Re: Release dates/nomenclature -- Revised proposal

2007-09-05 Thread Torsten Rahn

 October 22, 2007: Tagging Beta4 (Again, a Monday)
 October 30 2007: Release Beta4 (Beta4 is only 2 weeks long)

Having a Beta for two weeks long makes no sense.
Betas are made to get feedback from the people who are meant to test the beta 
release. 
Subtract 8 days for packaging from the 2 weeks and about 3-5 days until people 
manage to get the spare-time to look at the release. Then how much time do 
developers have left to fix the issue reported from the last Beta to get the 
feedback right in time into the next Beta? None. So in that case people will 
report the same bugs for the next Beta. Brilliant.

 November 13 2007: Tagging Release Candidate 1 (move closer to the freeze?)

Same issue there. Too little feedback. Releases are not only there to pressure 
developers to get working but also to incorporate feedback from the testers.

Personally I think the current suggested schedule is the first one that is 
close to being realistic. Except that I think that you won't be able to 
squeeze the release before Dec 20th without risking to delay the release 
highly likely again later on.

Torsten






November 5, 2007: Total Release Freeze (a Monday)

November 20 2007: Release Release Candidate 1

November 21 2007: Tagging Release Candidate 2 (to close to -rc1?)

Definetely too close to gather any useful feedback that 

-- 
 Torsten Rahn

 Tel.: 0 21 61 - 46 43 - 192

credativ GmbH, HRB Mönchengladbach 12080
Hohenzollernstr. 133, 41061 Mönchengladbach
Geschäftsführung: Dr. Michael Meskes, Jörg Folz
___
release-team mailing list
release-team@kde.org
https://mail.kde.org/mailman/listinfo/release-team


Re: Release dates/nomenclature

2007-09-05 Thread Dirk Mueller
On Sunday, 2. September 2007, Helio Chissini de Castro wrote:

 Probably i'm not the first one to say that, but the current issues have
 NOTHING TO DO with tools like krazy, but to lack of we do better management
 of project and failing in to explain to out developers what we need to do
 right now.

Thanks Helio, I think I feel similar. where do you think do we need better 
management? what has to be done?

I`m not asking out of total unawareness, I just want to know what we share the 
same goals, and that everyone on this list does his/her job to help getting 
progress on that. 

Sorry for the short mail, but I`m a bit in a hurry but still want to trigger 
further discussion here. 

Thanks,
Dirk
___
release-team mailing list
release-team@kde.org
https://mail.kde.org/mailman/listinfo/release-team


Re: Release dates/nomenclature -- Revised proposal

2007-09-04 Thread Sebastian Kügler
On Tuesday 04 September 2007 00:51:40 Allen Winter wrote:
 On Monday 03 September 2007 10:51:37 am Sebastian Kügler wrote:
  I've seen no reaction to this email, but I think the current schedule
  (release 20 dec) is broken (Dirk is on vacation, it's too close to
  christmas for PR).
 
  Can I at least get a shut up, you're irrelevant answer, maybe something
  more promising?

 Your counter-counter-proposal is close.
 Except that I try to schedule taggings on Wednesdays.
 That gives us 2 days to recover from BIC Mondays.

 Also, I like Dirk's idea of a libs release in late October.
 The libs release should include kdesupport, kdelibs, and kdepimlibs

 So please try again :)

Thanks. I'll send it shortly, I'm at Linuxconf.eu right now. Expect it 
somewhen during the next two days (I'll give it a shot on the train 
tomorrow).
-- 
sebas

 http://www.kde.org | http://vizZzion.org |  GPG Key ID: 9119 0EF9 


signature.asc
Description: This is a digitally signed message part.
___
release-team mailing list
release-team@kde.org
https://mail.kde.org/mailman/listinfo/release-team


Re: Release dates/nomenclature

2007-09-03 Thread Adriaan de Groot
On Saturday 01 September 2007 23:29, Allen Winter wrote:
 On Saturday 01 September 2007 11:45:07 am Thomas Zander wrote:
  On Saturday 01 September 2007 17:30:34 Matt Rogers wrote:
   That's no different than what we have now. The problem is that people
   seem to be too interested in fixing Krazy issues rather than fixing
   actual bugs. How do you propose we get them interested in fixing real
   bugs?
 
  Stop running the not-so-interresting krazy tests? ;)

 We can pull the plug on the EBN entirely.

That's like turning off compiler warnings because you don't want people to fix 
them either. It's possible, but I'd rather try to rely on the discipline of 
our developers to fix what is needed. The EBN has always been a reporter 
of the last mile of little fixes; I don't think we've ever suggested that 
issues reported by the EBN are related to functionality, real bugs (defects 
encountered at runtime) or releaseability.

The use of the words pull the plug garners a fuck it all, i'll reuse the 
domain for a LOLcats parody response from me. 

 I don't think that will help get bugs fixed, but at least
 it will reduce the number unneeded re-compiles.

 I had no idea Krazy/EBN was doing such so harm to the project.
 That was not the intention.

Communicate that to your developers. Give them priorities. Impress upon them 
the importance not of polishing th internals of the code until they shine, 
but on fixing the big ugly warts on the outside. There's a Sirius Cybernetics 
Corporation segue here, and I'm going to skip it. 

[ade]
___
release-team mailing list
release-team@kde.org
https://mail.kde.org/mailman/listinfo/release-team


Re: Release dates/nomenclature -- Revised proposal

2007-09-03 Thread Sebastian Kügler
I've seen no reaction to this email, but I think the current schedule (release 
20 dec) is broken (Dirk is on vacation, it's too close to christmas for PR).

Can I at least get a shut up, you're irrelevant answer, maybe something more 
promising?

On Saturday 01 September 2007 15:57:57 Sebastian Kügler wrote:
 On Saturday 01 September 2007 12:28:31 Kevin Ottens wrote:
   So let's aim for a release on my mother's birthday, 13th December.
 
  I vote for 6th December, it's my father birthday.

 That would be the tagging date then :-)

 How about this, then. It's mainly what Dirk and Allen proposed.

 Thursday, 6 September: Release Beta2 (already tagged)

 September 24, 2007: Tagging Beta3 (This is a Monday, does that make sense?)
 October 2 2007: Release Beta3

 October 22, 2007: Tagging Beta4 (Again, a Monday)
 October 30 2007: Release Beta4 (Beta4 is only 2 weeks long)

 November 5, 2007: Total Release Freeze (a Monday)

 November 13 2007: Tagging Release Candidate 1 (move closer to the freeze?)
 November 20 2007: Release Release Candidate 1

 November 21 2007: Tagging Release Candidate 2 (to close to -rc1?)
 November 27 2007: Release Release Candidate 2

 December 6  2007: Tagging final Release
 December 13 2007: Targeted Release Date


 * We should probably move the tagging and release for the following betas
   closer to each other, maybe tagging on Fridays, releasing on Tuesdays? Is
   that doable for testing and packaging?

 * Should we move tagging away from Monday? Or when will we suspend the BIC
   Monday?
-- 
sebas

 http://www.kde.org | http://vizZzion.org |  GPG Key ID: 9119 0EF9 


signature.asc
Description: This is a digitally signed message part.
___
release-team mailing list
release-team@kde.org
https://mail.kde.org/mailman/listinfo/release-team


Re: Release dates/nomenclature -- Revised proposal

2007-09-03 Thread Cornelius Schumacher
On Tuesday 04 September 2007 00:51, Allen Winter wrote:

 Also, I like Dirk's idea of a libs release in late October.

I like this idea as well.

 The libs release should include kdesupport, kdelibs, and kdepimlibs

Shouldn't it also include kdebase/runtime, so that it provides everything what 
is needed for third-party apps development?

-- 
Cornelius Schumacher [EMAIL PROTECTED]
___
release-team mailing list
release-team@kde.org
https://mail.kde.org/mailman/listinfo/release-team


Re: Release dates/nomenclature

2007-09-02 Thread Allen Winter
On Saturday 01 September 2007 6:55:57 pm Matt Rogers wrote:
 
 On Sep 1, 2007, at 4:29 PM, Allen Winter wrote:
 
  On Saturday 01 September 2007 11:45:07 am Thomas Zander wrote:
  On Saturday 01 September 2007 17:30:34 Matt Rogers wrote:
  That's no different than what we have now. The problem is that
  people
  seem to be too interested in fixing Krazy issues rather than fixing
  actual bugs. How do you propose we get them interested in fixing
  real
  bugs?
 
  Stop running the not-so-interresting krazy tests? ;)
 
  We can pull the plug on the EBN entirely.
  I don't think that will help get bugs fixed, but at least
  it will reduce the number unneeded re-compiles.
 
  I had no idea Krazy/EBN was doing such so harm to the project.
  That was not the intention.
 
  -Allen
 
 I don't think it does that much harm to the project. In fact, I would
 consider the EBN a good thing. However, it may be a good idea to
 suspend it temporarily. Perhaps the number of recompiles caused by
 people not fixing krazy issues will allow some of us to be more
 productive. It may also spur some of the newer developers to work on
 other things besides fixing Krazy issues, which in the grand scheme
 of things, are less important the closer we get to a freeze. (IMHO)
 

Ok, I turned the Krazy web site off.

I can't really do anything about the command line tool
except tell folks not to use it.

-Allen
___
release-team mailing list
release-team@kde.org
https://mail.kde.org/mailman/listinfo/release-team


Re: Release dates/nomenclature

2007-09-02 Thread Helio Chissini de Castro
On Sunday 02 September 2007, Allen Winter wrote:
 On Saturday 01 September 2007 6:55:57 pm Matt Rogers wrote:
  On Sep 1, 2007, at 4:29 PM, Allen Winter wrote:
   On Saturday 01 September 2007 11:45:07 am Thomas Zander wrote:
   On Saturday 01 September 2007 17:30:34 Matt Rogers wrote:
   That's no different than what we have now. The problem is that
   people
   seem to be too interested in fixing Krazy issues rather than fixing
   actual bugs. How do you propose we get them interested in fixing
   real
   bugs?
  
   Stop running the not-so-interresting krazy tests? ;)
  
   We can pull the plug on the EBN entirely.
   I don't think that will help get bugs fixed, but at least
   it will reduce the number unneeded re-compiles.
  
   I had no idea Krazy/EBN was doing such so harm to the project.
   That was not the intention.
  
   -Allen
 
  I don't think it does that much harm to the project. In fact, I would
  consider the EBN a good thing. However, it may be a good idea to
  suspend it temporarily. Perhaps the number of recompiles caused by
  people not fixing krazy issues will allow some of us to be more
  productive. It may also spur some of the newer developers to work on
  other things besides fixing Krazy issues, which in the grand scheme
  of things, are less important the closer we get to a freeze. (IMHO)

 Ok, I turned the Krazy web site off.

 I can't really do anything about the command line tool
 except tell folks not to use it.

 -Allen

Please, no.

I was planning to use it in a next week presentation in a major conference 
here an one of the important things introduced during our evolution.
 
Despite this, turn off Krazy is just a way to make more people talking 
about what's  happening. 

Sorry, but by blaming Krazy for the fact that we're lack of resources or for 
try force people to do something is almost the same to cut off the liberty of 
developers decide by thenselves what they want to do.

Probably i'm not the first one to say that, but the current issues have 
NOTHING TO DO with tools like krazy, but to lack of we do better management 
of project and failing in to explain to out developers what we need to do 
right now.

Seriously, shutting down krazy is more signal of weakness and confusion than a 
really help to project. PLease, get it online again.

My 2c's

-- 
Helio Chissini de Castro
KDE Developer
Brasil/South America Primary Contact
___
release-team mailing list
release-team@kde.org
https://mail.kde.org/mailman/listinfo/release-team


Re: Release dates/nomenclature

2007-09-02 Thread Andreas Pakulat
On 02.09.07 17:44:25, Helio Chissini de Castro wrote:
 On Sunday 02 September 2007, Allen Winter wrote:
  Ok, I turned the Krazy web site off.
 
  I can't really do anything about the command line tool
  except tell folks not to use it.
 
  -Allen
 
 Please, no.
 
 I was planning to use it in a next week presentation in a major conference 
 here an one of the important things introduced during our evolution.
  
 Despite this, turn off Krazy is just a way to make more people talking 
 about what's  happening. 
 
 Sorry, but by blaming Krazy for the fact that we're lack of resources or for 
 try force people to do something is almost the same to cut off the liberty of 
 developers decide by thenselves what they want to do.
 
 Probably i'm not the first one to say that, but the current issues have 
 NOTHING TO DO with tools like krazy, but to lack of we do better management 
 of project and failing in to explain to out developers what we need to do 
 right now.
 
 Seriously, shutting down krazy is more signal of weakness and confusion than 
 a 
 really help to project. PLease, get it online again.

I completely agree and there may be krazy issues that should be fixed
_before_ the release - for example dpointerness and maybe  some others. 

I also don't think this helps getting developers into fixing bugs, those
that were looking at the ebn and fixing krazy issues might as well just
use the commandline tool. IMHO there's a reason these people don't work
on real bugs but krazy stuff (in my case the factor is time, if I only
have some 30 minutes I rather fix a bunch of krazies instead of starting
something which I can't finish).

Andreas

-- 
Chess tonight.
___
release-team mailing list
release-team@kde.org
https://mail.kde.org/mailman/listinfo/release-team


Re: Release dates/nomenclature

2007-09-02 Thread Pradeepto Bhattacharya
Hi,

On 9/3/07, Allen Winter [EMAIL PROTECTED] wrote:
 There are several issues:
  1) people with limited experience trying to fix non-trivial stuff and making 
 things worse
  2) people fixing trivial stuff which causes lots of extra recompiles
  3) people getting their hands in places they don't belong

 All this causes slower development and unnecessary frustration, perhaps at 
 the cost
 of quality; but, maybe a .0 release isn't the time to be focusing on fine 
 polishing.

 I don't want to hurt the project.  Doctor, do no harm.

 We can revisit the issue for the point-releases.

 Maybe, just maybe you can turn off some tests and keep
the others running instead of this blanket shutdown of all tests ,
non? I mean, I am sure we can arrive at an consensus of which tests
are important for the long run of the project and important even for
our short term goals. Say for e.g. (imho), typo checks are not exactly
something very critical, right?

  So maybe people come to common ground here and decide
which Krazy tests can be suspended temporarily and which can run.

  Cheers!

Pradeepto
-- 
The KDE Project : http://www.kde.org
KDE India : http://in.kde.org
Mailing List : http://mail.kde.org/mailman/listinfo/kde-india
___
release-team mailing list
release-team@kde.org
https://mail.kde.org/mailman/listinfo/release-team


Re: Release dates/nomenclature

2007-09-02 Thread Thomas Zander
On Monday 03 September 2007 04:57:08 Pradeepto Bhattacharya wrote:
               So maybe people come to common ground here and decide
 which Krazy tests can be suspended temporarily and which can run.

Which do you propose?

-- 
Thomas Zander


pgptCUNjAFPFE.pgp
Description: PGP signature
___
release-team mailing list
release-team@kde.org
https://mail.kde.org/mailman/listinfo/release-team


Re: Release dates/nomenclature

2007-09-02 Thread Thomas Zander
On Monday 03 September 2007 02:11:06 Allen Winter wrote:
 There are several issues:
  1) people with limited experience trying to fix non-trivial stuff and
 making things worse 2) people fixing trivial stuff which causes lots of
 extra recompiles 3) people getting their hands in places they don't
 belong

I agree with this;
the mindset is still very much on polish before the codebase closes.  
Never mind we froze some time ago, people are still doing it.
With krazy showing all this low hanging fruit people get an urgency 
of Oh, there is so much left to do! but its all of the wrong kind. In 
the end people apparently become quite bad at simple procrastination :)

so shutting down at least the discovery of bugs that are not relevant in 
this phase of development (freeze and all) makes sense, and can be 
explained to people under the guise of; what use is it to report bugs 
that you shouldn't be fixing right now anyway.
-- 
Thomas Zander


pgpvDburItnXY.pgp
Description: PGP signature
___
release-team mailing list
release-team@kde.org
https://mail.kde.org/mailman/listinfo/release-team


Re: Release dates/nomenclature

2007-09-01 Thread Matt Rogers
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1


On Sep 1, 2007, at 4:29 PM, Allen Winter wrote:

 On Saturday 01 September 2007 11:45:07 am Thomas Zander wrote:
 On Saturday 01 September 2007 17:30:34 Matt Rogers wrote:
 That's no different than what we have now. The problem is that  
 people
 seem to be too interested in fixing Krazy issues rather than fixing
 actual bugs. How do you propose we get them interested in fixing  
 real
 bugs?

 Stop running the not-so-interresting krazy tests? ;)

 We can pull the plug on the EBN entirely.
 I don't think that will help get bugs fixed, but at least
 it will reduce the number unneeded re-compiles.

 I had no idea Krazy/EBN was doing such so harm to the project.
 That was not the intention.

 -Allen

I don't think it does that much harm to the project. In fact, I would  
consider the EBN a good thing. However, it may be a good idea to  
suspend it temporarily. Perhaps the number of recompiles caused by  
people not fixing krazy issues will allow some of us to be more  
productive. It may also spur some of the newer developers to work on  
other things besides fixing Krazy issues, which in the grand scheme  
of things, are less important the closer we get to a freeze. (IMHO)

- --
Matt


-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.6 (Darwin)

iD8DBQFG2e3+A6Vv5rghv0cRAnTNAJ0eg6w27VymESIEBMvPaNPIesQN6QCghdL2
3oAU7O6giP/QJcgxz7eiQog=
=8B7D
-END PGP SIGNATURE-
___
release-team mailing list
release-team@kde.org
https://mail.kde.org/mailman/listinfo/release-team


Re: Release dates/nomenclature

2007-08-31 Thread Allen Winter
On Friday 31 August 2007 1:13:17 pm Allen Winter wrote:
 On Friday 31 August 2007 1:07:32 pm Dirk Mueller wrote:
  On Wednesday, 29. August 2007, Allen Winter wrote:
  
defect in my mind, 'cause some people can actually plan all releases up
to 17 januari ;-) )
   Of course this all assumes that Dirk will around at Christmas time to
   actually to the tagging. :)
  
  I think it is highly unrealistic to get anything done by December 20th. My 
  holidays season starts December 15th btw, although it is slightly possible 
  that I could stay around for a few days. 
  
  From past years, I know that it will be unlikely to have enough developers 
  around that time to actually get release critical bugs fixed. 
  
  if we can`t get it ready tagging and announcement by December 15th, then we 
  should delay it into next year. 
  
 
 OK, I'll do some date-squeezing and move things back two weeks.
 How does December 12th sound?
 

Countering my proprosal, how about Dec 6?
We can make the Beta4 only 2 weeks long.
___
release-team mailing list
release-team@kde.org
https://mail.kde.org/mailman/listinfo/release-team


Re: Release dates/nomenclature

2007-08-31 Thread Dirk Mueller
On Wednesday, 29. August 2007, Allen Winter wrote:

  defect in my mind, 'cause some people can actually plan all releases up
  to 17 januari ;-) )
 Of course this all assumes that Dirk will around at Christmas time to
 actually to the tagging. :)

I think it is highly unrealistic to get anything done by December 20th. My 
holidays season starts December 15th btw, although it is slightly possible 
that I could stay around for a few days. 

From past years, I know that it will be unlikely to have enough developers 
around that time to actually get release critical bugs fixed. 

if we can`t get it ready tagging and announcement by December 15th, then we 
should delay it into next year. 


Dirk
___
release-team mailing list
release-team@kde.org
https://mail.kde.org/mailman/listinfo/release-team


Re: Release dates/nomenclature

2007-08-31 Thread Dirk Mueller
On Friday, 31. August 2007, Allen Winter wrote:

  OK, I'll do some date-squeezing and move things back two weeks.
  How does December 12th sound?
 Countering my proprosal, how about Dec 6?
 We can make the Beta4 only 2 weeks long.

what is part of the December release (12th or 6th doesn't count)?

I think if we delay the KDE desktop release into december (which I agree 
with), then we should release libs earlier. I think it is possible to release 
the libs by the end of october like planned. 

Releasing the libs earlier has two advantages: 

a) it is possible to add new features needed for apps without risking binary 
compatibility burdens becase we do a mistake last minute

b) we somehow meet our deadline. We're slipping already bad, and slipping has 
to stop. 

c) 3rd party applications who have already ported (digikam, amarok, ktorrent, 
there are probably many others) have something to require and build on. it 
wouldn't be bad for apps to be available before the desktop release either. 


Release date juggling aside, I'm quite concerned about the recent trend of 
blogging how bad KDE4 is. It is time to advertise that we have to eat our own 
dogfood, and that KDE4 will stay a vision forever if developers don't start 
to use it and fix the glaring bugs. I know that part of the pain is that 
there is no stable mail client for KDE4, and no IRC client at all. These are 
things that we should perhaps somehow address ;)

I can only say that I tried to switch my development desktop to KDE4, and it 
is still very painful, as barely anything works, and I'm busy compiling after 
krazy check changes instead of getting things done. This has to stop! It 
takes me almost two work days go get things recompiled, and I cannot do 
something during that time on it. 

This is a much more important topic than juggling tagging dates around. If we 
continue like that, we loose the bugfixers because there is no point in 
fixing KDE3.x bugs anymore, and we loose the app code monkeys, because there 
is no foundation they could base their work on. 



Dirk
___
release-team mailing list
release-team@kde.org
https://mail.kde.org/mailman/listinfo/release-team


Re: Release dates/nomenclature

2007-08-28 Thread Thomas Zander
On Tuesday 28 August 2007 16:36:34 Thomas Zander wrote:
 Because of the new reality that the 'customers' of KDE are not just the
 users downloading packages. Its also users that want to be able to get
 a kubuntu CD shipped to them by email.

ugh, shipped by snail mail naturally.

-- 
Thomas Zander


pgpIHt0PQM75e.pgp
Description: PGP signature
___
release-team mailing list
release-team@kde.org
https://mail.kde.org/mailman/listinfo/release-team


Re: Release dates/nomenclature

2007-08-28 Thread Jason Harris
Hello,

On Monday 27 August 2007 15:54, Troy Unrau wrote:
 So, I've brought this proposal up once before, but here it is again

Is it too early to declare consensus on this?  The proposed schedule (as 
revised by Allen) seems like a good plan to me.  I think it's important that 
the 4.0 release is well received, and I don't think people will be convinced 
by the 4.0.0 != 4.x line.  

regards,
Jason

-- 
KStars: KDE Desktop Planetarium
[EMAIL PROTECTED]
___
release-team mailing list
release-team@kde.org
https://mail.kde.org/mailman/listinfo/release-team