Firefox 4: list of supported platforms

[Mike Beltzner]

Hello,

I've been working on the list of supported platforms for Firefox 4 and wanted to provide an update on some decisions that have been reached.

* In February of this year, there was a discussion [1] in which we decided that the minimum supported version of Mac OS X would be 10.5 (Leopard)

* Early in the summer we decided to not yet provide supported 64-bit builds on Windows [2], and will instead work on delivering those in some future release.

* In July we decided that the minimum supported version of Windows would be Windows 2000 [3]

* After some evaluation [4] we decided this week to drop support for OSX/PPC; Firefox 4 Mac OS X binaries will ship as universal builds for i386/x86-64 only

* We are considering dropping support for i386 architectures which do not support SSE2 [5] (older Athlon CPUs, some VIA chipsets) but have not yet reached a final decision here.

I will maintain a de-facto list on this page: https://wiki.mozilla.org/Firefox/4/Platforms

cheers,
mike

[1]: http://groups.google.com/group/mozilla.dev.planning/browse_thread/thread/7d3a647586bab993
[2]: I can't find the link, so if you'd like to, treat this post or https://wiki.mozilla.org/Firefox/4/Platforms as canonical!
[3]: https://bugzilla.mozilla.org/show_bug.cgi?id=579078
[4]: https://wiki.mozilla.org/Platform/2010-08-31#Roundtable
[5]: https://bugzilla.mozilla.org/show_bug.cgi?id=593117
[6]: http://www.mozilla.com/firefox/system-requirements.html

[James May]

On 3 September 2010 07:51, Mike Beltzner <belt...@mozilla.com> wrote:

> [..]

> * We are considering dropping support for i386 architectures which do not
> support SSE2 [5] (older Athlon CPUs, some VIA chipsets) but have not yet
> reached a final decision here.
>

There was a thread on this back in December[1]. I think the outcome was that
it was a clear "collect more data".

On an anecdotal note, I know of several machines still in active use (some
even that have just even have just been updated do windows 7) that would be
disadvantaged by this change. Diverging from OS requirements (as I stated
before, Windows 7 supports non-sse2 systems) could be confusing.

HTH,

James

[1]
http://groups.google.com/group/mozilla.dev.platform/browse_thread/thread/9f5b54b0b4f388ce

[Jean-Marc Desperrier]

Ted Mielczarek wrote:
>> I think the outcome was that
>> > it was a clear "collect more data".

> Indeed, which is exactly what is happening:
> https://bugzilla.mozilla.org/show_bug.cgi?id=593117#c13

That's amazing how many people still run Firefox on a PIII. Also a lot
of Athlon.

I wonder if the only day they'll stop using their computer is when
Firefox won't run anymore on it ;-)

[woiferl]

So, does that mean 386 or SSE2? SSE2 goes much further than 386 - it
includes all Pentiums until P4 as well.

[bblackmoor]

On Sep 2, 5:51 pm, Mike Beltzner <beltz...@mozilla.com> wrote:
>
> * Early in the summer we decided to not yet provide supported 64-bit builds on Windows [2], and will instead work on delivering those in some future release.

Well, people have only been asking for that since 2008. So, no
hurry...

(Are you KIDDING? I hope this is just a bad joke.)

[Boris Zbarsky]

On 9/13/10 1:03 AM, Thomas Stache wrote:
> Don't expect your websites to render any faster, just because Firefox can address more memory!

Performance of 32-bit vs 64-bit code really depends on the workload; for
a lot of things we do the extra registers make a noticeable difference
and we get a speedup (even in spite of the higher cache pressure 64-bit
builds entail).

Last I measured, for DOM operations we're talking about 20%, at least on
Mac.

-Boris

[Mike Beltzner]

> Performance of 32-bit vs 64-bit code really depends on the workload; for

This isn't on-topic for this thread. Please stop.

cheers,
mike