtfyd7 Base
needs a very specific version of JRE 6, may be this is worth to be
mentioned somewhere because pointing to the latest jre 6 won't work for
Base users.


@Andreas:
Is that true, is there really a hard dependency for a specific JRE version?

Thanks

Marcus

Re: having problems with download

[Marcus (OOo)]

Am 08/29/2012 12:44 PM, schrieb PRABHU S:

respected sir/madam,

i want to use your most useful software openoffice. but when i download it,
it downloads only 99.90%. after that it gets stopped. please resolve this
issue and so many people like me will get to use this..


thanking you!


regards

PRABU.

note: downloaded using internet download manager from internode server.. it
splits the links into 8. except 1&  3, others are completed.


I don't know from where you want to download the software. But please 
make sure you download only from official and well-known locations. For 
Apache OpenOffice it is the following webpage:


http://www.openoffice.org/download/

Hope this help.

Marcus

Re: [BZ] Please add 3.4.1 as version to all components in BugZilla

[Marcus (OOo)]

Ping :-)



Am 08/25/2012 12:08 AM, schrieb Marcus (OOo):

Hi BZ admins,

as AOO 3.4.1 is now available, please add the 3.4.1 version into the BZ
instance of AOO.

Thanks

Marcus

Re: Something wrong in download page for Linux users

[Marcus (OOo)]

Am 08/27/2012 11:51 PM, schrieb drew:




Howdy

http://www.openoffice.org/download/test/analyze.html
64bit Ubuntu 11.04 - latest Chromium - still thinks I am 32Bit Debs :(


Please can you give me the first part of the data in the table
("Variables from the browser | Values")?

Thanks

Marcus



Copy/paste from the html page just now:
https://docs.google.com/spreadsheet/ccc?key=0Ah7ZNEXlmR0IdGdCRXZVbE5vdmZrdlc2TzhaUV81c3c


Thanks. Please see my answer to Ariel. It seems the difference in both 
versions is the root cause.


Marcus

Re: Something wrong in download page for Linux users

[Marcus (OOo)]

Am 08/27/2012 11:51 PM, schrieb Ariel Constenla-Haile:

On Mon, Aug 27, 2012 at 11:26:45PM +0200, Marcus (OOo) wrote:

Can you confirm this especially for Chrome?


Howdy

http://www.openoffice.org/download/test/analyze.html
64bit Ubuntu 11.04 - latest Chromium - still thinks I am 32Bit Debs :(


Please can you give me the first part of the data in the table
("Variables from the browser | Values")?


For me it's working with this Chrome version:

Name: google-chrome-beta
Arch: x86_64
Version : 22.0.1229.14
Release : 152690

recognized as

navigator.userAgent Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.4 (KHTML, 
like Gecko) Chrome/22.0.1229.14 Safari/537.4


May be Drew is running a Chrome version without the bug fix.
According to http://code.google.com/p/chromium/issues/detail?id=128167
it fixed upstream in https://bugs.webkit.org/show_bug.cgi?id=86778


Yes, that's my guess, too. You have version "22.0.1229.14" but Drew 
"18.0.1025.151". I don't know how often Chrome is updated but it looks 
reasonable older to justify this difference.


Thanks

Marcus

Re: Something wrong in download page for Linux users

[Marcus (OOo)]

Am 08/27/2012 08:52 PM, schrieb drew:

On Sun, 2012-08-26 at 20:18 +0200, Marcus (OOo) wrote:

Am 08/24/2012 11:52 PM, schrieb Marcus (OOo):

Am 08/23/2012 04:24 PM, schrieb Ariel Constenla-Haile:


Hi Marcus,

On Fri, May 18, 2012 at 01:49:19AM +0200, Marcus (OOo) wrote:

Am 05/18/2012 01:29 AM, schrieb Ariel Constenla-Haile:

On Fri, May 18, 2012 at 12:26:41AM +0200, Marcus (OOo) wrote:


The system and the browser are 64 bits, the package is 32 bits.


Interesting. The browser shows that the platform is i686 (= x86) and
the user agents says x86_64. Haven't seen this before.

OK, which value is right when you don't know the truth? ;-)


just blame it on Google :)

http://code.google.com/p/chromium/issues/detail?id=44905


Interesting, even Google software has old bugs. :-P


Duplicated by this one?
http://code.google.com/p/chromium/issues/detail?id=128167


Great. When this is solved somewhen, we can check our DL logic
again. I'll add this to the Wiki page.



FYI this is fixed now in Chrome, according to the browser values shown
by http://www.openoffice.org/download/test/analyze.html


Ah, thank you for the hint. I will analyze the data what needs to be
updated.

BTW:
Great to have this little test webpage online, isn't it? ;-)


This should work now with the recent change from Oliver.

Can you confirm this especially for Chrome?


Howdy

http://www.openoffice.org/download/test/analyze.html
64bit Ubuntu 11.04 - latest Chromium - still thinks I am 32Bit Debs :(


Please can you give me the first part of the data in the table 
("Variables from the browser | Values")?


Thanks

Marcus

Re: [DL Website] Prototype of an automatic table with download links

[Marcus (OOo)]

Am 08/27/2012 12:38 AM, schrieb Marcus (OOo):

Am 08/27/2012 12:12 AM, schrieb Dave Fisher:


On Aug 26, 2012, at 2:59 PM, Marcus (OOo) wrote:


Am 08/26/2012 11:37 PM, schrieb Marcus (OOo):

Am 08/26/2012 09:38 PM, schrieb Dave Fisher:


On Aug 26, 2012, at 11:05 AM, Marcus (OOo) wrote:


Hi all,

as promized on Friday I've adjusted the existing automatism to the
lastest AOO 3.4.1 parameters.

The most recent webpage is here:
http://ooo-site.staging.apache.org/download/test/other_print.html


There is one change needed with the source downloads. Please use the
Apache Mirrors and not dist directly for the packages. See [1].


Only for the source files, not for the hash files, right? Done.


Should the language packs be mixed with the full installation sets? If
so then maybe the language fields should span two rows?


To have all in a single table it's easier to point the user to resp.
the
user has only to remember this.

I'll think about the span thing.


As "fast solution" I've deleted every second language strings and
expanded the row highlighting.

Dave, what do you think?


Much better.

Maybe the go back to the top break in the language group should be
more frequent than every 10?


Ah, right. With the combined table (full install, langpacks, hashes)
it's still after 10 languages but actually a lot of big rows more which
is indeed to much. I'll change it tomorrow.


I've changed it in a way that max. 4 languages are printed and than a 
new sub-table header.


With the most high screen resolutions nowadays this should fit. On my 
screen the visible part begins with "Start of page" and it ends with this.


Marcus




[1] http://incubator.apache.org/openofficeorg/downloads.html


I'm open for opinions. Otherwise I can put it easily to the
production website as the new "other.html".

Re: [DL Website] Prototype of an automatic table with download links

[Marcus (OOo)]

Am 08/27/2012 12:12 AM, schrieb Dave Fisher:


On Aug 26, 2012, at 2:59 PM, Marcus (OOo) wrote:


Am 08/26/2012 11:37 PM, schrieb Marcus (OOo):

Am 08/26/2012 09:38 PM, schrieb Dave Fisher:


On Aug 26, 2012, at 11:05 AM, Marcus (OOo) wrote:


Hi all,

as promized on Friday I've adjusted the existing automatism to the
lastest AOO 3.4.1 parameters.

The most recent webpage is here:
http://ooo-site.staging.apache.org/download/test/other_print.html


There is one change needed with the source downloads. Please use the
Apache Mirrors and not dist directly for the packages. See [1].


Only for the source files, not for the hash files, right? Done.


Should the language packs be mixed with the full installation sets? If
so then maybe the language fields should span two rows?


To have all in a single table it's easier to point the user to resp. the
user has only to remember this.

I'll think about the span thing.


As "fast solution" I've deleted every second language strings and expanded the 
row highlighting.

Dave, what do you think?


Much better.

Maybe the go back to the top break in the language group should be more 
frequent than every 10?


Ah, right. With the combined table (full install, langpacks, hashes) 
it's still after 10 languages but actually a lot of big rows more which 
is indeed to much. I'll change it tomorrow.


Marcus




[1] http://incubator.apache.org/openofficeorg/downloads.html


I'm open for opinions. Otherwise I can put it easily to the
production website as the new "other.html".

Re: [DL Website] Prototype of an automatic table with download links

[Marcus (OOo)]

Am 08/26/2012 11:37 PM, schrieb Marcus (OOo):

Am 08/26/2012 09:38 PM, schrieb Dave Fisher:


On Aug 26, 2012, at 11:05 AM, Marcus (OOo) wrote:


Hi all,

as promized on Friday I've adjusted the existing automatism to the
lastest AOO 3.4.1 parameters.

The most recent webpage is here:
http://ooo-site.staging.apache.org/download/test/other_print.html


There is one change needed with the source downloads. Please use the
Apache Mirrors and not dist directly for the packages. See [1].


Only for the source files, not for the hash files, right? Done.


Should the language packs be mixed with the full installation sets? If
so then maybe the language fields should span two rows?


To have all in a single table it's easier to point the user to resp. the
user has only to remember this.

I'll think about the span thing.


As "fast solution" I've deleted every second language strings and 
expanded the row highlighting.


Dave, what do you think?

Marcus




[1] http://incubator.apache.org/openofficeorg/downloads.html


I'm open for opinions. Otherwise I can put it easily to the
production website as the new "other.html".

Re: [DL Website] Prototype of an automatic table with download links

[Marcus (OOo)]

Am 08/26/2012 09:38 PM, schrieb Dave Fisher:


On Aug 26, 2012, at 11:05 AM, Marcus (OOo) wrote:


Hi all,

as promized on Friday I've adjusted the existing automatism to the lastest AOO 
3.4.1 parameters.

The most recent webpage is here:
http://ooo-site.staging.apache.org/download/test/other_print.html


There is one change needed with the source downloads. Please use the Apache 
Mirrors and not dist directly for the packages. See [1].


Only for the source files, not for the hash files, right? Done.


Should the language packs be mixed with the full installation sets? If so then 
maybe the language fields should span two rows?


To have all in a single table it's easier to point the user to resp. the 
user has only to remember this.


I'll think about the span thing.


[1] http://incubator.apache.org/openofficeorg/downloads.html


I'm open for opinions. Otherwise I can put it easily to the production website as the new 
"other.html".


Marcus

Re: Something wrong in download page for Linux users

[Marcus (OOo)]

Am 08/24/2012 11:52 PM, schrieb Marcus (OOo):

Am 08/23/2012 04:24 PM, schrieb Ariel Constenla-Haile:


Hi Marcus,

On Fri, May 18, 2012 at 01:49:19AM +0200, Marcus (OOo) wrote:

Am 05/18/2012 01:29 AM, schrieb Ariel Constenla-Haile:

On Fri, May 18, 2012 at 12:26:41AM +0200, Marcus (OOo) wrote:


The system and the browser are 64 bits, the package is 32 bits.


Interesting. The browser shows that the platform is i686 (= x86) and
the user agents says x86_64. Haven't seen this before.

OK, which value is right when you don't know the truth? ;-)


just blame it on Google :)

http://code.google.com/p/chromium/issues/detail?id=44905


Interesting, even Google software has old bugs. :-P


Duplicated by this one?
http://code.google.com/p/chromium/issues/detail?id=128167


Great. When this is solved somewhen, we can check our DL logic
again. I'll add this to the Wiki page.



FYI this is fixed now in Chrome, according to the browser values shown
by http://www.openoffice.org/download/test/analyze.html


Ah, thank you for the hint. I will analyze the data what needs to be
updated.

BTW:
Great to have this little test webpage online, isn't it? ;-)


This should work now with the recent change from Oliver.

Can you confirm this especially for Chrome?

Thanks

Marcus




Variables from the browser Values

navigator.platform Linux x86_64
navigator.platform.toLowerCase() linux x86_64
navigator.language en-US
navigator.userLanguage undefined
navigator.systemLanguage undefined
navigator.userAgent Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.1
(KHTML, like Gecko) Chrome/21.0.1180.81 Safari/537.1
navigator.userAgent.toLowerCase() mozilla/5.0 (x11; linux x86_64)
applewebkit/537.1 (khtml, like gecko) chrome/21.0.1180.81 safari/537.1
navigator.javaEnabled() Yes


But the download page is providing 32 bits to download (though the text
says "Click to start downloading the most recent version for Linux
64-bit (RPM and English (US)"):


JavaScript functions from the DL scripts Return values
getLink( VERSION, LANGUAGE, MIRROR, SCHEMA )
http://sourceforge.net/projects/openofficeorg.mirror/files/stable/3.4.1/Apache_OpenOffice_incubating_3.4.1_Linux_x86_install-rpm_en-US.tar.gz/download

getArray( LANGUAGE ) here,English (US),English
(US),http://www.openoffice.org/download/other.html,y
getPlatform( LANGUAGE, SCHEMA ) Linux 64-bit (RPM)
getLanguage( LANGUAGE ) English (US)
getLanguageISO( LANGUAGE ) en-US
sourceforge_getLink( VERSION, LANGUAGE, SCHEMA )
http://sourceforge.net/projects/openofficeorg.mirror/files/stable/3.4.1/Apache_OpenOffice_incubating_3.4.1_Linux_x86_install-rpm_en-US.tar.gz/download

apache_getLink( VERSION, LANGUAGE, SCHEMA )
http://www.apache.org/dyn/closer.cgi/incubator/ooo/files/stable/3.4.1/Apache_OpenOffice_incubating_3.4.1_Linux_x86_install-rpm_en-US.tar.gz

apache_getChecksum( VERSION, LANGUAGE, SCHEMA, HASH )
http://www.apache.org/dist/incubator/ooo/files/stable/3.4.1/Apache_OpenOffice_incubating_3.4.1_Linux_x86_install-rpm_en-US.tar.gz.md5

mirrorbrain_getPlatformForMirror( LANGUAGE, SCHEMA )
Linux_x86_install-rpm
mirrorbrain_getFilename( VERSION, LANGUAGE, SCHEMA )
Apache_OpenOffice_incubating_3.4.1_Linux_x86_install-rpm_en-US.tar.gz
mirrorbrain_getExtension( LANGUAGE, SCHEMA ) .tar.gz
hasMirrorLink( LANGUAGE ) true


Google Chrome Info:

Name : google-chrome-beta
Arch : x86_64
Version : 21.0.1180.81
Release : 151980
Size : 125 M
Repo : installed
From repo : google-chrome
Summary : Google Chrome
URL : http://chrome.google.com/
License : Multiple, see http://chrome.google.com/

[DL Website] Prototype of an automatic table with download links

[Marcus (OOo)]

Hi all,

as promized on Friday I've adjusted the existing automatism to the 
lastest AOO 3.4.1 parameters.


The most recent webpage is here:
http://ooo-site.staging.apache.org/download/test/other_print.html

I'm open for opinions. Otherwise I can put it easily to the production 
website as the new "other.html".


Marcus

Re: proposed new directory structure for future releases

[Marcus (OOo)]

Am 08/26/2012 02:26 AM, schrieb Rob Weir:

On Sat, Aug 25, 2012 at 5:36 PM, Kay Schenk  wrote:



On 08/25/2012 01:28 PM, Andrea Pescetti wrote:


On 23/08/2012 22:06, Marcus (OOo) wrote:


Also IMHO no need to separate between the languages.
Furthermore, when you have to check for a potential error, then you you
just need to verify this in a single directory and not more.
/stable/VERSION/



The rest is OK, but removing even the languages subdirectories is
probably too much: we already have 20 languages and they could
theoretically reach 100+, so it would make sense to preserve the
individual subdirectories "en-US", "de", "it" and so on.

Regards,
Andrea.



+1 with Andrea on this one. We should leave the languages in separate
directories. And, if folks do "browsing" on SourceForge instead of using our
download logic, this will make things easier for that purpose as well.



+1

It also makes it easier if someone wants to burn a CD for just a
single language.


OK, so our most recent opinion is the following:

/stable/VERSION//...

Anyone else with different suggestions?

Marcus

[BZ] Please add 3.4.1 as version to all components in BugZilla

[Marcus (OOo)]

Hi BZ admins,

as AOO 3.4.1 is now available, please add the 3.4.1 version into the BZ 
instance of AOO.


Thanks

Marcus

Re: [RELEASE][3.4.1]: update on the current status ...

[Marcus (OOo)]

Am 08/23/2012 10:20 PM, schrieb Marcus (OOo):

Am 08/23/2012 06:07 PM, schrieb Oliver-Rainer Wittmann:

Hi,

On 23.08.2012 17:34, Andre Fischer wrote:

On 23.08.2012 17:06, Andreas Säger wrote:

Am 23.08.2012 15:52, Jürgen Schmidt wrote:

Hi,

as the so called release manager I would like to give an update
where we
are with the release and the final release preparation.

You all know we passed the PPMC and IPMC vote successful, the first
important step towards our release.



The download page suggests:


Download Apache OpenOffice 3.4.1

Click to start downloading the most recent version for Linux Debian
64-bit (DEB) and Deutsch.



which is correct.

When I follow the suggestion I get a file

Apache_OpenOffice_incubating_3.4.1_Linux_x86_install-deb_de.tar.gz


which is not correct. It does not install a runnable program

$ /opt/openoffice.org3/program/soffice
/opt/openoffice.org3/program/soffice: 84:
/opt/openoffice.org3/program/soffice:
/opt/openoffice.org3/program/../basis-link/program/pagein: not found
/opt/openoffice.org3/program/soffice: 1:
/opt/openoffice.org3/program/soffice:
/opt/openoffice.org3/program/../basis-link/ure-link/bin/javaldx: not
found
/opt/openoffice.org3/program/soffice: 120:
/opt/openoffice.org3/program/soffice:
/opt/openoffice.org3/program/soffice.bin: not found


I get the same error message (on Ubuntu 12.04 64bit) but it is
misleading. The
files exist (I checked), but they do not run. So, only one problem to
solve:
fix detection of correct Linux platform.



Together with Andre and Jürgen I had made a change to the download
script.
Please try again.


Thanks for diving in and fixing.

I'm sure that the now commented code part is in since, hm, ever. Were
there also problems in the past? Does someone remember?

Seems to be a case to investigate on the weekend when I've more spare time.


I've looked into the code and the part you have commented out.

It is logical that the DL script return always a 32-bit file because of 
the last "else" clause which is working as a fall-back regardless what 
was set in the "if" statements some line before.


However, I've no clue how that last "else" came in. So, I guess it was 
already online with AOO 3.4.0 or maybe longer. But nobody has noticed 
the error before? Hm, strange. I'm a bit puzzled.


So, thanks again for the fix.

Marcus

Re: Something wrong in download page for Linux users

[Marcus (OOo)]

Am 08/23/2012 04:24 PM, schrieb Ariel Constenla-Haile:


Hi Marcus,

On Fri, May 18, 2012 at 01:49:19AM +0200, Marcus (OOo) wrote:

Am 05/18/2012 01:29 AM, schrieb Ariel Constenla-Haile:

On Fri, May 18, 2012 at 12:26:41AM +0200, Marcus (OOo) wrote:


The system and the browser are 64 bits, the package is 32 bits.


Interesting. The browser shows that the platform is i686 (= x86) and
the user agents says x86_64. Haven't seen this before.

OK, which value is right when you don't know the truth? ;-)


just blame it on Google :)

http://code.google.com/p/chromium/issues/detail?id=44905


Interesting, even Google software has old bugs. :-P


Duplicated by this one?
http://code.google.com/p/chromium/issues/detail?id=128167


Great. When this is solved somewhen, we can check our DL logic
again. I'll add this to the Wiki page.



FYI this is fixed now in Chrome, according to the browser values shown
by http://www.openoffice.org/download/test/analyze.html


Ah, thank you for the hint. I will analyze the data what needs to be 
updated.


BTW:
Great to have this little test webpage online, isn't it? ;-)

Marcus




Variables from the browser  Values

navigator.platform  Linux x86_64
navigator.platform.toLowerCase()linux x86_64
navigator.language  en-US
navigator.userLanguage  undefined
navigator.systemLanguageundefined
navigator.userAgent Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.1 (KHTML, 
like Gecko) Chrome/21.0.1180.81 Safari/537.1
navigator.userAgent.toLowerCase()   mozilla/5.0 (x11; linux x86_64) 
applewebkit/537.1 (khtml, like gecko) chrome/21.0.1180.81 safari/537.1
navigator.javaEnabled() Yes


But the download page is providing 32 bits to download (though the text
says "Click to start downloading the most recent version for Linux
64-bit (RPM and English (US)"):


JavaScript functions from the DL scriptsReturn values
getLink( VERSION, LANGUAGE, MIRROR, SCHEMA )
http://sourceforge.net/projects/openofficeorg.mirror/files/stable/3.4.1/Apache_OpenOffice_incubating_3.4.1_Linux_x86_install-rpm_en-US.tar.gz/download
getArray( LANGUAGE )here,English (US),English 
(US),http://www.openoffice.org/download/other.html,y
getPlatform( LANGUAGE, SCHEMA ) Linux 64-bit (RPM)
getLanguage( LANGUAGE ) English (US)
getLanguageISO( LANGUAGE )  en-US
sourceforge_getLink( VERSION, LANGUAGE, SCHEMA )
http://sourceforge.net/projects/openofficeorg.mirror/files/stable/3.4.1/Apache_OpenOffice_incubating_3.4.1_Linux_x86_install-rpm_en-US.tar.gz/download
apache_getLink( VERSION, LANGUAGE, SCHEMA ) 
http://www.apache.org/dyn/closer.cgi/incubator/ooo/files/stable/3.4.1/Apache_OpenOffice_incubating_3.4.1_Linux_x86_install-rpm_en-US.tar.gz
apache_getChecksum( VERSION, LANGUAGE, SCHEMA, HASH )   
http://www.apache.org/dist/incubator/ooo/files/stable/3.4.1/Apache_OpenOffice_incubating_3.4.1_Linux_x86_install-rpm_en-US.tar.gz.md5
mirrorbrain_getPlatformForMirror( LANGUAGE, SCHEMA )Linux_x86_install-rpm
mirrorbrain_getFilename( VERSION, LANGUAGE, SCHEMA )
Apache_OpenOffice_incubating_3.4.1_Linux_x86_install-rpm_en-US.tar.gz
mirrorbrain_getExtension( LANGUAGE, SCHEMA ).tar.gz
hasMirrorLink( LANGUAGE )   true


Google Chrome Info:

Name: google-chrome-beta
Arch: x86_64
Version : 21.0.1180.81
Release : 151980
Size: 125 M
Repo: installed
 From repo   : google-chrome
Summary : Google Chrome
URL : http://chrome.google.com/
License : Multiple, see http://chrome.google.com/

Re: Minor windows binary branding issue

[Marcus (OOo)]

Am 08/24/2012 09:29 PM, schrieb Issac Goldstand:

I just installed AOO for a friend and noticed that the start menu
created was still "OpenOffice.org".  Whomever hacks on the windows
installer might want to quick fix it.


Thats known and - at least a little bit - wanted.

In order to release the version in time and not to post-pone it for 
weeks we had 2 options:


1) Fix all bugs which are there or reported by users.

2) Fix the severe once only.

Of course, a wrong string in a menu - even when it touches the brand - 
is not that severe than crashes, data loss, etc.


We will what you have posted here and similar bugs step by step, so that 
the next version should look like much more like an own software 
compared with 3.4.0 and 3.4.1.


I hope you can understand this position.

Marcus