Re: Supertesters - Make, accept, nominations

[Neal H. Walfield]

At Thu, 16 Feb 2012 14:44:07 +0100,
Pali Rohár wrote:
> We (Maemo Community) really need people with ability to promote packages 
> which 
> do not maintain (e.g not on maintainer list). See how log took to promote 
> package rootsh, version 1.8 to Extras (more than year?). Old version in 
> Extras 
> had critical bug which cause not possible to uninstall package. And 
> maintainer 
> abandoned Maemo community, so nobody had permission for it!
> 
> I would like to see permission for promoting packages which I update and not 
> wait while old maintainer give me needed permisstion!

This is a good point.  It needs to be easier to adopt unmaintained
packages.
___
maemo-developers mailing list
maemo-developers@maemo.org
https://lists.maemo.org/mailman/listinfo/maemo-developers

Re: promotion to extras

[Neal H. Walfield]

Hi,

There has been support for and no objections to the below proposal.
How do we go about implementing this?

Thanks,

Neal

At Tue, 13 Dec 2011 14:00:54 +0100,
Neal H. Walfield wrote:
> 
> Thanks to everyone for the feedback.  I'm happy to learn that there is
> already infrastructure for differentiating super-testers from normal
> tesers.  I think promoting orphaned packages is a different discussion
> and I think this should be decided seperately.
> 
> My new proposal is:
> 
>  - quarantine for 10 days, promote after: 1 tester + 1 super-tester approve
>  - quarantine for 10 days, promote after: 3 testers approve
> 
> Thanks,
> 
> Neal
> ___
> maemo-developers mailing list
> maemo-developers@maemo.org
> https://lists.maemo.org/mailman/listinfo/maemo-developers
> 
___
maemo-developers mailing list
maemo-developers@maemo.org
https://lists.maemo.org/mailman/listinfo/maemo-developers

Re: promotion of orphaned packages

[Neal H. Walfield]

At Wed, 14 Dec 2011 13:32:47 +0100,
Nicolai Hess wrote:
> Yes, but who decides to transfer the maintainance? Can this only be done by
> the current
> maintainer?
> I doubt most maintainers of orphaned packages are listening to this
> request.

As I understand it, the maintainer gets an email.  If they ignore it
for, say, a week, I bet you could ask Niels Breet to authorize the
request.

Neal

___
maemo-developers mailing list
maemo-developers@maemo.org
https://lists.maemo.org/mailman/listinfo/maemo-developers

Re: promotion of orphaned packages

[Neal H. Walfield]

Hi Nicolai,

> Is it possible to transfer the maintainance for orphaned packages to other,
> still active, developers?
> If so, I would volunteer myself for maintainance of some orphaned packages.

If you look at the package's page (e.g.,
http://maemo.org/packages/view/feedingit/), there is a "Request to be
maintainer of this package" button.

Neal
___
maemo-developers mailing list
maemo-developers@maemo.org
https://lists.maemo.org/mailman/listinfo/maemo-developers

Re: promotion to extras

[Neal H. Walfield]

Thanks to everyone for the feedback.  I'm happy to learn that there is
already infrastructure for differentiating super-testers from normal
tesers.  I think promoting orphaned packages is a different discussion
and I think this should be decided seperately.

My new proposal is:

 - quarantine for 10 days, promote after: 1 tester + 1 super-tester approve
 - quarantine for 10 days, promote after: 3 testers approve

Thanks,

Neal
___
maemo-developers mailing list
maemo-developers@maemo.org
https://lists.maemo.org/mailman/listinfo/maemo-developers

promotion to extras

[Neal H. Walfield]

Hi,

I'm afraid that extras is dying due to a lack of interest in testing
packages.

Last month, I uploaded some Woodchuck-related packages to
extras-devel.  I asked on #maemo a few times for people to test my
packages.  I asked on this mailing list.  And, I asked on
talk.maemo.org.  In the end, most of the testers were people who I
know personally and asked to do me a favor and vote for my package.  I
doubt that they actually tested it.  This is a waste of time: it
doesn't improve the quality of packages and my time could be better
spent working on my package.

There is one person who appears to test packages: Harald Schmidt.  He
has been designated a tester and appears to periodically check
extras-testing for new packages.

Given this, I propose that we use the following criteria for promoting
to extras:

 - 10 days incubation (unchanged)
 - maintainer + 1 designated "testers" (e.g., Harald Schmidt) thumbs up
   or maintainer + 3 non-testers

Thoughts?

Neal

___
maemo-developers mailing list
maemo-developers@maemo.org
https://lists.maemo.org/mailman/listinfo/maemo-developers

Re: Woodchuck package review

[Neal H. Walfield]

Thanks for the suggestion.  I prefer reading mailing lists and blogs
and I assume everyone is like me :-).  I've created a thread on TMO:

  http://talk.maemo.org/showthread.php?t=79707

Neal

At Sat, 05 Nov 2011 12:33:06 -0200,
Roberto Colistete Jr. wrote:
> 
>  Hi Neal,
> 
>  Why not announce Woodchuck on Talk Maemo.org (section "Applications") ?
> 
>  Hundreds of users will know about it.
> 
>  Best regards,
> 
>          Roberto
> 
> Em 05-11-2011 10:50, Neal H. Walfield escreveu:
> > Hi, All,
> >
> > I'd like to upload Woodchuck to extras.  Woodchuck is stable and
> > usable and I think that it offers great value to users of mobile
> > devices.  Since my favorite programs are now Woodchuck-enabled, I've
> > observed a change in my behavior.  Because, gPodder, FeedingIt and
> > Khwetteur automatically download content, I rely a lot less on my
> > cellular provider, and, since the data I want is already there, the
> > content is more quickly available and poor cellular coverage rarely
> > bothers me.
> >
> > To get Woodchuck into extras, I need 6 people to review the packages.
> > Here are the links to the packages:
> >
> >
> > http://maemo.org/packages/package_instance/view/fremantle_extras-testing_free_armel/murmeltier/0.4~2002-7/
> >
> > http://maemo.org/packages/package_instance/view/fremantle_extras-testing_free_armel/pywoodchuck/0.4~2002-7/
> >
> > http://maemo.org/packages/package_instance/view/fremantle_extras-testing_free_armel/apt-woodchuck/0.4~2002-7/
> >
> > http://maemo.org/packages/package_instance/view/fremantle_extras-testing_free_armel/smart-storage-logger/0.4~2002-7/
> >
> > http://maemo.org/packages/package_instance/view/fremantle_extras-testing_free_armel/libgwoodchuck/0.4~2002-7/
> >
> > Thanks for your help!
> >
> > Neal
> > ___
> > maemo-developers mailing list
> > maemo-developers@maemo.org
> > https://lists.maemo.org/mailman/listinfo/maemo-developers
> 
> 
___
maemo-developers mailing list
maemo-developers@maemo.org
https://lists.maemo.org/mailman/listinfo/maemo-developers

Re: Woodchuck package review

[Neal H. Walfield]

Hi, Andrew,

At Sat, 5 Nov 2011 13:00:20 +,
and...@bleb.org wrote:
> For the packahes which aren't in "Section: user/..." they will be
> promoted automatically as dependencies of things which relu on them.

I wasn't aware of that, thanks!  That means there are only three
packages that need review:

  
http://maemo.org/packages/package_instance/view/fremantle_extras-testing_free_armel/murmeltier/0.4~2002-7/
  
http://maemo.org/packages/package_instance/view/fremantle_extras-testing_free_armel/apt-woodchuck/0.4~2002-7/
  
http://maemo.org/packages/package_instance/view/fremantle_extras-testing_free_armel/smart-storage-logger/0.4~2002-7/

Thanks,

Neal
___
maemo-developers mailing list
maemo-developers@maemo.org
https://lists.maemo.org/mailman/listinfo/maemo-developers

Woodchuck package review

[Neal H. Walfield]

Hi, All,

I'd like to upload Woodchuck to extras.  Woodchuck is stable and
usable and I think that it offers great value to users of mobile
devices.  Since my favorite programs are now Woodchuck-enabled, I've
observed a change in my behavior.  Because, gPodder, FeedingIt and
Khwetteur automatically download content, I rely a lot less on my
cellular provider, and, since the data I want is already there, the
content is more quickly available and poor cellular coverage rarely
bothers me.

To get Woodchuck into extras, I need 6 people to review the packages.
Here are the links to the packages:

  
http://maemo.org/packages/package_instance/view/fremantle_extras-testing_free_armel/murmeltier/0.4~2002-7/
  
http://maemo.org/packages/package_instance/view/fremantle_extras-testing_free_armel/pywoodchuck/0.4~2002-7/
  
http://maemo.org/packages/package_instance/view/fremantle_extras-testing_free_armel/apt-woodchuck/0.4~2002-7/
  
http://maemo.org/packages/package_instance/view/fremantle_extras-testing_free_armel/smart-storage-logger/0.4~2002-7/
  
http://maemo.org/packages/package_instance/view/fremantle_extras-testing_free_armel/libgwoodchuck/0.4~2002-7/

Thanks for your help!

Neal
___
maemo-developers mailing list
maemo-developers@maemo.org
https://lists.maemo.org/mailman/listinfo/maemo-developers

Re: announcing libconnui-dev-0.1 and operator-name-cbs-widget-0.1

[Neal H. Walfield]

At Fri, 28 Oct 2011 21:24:06 +0300,
Aapo Rantalainen wrote:
> 
> 2011/10/28 Neal H. Walfield :
> > As far as I can tell, your code is under the GPLv2 or later.  If that
> > code is using a GPL-incompatible library, you've got a problem.
> Hmm,
> GPL code can dynamically use GPL-incompatible libraries legally. Think
> about GPL game using (closed source) DirectX.
> Or GPL QT-application running on Windows.

Running on Windows is not a problem: if you just make system calls or
perform an IPC, you haven't created a derivative work.  If you mean a
program that uses essential system libraries, that use falls under the
system library exception:

  http://www.gnu.org/licenses/gpl-faq.html#GPLIncompatibleLibs

I don't know if DirectX would be included by this exception (I don't
know how Direct X works).

Perahps the use of libconnui would be covered by this exception.

Neal

___
maemo-developers mailing list
maemo-developers@maemo.org
https://lists.maemo.org/mailman/listinfo/maemo-developers

Re: announcing libconnui-dev-0.1 and operator-name-cbs-widget-0.1

[Neal H. Walfield]

At Sat, 29 Oct 2011 01:25:30 +0800,
Jonathan Wilson wrote:
> Its a set of header files to allow developers to use the functionality 
> inside these closed-source libraries, some of the functionality seems 
> useful to me (bits of functionality are being used in 
> operator-name-cbs-widget-0.1) and I am guessing other devs may also like it.

As far as I can tell, your code is under the GPLv2 or later.  If that
code is using a GPL-incompatible library, you've got a problem.

Neal
___
maemo-developers mailing list
maemo-developers@maemo.org
https://lists.maemo.org/mailman/listinfo/maemo-developers

Re: Looking for anyone with a N900 running Maemo5 to run a shell script & get some diagnostics for me

[Neal H. Walfield]

Hi, Jonathan,

I'm happy to help, but you've got to make it easier: I needed to
manually fix the line breaks.

At Thu, 27 Oct 2011 00:19:41 +0800,
Jonathan Wilson wrote:
> 
> What I want is the output from this shell script.

get_service_provider_name
method return sender=:1.20 -> dest=:1.318 reply_serial=2
   string "blau"
   uint32 0
   uint32 0
   int32 0
get_service_provider_info
method return sender=:1.20 -> dest=:1.319 reply_serial=2
   array [
   ]
   int32 1003
country 262
operator 3
NET_NITZ_FULL_OPER_NAME
method return sender=:1.20 -> dest=:1.321 reply_serial=2
   string ""
   int32 1000
NET_HARDCODED_LATIN_OPER_NAME
method return sender=:1.20 -> dest=:1.322 reply_serial=2
   string "E-Plus"
   int32 0


> I also want to know what 
> the name of your service provider is (i.e. the company you pay for
> service) 

blau

> as well as the name of your operator (i.e. the company that actually 
> operates the cell towers and/or matches the MNC/MCC that get output by the 
> script as "country" and "operator")

E-Plus

> and I want to know whether you were 
> connected to 2G or 3G when you ran the script

3G

> as well as what operator name 
> was displayed on the home screen

blau

> and whether you were roaming or on your 
> home operator at the time.

home operator

> (if there is someone out there who is in a 
> position to get output from when you are roaming rather than on your home 
> network, that would be nice)

Looking through my logs, here is what I say when roaming in Austria:

method call sender=:1.33 -> dest=com.nokia.phone.net serial=305
   path=/com/nokia/phone/net; interface=Phone.Net;
   member=get_operator_name
   byte 0
   uint32 1
   uint32 262
method return sender=:1.20 -> dest=:1.33 reply_serial=305
   string "T-Mobile D"
   int32 0

I didn't find any calls to get_service_provider_name or
get_service_provider_info.

HTH
Neal
___
maemo-developers mailing list
maemo-developers@maemo.org
https://lists.maemo.org/mailman/listinfo/maemo-developers

Re: Looking for some help with autoconf/automake/dpkg (and general feedback on my code)

[Neal H. Walfield]

At Tue, 25 Oct 2011 14:07:13 +0800,
Jonathan Wilson wrote:
> I suspect also I need more in the postinst file in terms of verifying the 
> correct binary and things, any help would be appreciated.

You should never modify a binary in place.  Instead, copy the binary,
modify it and then atomically replace it using mv.  If a binary is in
use by a program and not all of it is loaded in memory and you modify
it, it may end up using neither the original nor the new version but
some inconsistent one...

Neal
___
maemo-developers mailing list
maemo-developers@maemo.org
https://lists.maemo.org/mailman/listinfo/maemo-developers

Re: Looking for some help with autoconf/automake/dpkg (and general feedback on my code)

[Neal H. Walfield]

At Tue, 25 Oct 2011 14:07:13 +0800,
Jonathan Wilson wrote:
> 
> I have uploaded my code and files to
> http://www.cncmods.net/files/operator-name-cbs-widget-0.1.zip
> 
> I am looking for general feedback on this code and packaging but 
> specifically I am having problems getting it to create, include in the .deb 
> file and run the cbspatch binary

I find cdbs to be very easy to use.  This will likely be enough for
your debian/rules file (it works for my autoconf-based project):

 #!/usr/bin/make -f
 
 include /usr/share/cdbs/1/rules/debhelper.mk
 include /usr/share/cdbs/1/class/autotools.mk

Check out

http://hssl.cs.jhu.edu/~neal/woodchuck/src/branches/master/#files:debian

> I suspect also I need more in the postinst file in terms of verifying the 
> correct binary and things, any help would be appreciated.

I'd check the md5sum of the binary, particularly as cbspatch doesn't
check for any signatures.

Also, cbspatch does not return 0, so your postinst script might fail
spuriously.

Neal

___
maemo-developers mailing list
maemo-developers@maemo.org
https://lists.maemo.org/mailman/listinfo/maemo-developers

Re: looking for info from Phone.SIM.get_service_provider_info, Phone.SIM.get_service_provider_name and getting the proper operator name

[Neal H. Walfield]

At Mon, 24 Oct 2011 14:49:42 +0800,
Jonathan Wilson wrote:
> 
> I am trying to find out how to get the proper operator name as displayed by 
> the home widget.
> 
> I know about Phone.Net.operator_name_change and Phone.Net.get_operator_name 
> and the other Phone.Net calls but that returns the actual operator (in my 
> case "YES OPTUS") and not the service provider (in my case its "TPG")
> 
> The stock operator widget uses calls to Phone.SIM.get_service_provider_info 
> and Phone.SIM.get_service_provider_name to get the correct service provider 
> but I cant find any info on what those dbus calls do or return.

At Mon, 24 Oct 2011 17:38:41 +0800,
Jonathan Wilson wrote:
> 
> I run
> dbus-send --system --type=method_call --print-reply 
> --dest=com.nokia.phone.SIM /com/nokia/phone/SIM 
> Phone.Sim.get_service_provider_name
> and get
> string "TPG"
> uint32 0
> uint32 0
> int32 0

I pay blau (my operator?) who just is a reseller for the eplus network
(my service provider?).

   string "blau"
   uint32 0
   uint32 0
   int32 0

This is what is shown my the home widget.

> I run
> dbus-send --system --type=method_call --print-reply 
> --dest=com.nokia.phone.SIM /com/nokia/phone/SIM 
> Phone.Sim.get_service_provider_info
> and get
> array [
> ]
> int32 1003

   array [
   ]
   int32 1003

> Problem is, there is no information available as to what the parameters are 
> (other than the operator name obviously)
> Hence why I need information.
> 
> Perhaps if other people post the output from running the same commands, it 
> might make it easier to figure out what we are looking at.
> That or someone needs to reverse engineer libconnui_cellular.so :)

My suggestion was to eavesdrop on what the plug-in is doing.  You
could just eaves drop on the whole system bus and then grep of the
strings you are looking for.  I have a few weeks worth of such logs
and I found one relevant reference:

  method call sender=:1.33 -> dest=com.nokia.phone.net serial=293
  path=/com/nokia/phone/net; interface=Phone.Net;
  member=get_operator_name
 byte 0
 uint32 3
 uint32 262
  method return sender=:1.20 -> dest=:1.33 reply_serial=293
 string "E-Plus"
 int32 0


HTH,

Neal
___
maemo-developers mailing list
maemo-developers@maemo.org
https://lists.maemo.org/mailman/listinfo/maemo-developers

Re: looking for info from Phone.SIM.get_service_provider_info, Phone.SIM.get_service_provider_name and getting the proper operator name

[Neal H. Walfield]

At Mon, 24 Oct 2011 14:49:42 +0800,
Jonathan Wilson wrote:
> The stock operator widget uses calls to Phone.SIM.get_service_provider_info 
> and Phone.SIM.get_service_provider_name to get the correct service provider 
> but I cant find any info on what those dbus calls do or return.

I guess you tried eavesdropping.  What kind of problems did you have?
Could you capture the traffic, but not decode the output?  Could you
post the traffic here.  May be someone could help you.

Neal
___
maemo-developers mailing list
maemo-developers@maemo.org
https://lists.maemo.org/mailman/listinfo/maemo-developers

Woodchuck-Related User Behavior and Network Environment Study

[Neal H. Walfield]

Hello world!

We at the Hopkins Storage Systems Lab (HSSL) are trying to understand
how you, the tech elite, access and use data and how you access the
Internet on mobile devices.  We are interested in you, because we
think that your data-use habits may be indicative of what might become
common in a few years time.

We want to understand how you access and use data on mobile devices to
improve the user experience on mobile devices.  Specifically, we want
to:

  - Improve disconnected operation;
  - Make accessing data faster;
  - Increase battery life;
  - Reduce network connectivity costs; and,
  - Simplify data management.

We suspect that significant amounts of data that you use are
downloaded on demand and that this data could be effectively
prefetched.  Although prefetching sounds easy enough, there are a
number of issues that need to be considered: when should data be
prefetched? what data should be prefetched? how do we avoid exhausting
free space? how do we enable applications to coordinate the use of
shared resources?

To this end, we are conducting a user study.  We'd like you to
participate by running our data collection software, which gathers
information about the data you use, your network connectivity, and
your battery use.

To help by running the data collection software, which should take
about 10 minutes to install and not require any further interactions
on your part, please visit, on your N900:

  http://hssl.cs.jhu.edu/~neal/woodchuck/smart-storage-logger.install

or:

  http://tinyurl.com/wcssl

For more information about Woodchuck, please visit:

  http://hssl.cs.jhu.edu/~neal/woodchuck/

Much of the data that we collect will be anonymized.  No personally
identifying data will be published.  Data collection will last for
approximately one year.

Anyone with a compatible device may run the data collection software.
Your participation in this experiment is entirely voluntary.  Should
you choose to participate, your data will be kept confidential to the
extent possible by law.  Only researchers involved in this study will
see collected data.  Published data will not include identifying
artifacts (i.e., we will make every effort to prevent the identify of
participants from being determined from the data we publish).
Encryption will be used to transfer collected data and to verify the
server to which that data is uploaded.

If you have any questions or concerns, feel free to contact me at
n...@cs.jhu.edu or Randal Burns, the principle investigator, at
ran...@cs.jhu.edu.  Your assistance in helping us meet our research
goals would be greatly appreciated.


Thanks for your help!

Neal Walfield


P.S. This study is research.  You will not receive any direct benefits
from participating in this study.  This study may benefit society if
the results lead to a better understanding of how data is used on
mobile devices.  The study is taking place at the Whiting School of
Engineering at the Johns Hopkins University in the United States.  The
principle investigator is Randal Burns:

  Email: randal(at)cs.jhu.edu
  Phone: 410.516.7708
  Mailing Address:
Department of Computer Science 
The Johns Hopkins University 
222 New Engineering Building 
Baltimore, MD 21218
USA

Approved by HIRB on November 18, 2010
HIRB Study number: 111910
___
maemo-developers mailing list
maemo-developers@maemo.org
https://lists.maemo.org/mailman/listinfo/maemo-developers

Re: gpg-agent in scratchbox?

[Neal H. Walfield]

Hi, Rainer,

At Sun, 2 Oct 2011 22:59:17 +0200,
Rainer Dorsch wrote:
> when building my packages (navit), dpkg-buildpackage asks for my gpg 
> passphrase. Outstide of scratchbox this is handled through gpg-agent, but 
> inside scratchbox that does not seem to work.
> 
> Do all package builders enter manually their passphrases, or does sombody 
> have 
> a working gpg-agent setup? If yes, can you please provide details?

I don't use gpg-agent in scratchbox.  Instead, I pass '-us -uc' to
dpkg-buildpackage to tell it not to sign the .dsc and .changes file.
For builds that I want to actually publish, I copy the .dsc and
.changes files over to the computer with my gpg key and use debsign to
generate the signatures.

HTH,

Neal
___
maemo-developers mailing list
maemo-developers@maemo.org
https://lists.maemo.org/mailman/listinfo/maemo-developers

Woodchuck: A Download Manager for Delay Tolerant Data

[Neal H. Walfield]

Hi,

I've recently released the first publically announced version of
Woodchuck, a download manager for delay tolerant data, such as blogs,
podcasts, email, social network updates, weather reports, and
calendaring data.  Woodchuck aims to save battery power, manage data
caps and hide spotty network coverage and high latency.  It does this
by monitoring network conditions and user behavior and, when
conditions are good, tells applications (starting them if necessary)
to perform an update.

  http://hssl.cs.jhu.edu/~neal/woodchuck/blog/2011/08/02/version_0.3_released/

Binary packages for the N900 are available:

  http://hssl.cs.jhu.edu/~neal/woodchuck/woodchuck.install

Woodchuck requires application support.  To gain experience with the
API I've ported FeedingIt, an RSS reader:

  http://hssl.cs.jhu.edu/~neal/woodchuck/feedingit.install

I am currently finishing the port of gPodder, and plan to port a few
more applications to make sure the API is sound.

I'm sending this email for two reasons.  First, I'm looking for
feedback on the API (and the project in general).  Second, as
Woodchuck's effectiveness is proportional to the number of
applications that actually use it, I'd like to offer to help people
port their applications to use Woodchuck.  Extensive documentation of
the API, the glib-based C library, and the Python modules can be found
here:

  http://hssl.cs.jhu.edu/~neal/woodchuck/doc/


Thanks,

Neal
___
maemo-developers mailing list
maemo-developers@maemo.org
https://lists.maemo.org/mailman/listinfo/maemo-developers

Re: Best development environment for n810

[Neal H. Walfield]

At Tue, 26 Jul 2011 16:46:00 +0200,
warnh...@zedat.fu-berlin.de wrote:
> I've got an n810 some time ago and now want to spend some time in
> developing some small applications (DB (sqlite) with python. The tablet is
> OS2008 (4.0) and the development box is Debian squeeze x86-64. What will
> be the best choice? Upgrading the tablet, to what version, caveats with
> desktop pc, using a virtual box?

I personally find developing python easiest directly on the device.  I
use emacs' tramp mode to get my usual development environment.

Neal
___
maemo-developers mailing list
maemo-developers@maemo.org
https://lists.maemo.org/mailman/listinfo/maemo-developers

Re: Can anyone tell me how the N900 vibrator driver works?

[Neal H. Walfield]

At Fri, 08 Jul 2011 14:03:48 +0800,
Jonathan Wilson wrote:
> 
> ICD2 policy plugins never went anywhere, I dont think I can make them do 
> what I thought they could and would need to clone the entire icd2 binary 
> (and given how hard cloning any binary is, even simple ones, that's not 
> something I plan to try)

Rather than cloning ICD2, it might be easier to just reimplement the
ICD2 dbus interface [1].  This could be done either within
NetworkManager or as another program that just translates the ICD2
interface into NetworkManager calls.

  [1] http://maemo.org/api_refs/5.0/5.0-final/icd2/group__dbus__api.html

>From experience, working with ICD2 is rather annoying due to a lack of
documentation and subtle bugs (e.g.,
https://bugs.maemo.org/show_bug.cgi?id=10710).

Based on my experience abstracting NetworkManager and ICD2's
interfaces for Woodchuck [2], I expect such an implementation to be
doable, but require a few thousand lines of code.

  [2] http://maemo.org/api_refs/5.0/5.0-final/icd2/group__dbus__api.html

Neal
___
maemo-developers mailing list
maemo-developers@maemo.org
https://lists.maemo.org/mailman/listinfo/maemo-developers

Re: How to get dbus-monitor to listen to more dbus messages?

[Neal H. Walfield]

At Sun, 03 Jul 2011 22:00:13 +0800,
Jonathan Wilson wrote:
> 
> I am trying to get dbus-monitor to listen to some dbus messages, 
> (specifically the tklock_open and tklock_close messages sent from mce to 
> systemui for the device lock state) and no combination of arguments to 
> dbus-monitor seems to get dbus-monitor to output them.

I put the following in /etc/dbus-1/system-local.conf (this file is
automatically loaded via system.conf if it exists).  I've been using
it for about a year.


  








  


___
maemo-developers mailing list
maemo-developers@maemo.org
https://lists.maemo.org/mailman/listinfo/maemo-developers

Re: Neighboring Cell Towers

[Neal H. Walfield]

It seems a patch to get neighboring cell towers was recently posted to
the ofono list by a Nokia developer:

  http://lists.ofono.org/pipermail/ofono/2010-December/006829.html

I will try to extract the relevant functionality and see if I can't
get it to run on the N900.

Neal
___
maemo-developers mailing list
maemo-developers@maemo.org
https://lists.maemo.org/mailman/listinfo/maemo-developers

Re: Neighboring Cell Towers

[Neal H. Walfield]

At Wed, 23 Feb 2011 23:37:39 +,
Ian Stirling wrote:
> 
> Neal H. Walfield wrote:
> > Hi,
> > 
> > I'm writing a context-aware application.  I want to learn the current
> > location using any visible cell towers and their respective signal
> 
> > I can get the currently associated cell tower using the Phone.Net API,
> > but I have not found an interface to obtain the visible, non-connected
> > cell towers.  Is there one?  Do I have to interface with the SIM card
> > or modem directly?  If so, do you have any pointers on how I should go
> 
> As far as I'm aware, there is no way to do this.
> 
> ssh root@phone
> 
> Nokia-N900:~# pnatd
> at+cops=?
> +COPS: (2,"T-Mobile",,"23430",0),(2,"T-Mobile",,"23430",2),
> (3,"O2 - UK",,"23410",0),
> (3,"3 UK",,"23420",2),,(0,1,3),(0,2)
> at+csq
> +CSQ: 6,99
> 
> 
> Is about the closest we get.
> There isn't as I understand it a way to get the unconnected cell tower
> info.

I hadn't thought of looking for AT commands.

The canonical command [1] to get location information appears to be
+CMOLRG (Mobile Originated GPS Location Request).  Unfortunately the
N900 does not support this.

  [1] http://www.shapeshifter.se/wp-content/uploads/2008/04/27007-830.pdf

Another option, which I found, is +CENG.  Unfortunately, the N900 does
not support this one either.

  http://www.edaboard.com/thread121038.html

___
maemo-developers mailing list
maemo-developers@maemo.org
https://lists.maemo.org/mailman/listinfo/maemo-developers

Neighboring Cell Towers

[Neal H. Walfield]

Hi,

I'm writing a context-aware application.  I want to learn the current
location using any visible cell towers and their respective signal
strength.  I realize I could use something like ACWP (Assisted
Complementary Wireless Positioning) to get the location, but that
requires network access and I don't actually care about the actual
longitude and latitude; I need an environmental fingerprint.

I can get the currently associated cell tower using the Phone.Net API,
but I have not found an interface to obtain the visible, non-connected
cell towers.  Is there one?  Do I have to interface with the SIM card
or modem directly?  If so, do you have any pointers on how I should go
about doing this?  What might be useful is the data that
/usr/libexec/location-daemon sends to, e.g., supl.nokia.com.  Any
ideas on how it harvests the data?

Thanks for your help,

Neal
___
maemo-developers mailing list
maemo-developers@maemo.org
https://lists.maemo.org/mailman/listinfo/maemo-developers

Re: gprs_rx_bytes vs gprs_home_rx_bytes

[Neal H. Walfield]

Hi, Andrew,

At Wed, 30 Jun 2010 08:59:55 -0400,
Andrew Olmsted wrote:
> As far as I know, "gprs_rx_bytes" is the old pre-PR1.2 key.  This key will 
> not 
> change in PR1.2.

What do you mean by it will not change?  It will no longer be updated
(i.e., they broke the ABI?) Or, that its meaning has stayed the same
(and it is both home and roaming)?

Thanks,

Neal
___
maemo-developers mailing list
maemo-developers@maemo.org
https://lists.maemo.org/mailman/listinfo/maemo-developers

gprs_rx_bytes vs gprs_home_rx_bytes

[Neal H. Walfield]

What is the difference between gprs_home_rx_bytes vs gprs_rx_bytes?  I
thought they were for data transferred on the home network and when
roaming, respectively, however, the gconf values don't match what I
find in Settings -> Phone:

  Nokia-N900-42-11:~# gconftool-2 -R /system/osso/connectivity/network_type | 
grep bytes
gprs_home_tx_bytes = 4034987
gprs_rx_bytes = 709842177
gprs_home_rx_bytes = 31515124
gprs_tx_bytes = 62829876

My "Home network data counter" matches:

  Sent: 4.03 MB
  Receiverd: 31.5 MB

However, my "Roaming network data counter" doesn't:

  Sent: 0 B
  Received 0 B

Are gprs_home_{rx,tx}_bytes since the last reset and
gprs_{rx,tx}_bytes since forever?

Thanks,

Neal
___
maemo-developers mailing list
maemo-developers@maemo.org
https://lists.maemo.org/mailman/listinfo/maemo-developers

Re: Statistics about recently disconnected network connections

[Neal H. Walfield]

At Wed, 16 Jun 2010 19:33:32 +0300,
Jukka Rissanen wrote:
> 
> On 16 June 2010 16:58, Neal H. Walfield  wrote:
> > Hi,
> >
> > Currently, I figure out when a network connection is disconnecting by
> > listening for ICD_DBUS_API_STATE_SIG signals and checking if the
> > network connection status is ICD_STATE_DISCONNECTING.  When this
> > happens, I send an ICD_DBUS_API_STATISTICS_REQ to ICD2 specifying the
> > disconnecting network connection.  I get a reply that indicates that I
> > should see an ICD_DBUS_API_STATISTICS_SIG signal (one argument, value
> > 1).  I don't see any statistics, however.
> 
> Reason is that when DISCONNECTING signal is sent, there is very little
> time to do anything before the connection is disconnected. So I
> suspect that the connection was disconnected before the signal was
> sent.

Except that, as I said, the reply to my ICD_DBUS_API_STATISTICS_REQ
indicates that one ICD_DBUS_API_STATISTICS_SIG signal was sent.  If,
when ICD2 gets my request and it sends no signals, then it should
reply with a value of 0.


Thanks,

Neal
___
maemo-developers mailing list
maemo-developers@maemo.org
https://lists.maemo.org/mailman/listinfo/maemo-developers

Statistics about recently disconnected network connections

[Neal H. Walfield]

Hi,

Using ICD_DBUS_API_STATISTICS_REQ, I'm able to get statistics about
established network connections.  I'd like to get network statistics
(specifically bytes sent and received) about recently terminated
network connections.

Currently, I figure out when a network connection is disconnecting by
listening for ICD_DBUS_API_STATE_SIG signals and checking if the
network connection status is ICD_STATE_DISCONNECTING.  When this
happens, I send an ICD_DBUS_API_STATISTICS_REQ to ICD2 specifying the
disconnecting network connection.  I get a reply that indicates that I
should see an ICD_DBUS_API_STATISTICS_SIG signal (one argument, value
1).  I don't see any statistics, however.

  (Cf. )

To check that I'm correctly invoking ICD_DBUS_API_STATE_SIG, I changed
my code to do a stat when it sees that a connection is established
(ICD_DBUS_API_STATE_SIG signal with status equal to
ICD_STATE_CONNECTED).  ICD_DBUS_API_STATISTICS_REQ again returns 1,
but this time I indeed get statistics about the indicated connection.

What is the right way to get statistics about a closed network
connection?  Is this a bug in ICD2 (icd2 is version
0.87+fremantle9+0m5)?  If so, any suggestions for a work around?  I'd
really like to avoid aggressively polling.

Thanks,

Neal

Example code:


  DBusHandlerResult network_callback (DBusConnection *connection,
  DBusMessage *message,
  void *user_data)
  {
if (dbus_message_is_signal (message,
ICD_DBUS_API_INTERFACE,
ICD_DBUS_API_STATE_SIG))
  {
char *service_type = NULL;
uint32_t service_attributes = 0;
char *service_id = NULL;
char *network_type = NULL;
uint32_t network_attributes = 0;
char *network_id = NULL;
int network_id_len = 0;
char *conn_error = NULL;
int32_t status = 0;

DBusError error;
dbus_error_init (&error);

if (! dbus_message_get_args (message, &error,
 DBUS_TYPE_STRING, &service_type,
 DBUS_TYPE_UINT32, &service_attributes,
 DBUS_TYPE_STRING, &service_id,
 DBUS_TYPE_STRING, &network_type,
 DBUS_TYPE_UINT32, &network_attributes,
 DBUS_TYPE_ARRAY, DBUS_TYPE_BYTE,
 &network_id, &network_id_len,
 DBUS_TYPE_STRING, &conn_error,
 DBUS_TYPE_UINT32, &status,
 DBUS_TYPE_INVALID))
  {
debug (0, "Failed to grok "ICD_DBUS_API_STATE_SIG" reply: %s",
   error.message);
dbus_error_free (&error);
return DBUS_HANDLER_RESULT_NOT_YET_HANDLED;
  }

if (status == ICD_STATE_DISCONNECTING)
  {
DBusMessage *message = dbus_message_new_method_call
  (/* Service.  */ ICD_DBUS_API_INTERFACE,
   /* Object.  */ ICD_DBUS_API_PATH,
   /* Interface.  */ ICD_DBUS_API_INTERFACE,
   /* Method.  */ ICD_DBUS_API_STATISTICS_REQ);

dbus_message_append_args
  (message,
   DBUS_TYPE_STRING, &service_type,
   DBUS_TYPE_UINT32, &service_attributes,
   DBUS_TYPE_STRING, &service_id,
   DBUS_TYPE_STRING, &network_type,
   DBUS_TYPE_UINT32, &network_attributes,
   DBUS_TYPE_ARRAY, DBUS_TYPE_BYTE,
   &network_id, network_id_len,
   DBUS_TYPE_INVALID);

DBusMessage *reply
  = dbus_connection_send_with_reply_and_block
  (connection, message, 60 * 1000, &error);
if (dbus_error_is_set (&error))
  {
debug (0, "Error sending to ICd2: %s", error.message);
dbus_error_free (&error);
goto stat_done;
  }

int connected = 0;
if (! dbus_message_get_args (reply, &error, 
 DBUS_TYPE_UINT32, &connected,
 DBUS_TYPE_INVALID))
  {
debug (0, "Error parsing reply from ICd2: %s",
   error.message);
dbus_error_free (&error);
goto stat_done;
  }
else
  debug (0, "connected: %d", connected);

  stat_done:
dbus_message_unref (message);
dbus_message_unref (reply);
  }
  }
else if (dbus_message_is_signal (message,
 ICD_DBUS_API_INTERFACE,
 ICD_DBUS_API_STATISTICS_SIG))
  {
cha

hardware keyboard "open" event

[Neal H. Walfield]

Hi,

In an application I'm working on, I'd like to show the search tool bar
if the user "opens" the hardware keyboard.  Is there a way to detect
this?  I've tried searching but have had no success; perhaps I'm using
the wrong keywords.

Thanks,

Neal
___
maemo-developers mailing list
maemo-developers@maemo.org
https://lists.maemo.org/mailman/listinfo/maemo-developers

Re: row-activate signal for HildonTouchSelector

[Neal H. Walfield]

At Mon, 04 Jan 2010 16:56:17 +0100,
Neal H. Walfield wrote:
> After some investigation, it seems that setting the
> HildonTouchSelector widget to be in HILDON_UI_MODE_NORMAL causes
> "changed" to exhibit the above behavior.
> 
> I now have a new problem.  When calling
> hildon_touch_seletor_select_iter, the "changed" signal is emitted but
> hildon_touch_selector_get_last_activated_row does not reflect the
> newly selected row.  Any ideas?

In case someone is interested in the resolution of this, I've reported
this as bug #7648.

https://bugs.maemo.org/show_bug.cgi?id=7648

Neal


___
maemo-developers mailing list
maemo-developers@maemo.org
https://lists.maemo.org/mailman/listinfo/maemo-developers

Re: row-activate signal for HildonTouchSelector

[Neal H. Walfield]

At Mon, 04 Jan 2010 13:58:52 +0100,
Neal H. Walfield wrote:
> 
> I want to use the HildonTouchSelector widget to enable the user to
> select an item from a list.  I want an action to occur when the user
> "clicks" on an item.  I don't want the user to have to press another
> button.
> 
> The HildonTouchSelector widget provides a "changed" signal.  This is
> emtited when the active item changes.  This appears to be due to the
> user selecting an item (even the preselected one) or an item (perhaps
> the active item?) in the model changing via, e.g., a "row-changed"
> signal.
> 
> I need to distinguish between these two cases.  I only want to act if
> the user actually "clicked" on an item.  That is, I want the behavior
> that the media player implements, e.g., in its "select artist" view.
> Will someone please suggest how to best achieve this?

After some investigation, it seems that setting the
HildonTouchSelector widget to be in HILDON_UI_MODE_NORMAL causes
"changed" to exhibit the above behavior.

I now have a new problem.  When calling
hildon_touch_seletor_select_iter, the "changed" signal is emitted but
hildon_touch_selector_get_last_activated_row does not reflect the
newly selected row.  Any ideas?

Thanks,

Neal

___
maemo-developers mailing list
maemo-developers@maemo.org
https://lists.maemo.org/mailman/listinfo/maemo-developers

row-activate signal for HildonTouchSelector

[Neal H. Walfield]

I want to use the HildonTouchSelector widget to enable the user to
select an item from a list.  I want an action to occur when the user
"clicks" on an item.  I don't want the user to have to press another
button.

The HildonTouchSelector widget provides a "changed" signal.  This is
emtited when the active item changes.  This appears to be due to the
user selecting an item (even the preselected one) or an item (perhaps
the active item?) in the model changing via, e.g., a "row-changed"
signal.

I need to distinguish between these two cases.  I only want to act if
the user actually "clicked" on an item.  That is, I want the behavior
that the media player implements, e.g., in its "select artist" view.
Will someone please suggest how to best achieve this?

Thanks,

Neal
___
maemo-developers mailing list
maemo-developers@maemo.org
https://lists.maemo.org/mailman/listinfo/maemo-developers

Re: Kernel level checkpoint & restart

[Neal H. Walfield]

At Thu, 14 Aug 2008 15:24:03 +0100,
Andrew Flegg wrote:
> Today's LWN has an interesting article on a patchset being developed
> to give checkpoint & restart functionality to Linux:
> 
> http://lwn.net/SubscriberLink/293575/03bb471d0d0b9261/
> 
> This'd effectively allow the kernel to freeze a process and restore it
> at a later date. The immediate thought in my mind was using this to
> help automatically deal with low-memory, low-battery and standby
> problems in Maemo.
> 
> It's still very immature, but something Maemo kernel hackers may be
> interested in.

You may want to check out DMTCP, which I have heard is fairly stable:

  http://dmtcp.sourceforge.net/

and its associated paper:

  "Transparent User-Level Checkpointing for the Native POSIX Thread
  Library for Linux"
  by Michael Rieker, Jason Ansel and Gene Cooperman,
  PDPTA 2006

  http://www.ccs.neu.edu/home/gene/papers/pdpta06.pdf


Neal

___
maemo-developers mailing list
maemo-developers@maemo.org
https://lists.maemo.org/mailman/listinfo/maemo-developers

Re: swap not a substitute for virtual memory

[Neal H. Walfield]

At Thu, 17 Apr 2008 11:02:53 -0700 (PDT),
Allen Brown wrote:
> 
> The subject line is incorrect.  Swap is the portion of virtual
> memory which is not physical.  What you should have said is
> "swap not a substitute for *physical* memory".

Yes.
___
maemo-developers mailing list
maemo-developers@maemo.org
https://lists.maemo.org/mailman/listinfo/maemo-developers

Re: [RFC] Maemo package guidelines: mandatory categories

[Neal H. Walfield]

At Thu, 17 Apr 2008 13:52:07 +0100,
Graham Cobb wrote:
> The second problem is the real problem: categories are random, overlapp or 
> are 
> just variant words for the same thing and are not translated.  As someone 
> suggested when this was last discussed, some months ago, I believe there 
> should be a Wiki page which lists all the package names the community finds 
> acceptable.  That list should be editable by anyone who has upload rights and 
> who thinks they need a new category.  If the addition of the new category is 
> disputed, it would be discussed here and the community would come to a 
> consensus.

This is an interesting idea and has made me think of the following:
there could be a file that list the name of a category and synonyms
for that category.  The AM could download this file regularly (i.e.,
at update time) and sort applications according to the list of
category names.  This fixes all the problems that you noted except the
translation problem.  The translation problem could be fixed in that
for each language, the category name is different but the synonyms are
the same.  But, I don't know anything about localization.

Neal
___
maemo-developers mailing list
maemo-developers@maemo.org
https://lists.maemo.org/mailman/listinfo/maemo-developers

Re: playbin problem

[Neal H. Walfield]

At Thu, 17 Apr 2008 15:17:33 +0530,
Sanjay Gupta wrote:
> Can someone please tell me if playbin is working properly in maemo SDK
> environent?

It only works in Chinook.  Anything earlier including Bora doesn't
work.  You can try using playbinmaemo instead, which is in maemo
extras.
___
maemo-developers mailing list
maemo-developers@maemo.org
https://lists.maemo.org/mailman/listinfo/maemo-developers

AM limitations

[Neal H. Walfield]

At Thu, 17 Apr 2008 13:32:52 +0300,
Marius Vollmer wrote:
> 
> "ext Niels Breet" <[EMAIL PROTECTED]> writes:
> 
> > There is no official list of categories, which has brought us to state
> > we are in now.
> 
> There is, http://hildon-app-mgr.garage.maemo.org/packaging-stable.html:
> 
> Segments and Sections
> 
> By default, the AM only shows packages in certain segments to the
> user. This has been done to hide the existence of the hundreds of
> system packages that make up the operating system itself. The AM is,
> at this point, not intended to let the user manage the whole system,
> only a smaller set of third party applications.
> 
> The AM only shows packages in the user segment. Thus, your Section
> field in the control file should be of the form

It's worse than this (at least on Bora).  Downloading a .deb and
opening it in the AM results in an error if the .deb doesn't have
Section user/FOO.
___
maemo-developers mailing list
maemo-developers@maemo.org
https://lists.maemo.org/mailman/listinfo/maemo-developers

swap not a substitute for virtual memory

[Neal H. Walfield]

At Wed, 16 Apr 2008 13:18:29 -0400,
Ian Lawrence wrote:
> 
> Hi,
> >  The issue is more that the whole "Web 2.0" hasn't been planned to be run
> >  in 64MB of RAM.  There are sites out there where a single Flash object
> >  or JavaScript script consumes more memory.  Even 128MB is tight.
> But, I mean it is possible to increase the amount of memory available
> for a 770 using an MMC card, right?. I am not sure what the upper
> limit is but 128MB + 64MB of system memory seems feasible.

You are confused about what swap is.  Swap allows the memory manager
to page (save) anonymous memory to backing store (which in this case
is your disk).

If union of the working set of the active applications (the memory
they need to run with only a trivial amount of faults, i.e., page-ins)
exceeds the amount of *RAM*, then your system thrashes.  As disk is
several orders of magnitude slower than RAM, in such a case, your
system makes no useful progress and will appear to you to have frozen.

The advantage of swap then, is that if you are running e.g. Mozilla
and its working set is large, then the memory manager can page the
anonymous memory of background applications (as well as the anonymous
memory that Mozilla uses, which is hopefully inactive).

Without swap, the only data that the memory manager can page is that
associated with files.  So, program and library text, data files, etc.
The problem is that most programs don't use the disk format for the
data they use: on bringing it into memory, they convert it to a form
useful for in-memory operations.  Thus, a lot of the memory on your
system will be occupied by such computed data, which, without swap, is
locked in memory until it is explicitly freed.


For what it is worth, after I removed /usr/bin/metalayer-crawler, my
system went from being completely unusable for anything but being a
broken alarm clock to being relatively useful.
___
maemo-developers mailing list
maemo-developers@maemo.org
https://lists.maemo.org/mailman/listinfo/maemo-developers

Re: Continuous reboot problem with the N770 hacker edition

[Neal H. Walfield]

At Fri, 20 Apr 2007 18:26:42 +0200,
Rainer Dorsch <[EMAIL PROTECTED]> wrote:
> I used the N770 with the hacker edition of OS2007 heavily last week and it 
> broke:
> The device is rebooting constantly, even putting it into RD mode did not help 
> getting rid of this. The sequence is
> 1. blue boot progress bar
> 2. Nokia hands and sound
> 3. almost immidiately after the sound, a white stripe appears on the left of 
> the display. Looks like this is the Maemo menue bar. Then the screen becomes 
> white
> 4. after approx. 19 seconds the device reboots and continues with 1.
> 
> I interrupt the reboot cycles by removing the battery.

I'm using Gergale and this happens to me about once every week or two.
I had enough time after the device boots to run xterm and top.  The
problem was jffs2_gcd_mtd4 was taking 100% of the CPU and a fair
amount of memory.  This led me to suspect that the OOM killer was
killing an essential task and thereby causing the watchdog timer to
reset the device.  To prevent this, I simply disabled the watchdog
time.  Now when this happens, jffs2_gcd_mtd4 burns CPU (and thus is
unusable) but it is a few hours between resets.  After about day or
two, the device is happy again.

Neal

___
maemo-developers mailing list
maemo-developers@maemo.org
https://maemo.org/mailman/listinfo/maemo-developers

Re: Plash:tools for practical least privilege

[Neal H. Walfield]

At Wed, 14 Mar 2007 07:35:24 +0100,
"Kees Jongenburger" <[EMAIL PROTECTED]> wrote:
> 
> Hello
> 
> We have talked about security before when talking about repositories
> and such. J2me and Applets use a sandbox model that I happen to like.
> I came across this product http://plash.beasts.org/ . I wonder if
> people have experimented
> with this or have ideas about sandboxes. I am searching for AppArmour
> kind of applications that would be lightweight enough to run on
> embedded devices.

Check out OLPC's security framework, Bitfrost [1].  It is also based
on POLA principles.

[1] http://dev.laptop.org/git.do?p=security;a=blob;hb=HEAD;f=bitfrost.txt
___
maemo-developers mailing list
maemo-developers@maemo.org
https://maemo.org/mailman/listinfo/maemo-developers

Application not appearing in task manager

[Neal H. Walfield]

GPE Starling does not show up in the task manager.  It starts fine but
I cannot figure out why it doesn't appear in the task bar.  The code
(and package) is here: http://linuxtogo.org/~neal/maemo/ .  (The
dependencies can be found at
http://downloads.kernelconcepts.de/maemo2/ .)

Any insights would be appreciated.


Here is the output of dbus-monitor when starting Starling from the
application menu:

  signal sender=org.freedesktop.DBus -> dest=(null destination)
  interface=org.freedesktop.DBus; member=NameOwnerChanged
   string ":1.62"
  string ""
  string ":1.62"
  method call sender=:1.62 -> dest=org.freedesktop.DBus
  interface=org.freedesktop.DBus; member=Hello
  signal sender=org.freedesktop.DBus -> dest=(null destination)
  interface=org.freedesktop.DBus; member=NameOwnerChanged
   string "com.nokia.starling"
  string ""
  string ":1.62"
  method call sender=:1.62 -> dest=org.freedesktop.DBus
  interface=org.freedesktop.DBus; member=RequestName
   string "com.nokia.starling"
  uint32 1

It looks okay but I'm not a d-bus expert.

The perhaps relevant start up code:

  #define APPLICATION_DBUS_SERVICE "starling"
  
  ...
  
  #ifdef IS_HILDON
osso_context_t *osso_context;
  
/* Initialize maemo application */
osso_context = osso_initialize(APPLICATION_DBUS_SERVICE, "1.0",
TRUE, NULL);
  
/* Check that initialization was ok */
if (osso_context == NULL)
  {
g_critical ("Failed to initialize OSSO context!");
return OSSO_ERROR;
  }
  #endif
  
  ...

  char *iconname = DATADIR "/pixmaps/bass.png";
  GdkPixbuf *icon = gdk_pixbuf_new_from_file (iconname, &err);
  if (err)
{
  fprintf (stderr, "Loading %s: %s\n", iconname, err->message);
  g_error_free (err);
}
  else
gtk_window_set_default_icon (icon);



/usr/share/applications/hildon/starling.desktop:

[Desktop Entry]
Encoding=UTF-8
Version=1.0
Type=Application
Name=Audio player
Comment=Starling
Exec=/usr/bin/starling
Terminal=false
Icon=bass
X-HildonDesk-ShowInToolbar=true
X-Osso-Type=application/x-executable
X-Osso-Service=starling
StartupWMClass=starling


/usr/share/dbus-1/services/com.nokia.starling.service:

[D-BUS Service]
Name=com.nokia.starling
Exec=/usr/bin/starling


___
maemo-developers mailing list
maemo-developers@maemo.org
https://maemo.org/mailman/listinfo/maemo-developers

Re: [maemo-developers] Package architecture armel vs gnueabi-linux-arm

[Neal H. Walfield]

At Sat, 24 Feb 2007 12:59:34 +0100,
"Neal H. Walfield" <[EMAIL PROTECTED]> wrote:
> 
> At Mon, 11 Sep 2006 13:00:32 +0300,
> Marius Vollmer <[EMAIL PROTECTED]> wrote:
> > 
> > "ext Niels Breet" <[EMAIL PROTECTED]> writes:
> > 
> > > After doing some upgrades to Sardine, dpkg fails on a lot of packages
> > > because of:
> > > package architecture (armel) does not match system (gnueabi-linux-arm)
> > >
> > > Does this mean that you have introduced a new architecture and as a 
> > > result of
> > > that all current packages are incompatible? (Again)
> > 
> > Nonono, sorry.  Dpkg in Sardine was somehow miscompiled for Sardine,
> > but I haven't had time to fix it.  Will do it immediately...
> 
> I have installed the following rootstrap:
> 
>   
> http://repository.maemo.org/stable/gregale/armel/Maemo_Dev_Platform_v2.2_armel-rootstrap.tgz
> 
> and am experiencing this same problem.  Did this somehow creep back
> in?  Is there an easy fix?

The problem is that the debian-sarge devkit provides dpkg et al.,
which take precedence over the ones provided by the rootstrap.  The
solution was to install just the debian devkit not both the debain and
debian-sarge devkits.
___
maemo-developers mailing list
maemo-developers@maemo.org
https://maemo.org/mailman/listinfo/maemo-developers

Re: Programmatically Enabling Hardware Keys

[Neal H. Walfield]

At Sun, 25 Feb 2007 07:52:25 +0100,
Zoran Kolic <[EMAIL PROTECTED]> wrote:
> In x server it is possible to use "xev" to find correct button identifi-
> cation. Then, using xmodmap, to map them to appropriate function. I'm not
> sure does it work on 770; have to take a look on binaries.

That doesn't help me when the lid is put on: these key presses are
suppressed.  My question is how to stop this.
___
maemo-developers mailing list
maemo-developers@maemo.org
https://maemo.org/mailman/listinfo/maemo-developers

Programmatically Enabling Hardware Keys

[Neal H. Walfield]

I'm using GPE's Starling to be able to use my 770 as an mp3 player.  I
would like it if Starling, when it is the foreground application,
could provent Maemo from disabling/locking the hardware keys when the
lid is closed.  I've done a fair amount of searching but I haven't
found anything.  Hints?

Thanks,
Neal

___
maemo-developers mailing list
maemo-developers@maemo.org
https://maemo.org/mailman/listinfo/maemo-developers

Re: [maemo-developers] Package architecture armel vs gnueabi-linux-arm

[Neal H. Walfield]

At Mon, 11 Sep 2006 13:00:32 +0300,
Marius Vollmer <[EMAIL PROTECTED]> wrote:
> 
> "ext Niels Breet" <[EMAIL PROTECTED]> writes:
> 
> > After doing some upgrades to Sardine, dpkg fails on a lot of packages
> > because of:
> > package architecture (armel) does not match system (gnueabi-linux-arm)
> >
> > Does this mean that you have introduced a new architecture and as a result 
> > of
> > that all current packages are incompatible? (Again)
> 
> Nonono, sorry.  Dpkg in Sardine was somehow miscompiled for Sardine,
> but I haven't had time to fix it.  Will do it immediately...

I have installed the following rootstrap:

  
http://repository.maemo.org/stable/gregale/armel/Maemo_Dev_Platform_v2.2_armel-rootstrap.tgz

and am experiencing this same problem.  Did this somehow creep back
in?  Is there an easy fix?

Thanks,
Neal
___
maemo-developers mailing list
maemo-developers@maemo.org
https://maemo.org/mailman/listinfo/maemo-developers

gstreamer's playbin

[Neal H. Walfield]

The following works for me:

 gst-launch-0.10 -v file:///home/user/MyDocs/.sounds/Take_Me_Down.mp3 \
   ! dspmp3sink

But this:

 gst-launch-0.10 -v playbin \
   uri=file:///home/user/MyDocs/.sounds/Take_Me_Down.mp3

just *hangs* at "Pipeline is PREROLLING ...".  Someone mentioned on
irc that playbin is simply broken on maemo.  Is that the case?  Do I
have to hand code my pipelines and implement my own file type
detection?

For what it is worth, I'm using OS2006-3.

Thanks,
Neal
___
maemo-developers mailing list
maemo-developers@maemo.org
https://maemo.org/mailman/listinfo/maemo-developers

Re: [maemo-developers] library updates

[Neal H. Walfield]

At Thu, 20 Jul 2006 18:06:31 +0100,
Tomas Frydrych wrote:
> > how are updates of library packages which are not visible in the
> > application manager expected to work? We just noticed that when updating
> > gpe-calendar the updated libeventdb does not get installed. libeventdb
> > is visible because its section currently is "user/libs", so in this case
> > i can install the update but what if the package is not visible?
> 
> I think you just need to make sure that the main package includes
> appropriate version number in the dependency declaration, not just the
> library name; at least that seems to work for me.

That's a hack and a poor one at that.  Versioned dependencies are used
when the applications depends on a particular ABI version of a
library.  I think it is a bad idea to abuse this.

What florian wants is `apt-get upgrade' functionality, i.e. install
updates for any installed package.  The application manager could use
the following predicate to determine if an application is up to date:
if and application and all of its dependencies (direct and indirect)
are the latest version available (respecting APT::Default-Release and
the pin settings, of course), it is up to date.  This would cause a
bit of confusion when a new version of a commonly used library, e.g.,
GTK+, becomes available as then pretty much all applications
indirectly depend on it.  These can be exclude by having a base system
container.
___
maemo-developers mailing list
maemo-developers@maemo.org
https://maemo.org/mailman/listinfo/maemo-developers

Re: [maemo-developers] maemo 2.0 home applet

[Neal H. Walfield]

At Sat, 17 Jun 2006 19:43:01 -0700,
Collin R. Mulliner wrote:
> I'm currently trying to port my IpHome applet to Maemo 2.0, and I can't
> get it to display. I added the .desktop file
> to /user/share/applications/hildon-home/ and my applet shows up in the
> applet selction list. When I select it, I see the layout editor but my
> applet is not displayed. I looked into the home applet samples but
> couldn't find any difference to the old interface. What am I
> missing?

I didn't use the Maemo 1.1 interface so I can't comment if it has
changed but you can find the details of the 2.0 interface here [1].

> Btw. how is the size of an applet determined since the height and width
> are no longer part of the config file??

With the hildon_home_applet_lib_get_requested_width callback.


[1] 
http://maemo.org/lxr/source/maemo-af-desktop/hildon-home/hildon-home-plugin-loader.c

___
maemo-developers mailing list
maemo-developers@maemo.org
https://maemo.org/mailman/listinfo/maemo-developers