Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Status of Firefox for Maemo

50 views
Skip to first unread message

Mark Finkle

unread,
Jul 13, 2011, 5:38:35 PM7/13/11
to dev-platfo...@lists.mozilla.org, release, mobile-dev
To make responsible use of our limited resources, it's time to discuss the future status of the Firefox for Maemo 5 (GTK and Qt) releases and infrastructure. Mozilla is currently treating the Maemo builds with a medium to high priority:
* Builds, unit tests and performance tests are executed for all changes landed in mozilla-central (and other project branches)
* Nightlies are created and tested
* Aurora, Beta and Final releases are created and tested
* Final releases are in the Ovi Store
* Builds are not considered "tier 1", which means any breakage must be fixed immediately, but breakage is still fixed as soon as feasible.

This level of effort is not easy to maintain, so we need to justify the costs. Given Maemo has been end-of-lifed, there is little desire to continue at the current level of effort. I'd like to create some proposals for Mozilla's roadmap for Firefox for Maemo. I'll go first.

Baseline Proposal:

Let's move Firefox for Maemo to a Tier 3 build configuration. Tier 3 is defined as:

"Tier-3 platforms have a maintainer or community which attempt to keep the platform working. These platforms may or may not work at any time, and often have little test coverage"

from https://developer.mozilla.org/en/Supported_build_configurations

Firefox for Maemo has a somewhat active community. Bugs get filed and patches get created. This is a good thing. This is the baseline proposal. Essentially, this means:
* The code is not removed from the Mozilla source tree
* Developers could pull the source and make builds
* Bugs and patching are the core responsibility of the community, not Mozilla
* Any L10N outside of the core source code would also fall to the community to manage
* Mozilla would stop doing any builds and testing (automated or otherwise) - this falls to the community to manage

-----

There are other levels of effort we can discuss as well. Mozilla has build infrastructure in place and might be able to continue creating builds will little effort. Maybe not. Mozilla has a large pool of N900 devices used for automated testing. Maybe these could be managed by a community group and not Mozilla. Maybe not.

The main point you should take away from this posting is that the staus quo can not continue. Feedback, corrections and additional information is welcome.

Mark Finkle

Mark Finkle

unread,
Jul 13, 2011, 5:49:00 PM7/13/11
to dev-platfo...@lists.mozilla.org, release, mobile-dev
Whoops. Corrected:

* Builds are not considered "tier 1", which means any breakage is not fixed immediately, but breakage is still fixed as soon as feasible.

Doug Turner

unread,
Jul 13, 2011, 5:51:57 PM7/13/11
to Mark Finkle, dev-platfo...@lists.mozilla.org, release, mobile-dev
I support this decision. Maemo hasn't be a priority in practice and
it is good to see we are taking about it.

> _______________________________________________
> dev-platforms-mobile mailing list
> dev-platfo...@lists.mozilla.org
> https://lists.mozilla.org/listinfo/dev-platforms-mobile
>

Robert Kaiser

unread,
Jul 13, 2011, 8:11:27 PM7/13/11
to
Mark Finkle schrieb:

> * Mozilla would stop doing any builds and testing (automated or otherwise) - this falls to the community to manage

This will make the community shrink considerably and probably mostly die
out - but I guess that's the actual intention. After all, us
freedom-loving Maemo/MeeGo freaks are not an interesting market due to
the sheer number of followers of the various more closed up mobile
religions out there (and I'm trying hard to hold back any [de]valuing
attributes).

Robert Kaiser


--
Note that any statements of mine - no matter how passionate - are never
meant to be offensive but very often as food for thought or possible
arguments that we as a community should think about. And most of the
time, I even appreciate irony and fun! :)

Fabrice Desré

unread,
Jul 14, 2011, 4:22:52 AM7/14/11
to
On Thu, 14 Jul 2011 02:11:27 +0200, Robert Kaiser wrote:
> This will make the community shrink considerably and probably mostly die
> out - but I guess that's the actual intention. After all, us
> freedom-loving Maemo/MeeGo freaks are not an interesting market due to
> the sheer number of followers of the various more closed up mobile
> religions out there (and I'm trying hard to hold back any [de]valuing
> attributes).

Robert,

Don't you think the lack of credible hardware running these platforms is
the main reason for MoCo lowering its official support?

Freedom-loving freaks should work on getting meego running on top of
cyanogenmod kernels to reach current android users IMHO (personally I
would love this).

Fabrice

Robert Kaiser

unread,
Jul 14, 2011, 7:08:57 AM7/14/11
to
Fabrice Desré schrieb:

> Freedom-loving freaks should work on getting meego running on top of
> cyanogenmod kernels to reach current android users IMHO (personally I
> would love this).

The crappy fork called the "android kernel" is nothing us freaks will
want to run. There's no reasons (once the GPL requirement to make free
drivers available is being followed) not to run a real Linux kernel and
a full MeeGo on the same devices - just that there's no good handset UX
for MeeGo yet. The community will probably fill that gap one day - or
some vendor may help, as there are some vendors doing R&D with MeeGo
nowadays (no officially release hardware yet, though, but Mozilla hasn't
stated any support or interest for MeeGo anyhow, and they are going with
WebKit-based browsers instead, which probably are a better solution for
them).

Armen Zambrano Gasparnian

unread,
Jul 14, 2011, 9:44:08 AM7/14/11
to Doug Turner, Mark Finkle, dev-platfo...@lists.mozilla.org, release, mobile-dev
Hi,
In order to end such a support on good terms with the community we would
want to find a way to help them support themselves.
I was thinking that creating a document with all the information to
create their releases would make them move forward.
For instance, for each Firefox release we have a "contrib" dir [1] that
they could upload to.

What do you think?

cheers,
Armen

[1]
http://ftp.mozilla.org/pub/mozilla.org/firefox/nightly/6.0b2-candidates/build1/contrib/

On 11-07-13 5:51 PM, Doug Turner wrote:
> I support this decision. Maemo hasn't be a priority in practice and
> it is good to see we are taking about it.
>
> On Wed, Jul 13, 2011 at 2:38 PM, Mark Finkle<mfi...@mozilla.com> wrote:
>> To make responsible use of our limited resources, it's time to discuss the future status of the Firefox for Maemo 5 (GTK and Qt) releases and infrastructure. Mozilla is currently treating the Maemo builds with a medium to high priority:
>> * Builds, unit tests and performance tests are executed for all changes landed in mozilla-central (and other project branches)
>> * Nightlies are created and tested
>> * Aurora, Beta and Final releases are created and tested
>> * Final releases are in the Ovi Store
>> * Builds are not considered "tier 1", which means any breakage must be fixed immediately, but breakage is still fixed as soon as feasible.
>>
>> This level of effort is not easy to maintain, so we need to justify the costs. Given Maemo has been end-of-lifed, there is little desire to continue at the current level of effort. I'd like to create some proposals for Mozilla's roadmap for Firefox for Maemo. I'll go first.
>>
>> Baseline Proposal:
>>
>> Let's move Firefox for Maemo to a Tier 3 build configuration. Tier 3 is defined as:
>>
>> "Tier-3 platforms have a maintainer or community which attempt to keep the platform working. These platforms may or may not work at any time, and often have little test coverage"
>>
>> from https://developer.mozilla.org/en/Supported_build_configurations
>>
>> Firefox for Maemo has a somewhat active community. Bugs get filed and patches get created. This is a good thing. This is the baseline proposal. Essentially, this means:
>> * The code is not removed from the Mozilla source tree
>> * Developers could pull the source and make builds
>> * Bugs and patching are the core responsibility of the community, not Mozilla

>> * Any L10N outside of the core source code would also fall to the community to manage


>> * Mozilla would stop doing any builds and testing (automated or otherwise) - this falls to the community to manage
>>

>> -----
>>
>> There are other levels of effort we can discuss as well. Mozilla has build infrastructure in place and might be able to continue creating builds will little effort. Maybe not. Mozilla has a large pool of N900 devices used for automated testing. Maybe these could be managed by a community group and not Mozilla. Maybe not.
>>
>> The main point you should take away from this posting is that the staus quo can not continue. Feedback, corrections and additional information is welcome.
>>
>> Mark Finkle
>>
>> _______________________________________________
>> dev-platforms-mobile mailing list
>> dev-platfo...@lists.mozilla.org
>> https://lists.mozilla.org/listinfo/dev-platforms-mobile
>>


--
Armen Zambrano Gasparnian (armenzg)
Mozilla Corp Release Engineer
~ Jesus Christ is my Lord

Doug Turner

unread,
Jul 14, 2011, 9:59:50 AM7/14/11
to Armen Zambrano Gasparnian, Mark Finkle, Doug Turner, dev-platfo...@lists.mozilla.org, release, mobile-dev
I agree with your general idea. When we have a owner stand up and say "I want to support Maemo contrib builds", I'd be happy to help them get setup. However, until someone steps forward and wants to own this, I am not sure we need to do anything else. The documentation to setup, build, debug, run tests, it all over the wiki.

Doug

Oleg Romashin

unread,
Jul 18, 2011, 3:02:32 PM7/18/11
to
Maemo5 is dead, and I think it does not make any sense to waste too
much resources on supporting that.
I think it make more sense to setup Qt environment for Mobile FF,
based on Meego SDK, and Harmattan SDK, these platforms binary
incompatible but code base is the same. so ideally by developing Qt
Mobile FF port we will make it works for 2 platforms (Nokia/Harmattan
devices coming soon to the market), and Meego which is going to have
bunch of different devices this and next year.

Gervase Markham

unread,
Oct 15, 2011, 6:46:19 AM10/15/11
to
On 14/07/11 06:11, Robert Kaiser wrote:
> This will make the community shrink considerably and probably mostly die
> out - but I guess that's the actual intention. After all, us
> freedom-loving Maemo/MeeGo freaks are not an interesting market due to
> the sheer number of followers of the various more closed up mobile
> religions out there (and I'm trying hard to hold back any [de]valuing
> attributes).

I'm not sure this level of sarcasm is particularly helpful.

If Mozilla was only interested in supporting proprietary platforms, why
would we be starting the B2G project?

Gerv


Robert Kaiser

unread,
Oct 15, 2011, 9:02:49 AM10/15/11
to
Gervase Markham schrieb:
I believe that comment was before B2G was announced. And I currently
think B2G is the only hope we still have to make mobile a nice place.

Matej Cepl

unread,
Oct 15, 2011, 10:01:36 AM10/15/11
to
Dne 14.7.2011 02:11, Robert Kaiser napsal(a):
> This will make the community shrink considerably and probably mostly die
> out - but I guess that's the actual intention.

As a former N900 user and lover I have this to say:

- yes, only now when I am on Android I can see how good and well
behaving Maemo is (for more detailed treatment of my feelings tovards
Android see
http://matej.ceplovi.cz/blog/2011/09/my-first-hand-on-comparison-between-nexus-s-and-n900/),
- with one exception: one of the reasons I switched was that Firefox was
just not usable as day to day browser, 256MB RAM is just not enough
apparently; on my Nexus S it is just a bit slower than the system
browser, so it is good enough for being a default browser; on N900 I was
still struggling with a temptation to use microB (which is actually
really good browser, BTW).

So, I believe that killing Fennec on N900 is a right thing to do. It was
good when there was not (semi-)open mobile system available, but unless
N9 takes world by the storm (which I doubt), I think it makes sense to
concentrate on Android as the best OS we have.

Best,

Matěj

Shmerl

unread,
Oct 17, 2011, 12:39:20 PM10/17/11
to
On Oct 15, 10:01 am, Matej Cepl <mc...@redhat.com> wrote:
> So, I believe that killing Fennec on N900 is a right thing to do. It was
> good when there was not (semi-)open mobile system available, but unless
> N9 takes world by the storm (which I doubt), I think it makes sense to
> concentrate on Android as the best OS we have.
>
> Best,
>
> Matěj

I don't consider Android the best OS from many points, and especially
from free software perspective. While Maemo is definitely fading now,
Meego effort is not, and now it's represented by the Mer project,
which is collaborating with the KDE Plasma Active team. As well as
Tizen which is backed up by Intel and Samsung. Having Qt variant of
Firefox supported is therefore in Mozilla's interest.

Regards,

Hillel.

jed

unread,
Oct 24, 2011, 12:45:37 PM10/24/11
to
On Oct 17, 12:39 pm, Shmerl <shtetl...@gmail.com> wrote:
>
> I don't consider Android the best OS from many points, and especially
> from free software perspective. While Maemo is definitely fading now,
> Meego effort is not, and now it's represented by the Mer project,
> which is collaborating with the KDE Plasma Active team. As well as
> Tizen which is backed up by Intel and Samsung. Having Qt variant of
> Firefox supported is therefore in Mozilla's interest.
>
> Regards,
>
> Hillel.

Entirely agreed on this....
MeeGo (as it was structured before Nokia bailed-out*) was the most
open/free project of it's kind.
The MER project is currently seeking permission from LF to secure the
MeeGo name.
MER will collaborate with UX projects such as KDE Plasma & Cordia,
among others.
MER's focus is at the lower level, & will spread-out over a wider hw
base than MeeGo was in-reality.
Tizen will continue the work done by MeeGo-Core, they've also started
a UX framework based on HTML5*
They've all agreed to use Qt as their primary (or one of their
primary) native UX libraries/toolkits.
Therefore I think Fennec with a Qt-derived UX, is more important than
ever...

*Before they'd even really begun a transition from Maemo (except for
branding).
**Yet to be revealed...
0 new messages