Firefox2 desupport now only 3 months away

0 views
Skip to first unread message

John O'Duinn

unread,
Sep 22, 2008, 4:04:51 PM9/22/08
to dev-pl...@lists.mozilla.org, dev-apps...@lists.mozilla.org
An important milestone happened last week.

* 17-jun-2008: Firefox3 shipped, Firefox2 started its 6 month
end-of-life support cycle.
* 17-sep-2008: Firefox2 is half-way through its end-of-life cycle
* 17-dec-2008: Firefox2 will be formally de-supported.

It’ll be sad to see the ending of an era, and there’s still lots of
details to sort out, including how to deal with Thunderbird2 support.

However, when this happens, it will be a great relief for RelEng. We can
finally de-support and mothball those 28 machines and just as
importantly, we can simplify our automation / unittest / talos code by
cleaning out the special-case conditionals needed to handle Firefox2.

Only 3 more months to go…

take care
John.

Robert Kaiser

unread,
Sep 23, 2008, 8:27:20 AM9/23/08
to
John O'Duinn wrote:
> An important milestone happened last week.
>
> * 17-jun-2008: Firefox3 shipped, Firefox2 started its 6 month
> end-of-life support cycle.
> * 17-sep-2008: Firefox2 is half-way through its end-of-life cycle
> * 17-dec-2008: Firefox2 will be formally de-supported.
>
> It’ll be sad to see the ending of an era, and there’s still lots of
> details to sort out, including how to deal with Thunderbird2 support.

SeaMonkey 1.1.x runs into the same problem as Thunderbird, needing
security fixes probably for longer than December. I understand that
RelEng will be happy to get rid of the regular FF2 infrastructure but I
hope we'll have some way of keeping the 1.8 branch alive for a while
longer so that Thunderbird and SeaMonkey can get some overlap with our
1.9.1-based releases.

Robert Kaiser

Mark Finkle

unread,
Sep 23, 2008, 2:42:41 PM9/23/08
to
On Sep 23, 8:27 am, Robert Kaiser <ka...@kairo.at> wrote:
> John O'Duinn wrote:

> > It’ll be sad to see the ending of an era, and there’s still lots of
> > details to sort out, including how to deal with Thunderbird2 support.
>
> SeaMonkey 1.1.x runs into the same problem as Thunderbird, needing
> security fixes probably for longer than December. I understand that
> RelEng will be happy to get rid of the regular FF2 infrastructure but I
> hope we'll have some way of keeping the 1.8 branch alive for a while
> longer so that Thunderbird and SeaMonkey can get some overlap with our
> 1.9.1-based releases.

The XULRunner community also voiced the same concern at the Mozilla
Summit. There are a few deployed XULRunner apps that would benefit
from security releases after December, until they get completely moved
to XULRunner 1.9

Mike Beltzner

unread,
Sep 23, 2008, 5:55:33 PM9/23/08
to Robert Kaiser, dev-pl...@lists.mozilla.org

What do you mean by "alive"? There's different levels of service here,
from nightly buildbots to producing fully localized and branded
Firefox releases. I assume you mean:

- cvs stays up
- nightlies stay building
- any SeaMonkey buildbots running stay running (ie: we don't
dismantle those VMs)

That right?

cheers,
mike

Michael Connor

unread,
Sep 23, 2008, 5:57:36 PM9/23/08
to Mark Finkle, dev-pl...@lists.mozilla.org

Is this an issue related to communication? Or are we just hitting
"six months isn't enough time for platform consumers" here? We can at
least piggyback on Thunderbird-required work, but we should really
consider a different plan for dropping support for platform versions
than for Firefox.

(That should not be interpreted as "MoCo will continue to maintain
XULRunner beyond the corresponding Firefox end of life date" as MoCo
is a product-focused organization, not a platform provider. If the
consumers of XULRunner want extended platform lifecycles, they should
decide as a group how to best do that. The code is there, CVS is
there, and as we've shown with the 1.8.0 releases, we (mozilla.org)
are willing to enable longer lifecycles than what MoCo will support.)

TANSTAAFL applies to XULRunner, like anything else.

-- Mike

Robert Kaiser

unread,
Sep 23, 2008, 10:18:01 PM9/23/08
to
Mike Beltzner wrote:
> What do you mean by "alive"? There's different levels of service here,
> from nightly buildbots to producing fully localized and branded Firefox
> releases. I assume you mean:
>
> - cvs stays up
> - nightlies stay building
> - any SeaMonkey buildbots running stay running (ie: we don't dismantle
> those VMs)
>
> That right?

SeaMonkey nightly/build machines for branch are my private machines, so
no issue with that, and I don't think anyone's thinking on shutting down
cvs, so those are non-issues IMHO.
My issue is actually getting Gecko/Platform security fixes on the 1.8
branch, which is what SeaMonkey and Thunderbird at least will need for a
few months so we can do a few further security releases off the branch
while we don't have stable releases off 1.9.1 yet (and maybe 1-2
releases for overlap after we get our new finals out).

I actually suspect that this time Firefox will get a number of
complaints about ending the updates for that branch as well, due to Mac
OS X 10.3 and Win9x being completely desupported with this EOL, but
we've been through this discussion already.

Robert Kaiser

Simon Paquet

unread,
Sep 24, 2008, 4:16:21 AM9/24/08
to
Michael Connor wrote on 23. Sep 2008:

>> The XULRunner community also voiced the same concern at the
>> Mozilla Summit. There are a few deployed XULRunner apps that
>> would benefit from security releases after December, until
>> they get completely moved to XULRunner 1.9
>
> Is this an issue related to communication? Or are we just
> hitting "six months isn't enough time for platform consumers"
> here? We can at least piggyback on Thunderbird-required work,
> but we should really consider a different plan for dropping
> support for platform versions than for Firefox.
>
> (That should not be interpreted as "MoCo will continue to maintain
> XULRunner beyond the corresponding Firefox end of life date" as MoCo
> is a product-focused organization, not a platform provider.

Well, personally I think that MoCo has an obligation to provide longer
platform support for its former product Thunderbird, given that MoCo
was primarily responsible for the complete lack of further development
of Thunderbird for about a year, which was only fixed recently with
the formation of Mozilla Messaging.

I'm not advocating that the lifecycle of the 1.8 branch is prolonged
until TB3 has been out for six months, but ending its lifecycle in
December, when TB3 will probably be released only 1-2 months later
sounds like a pretty bad decision for me from a general mozilla.org
point of view to me.

That can easily kill a lot of goodwill in the market for Thunderbird.
Given that TB is supposed to generate revenues in the future, I really
wouldn't recommend to walk this route.

Simon

--
Thunderbird/Calendar Localization (L10n) Coordinator
Calendar website maintainer: http://www.mozilla.org/projects/calendar
Calendar developer blog: http://weblogs.mozillazine.org/calendar

Uli Link

unread,
Sep 24, 2008, 5:17:42 AM9/24/08
to
Robert Kaiser schrieb:

> I actually suspect that this time Firefox will get a number of
> complaints about ending the updates for that branch as well, due to Mac
> OS X 10.3 and Win9x being completely desupported with this EOL, but
> we've been through this discussion already.

And any UNIX flavour the vendor does not provide a recent enough
GKT2,glib2,cairo like HP-UX and AIX at the moment.

--
Uli Link

Chris Cooper

unread,
Sep 24, 2008, 9:38:50 AM9/24/08
to Simon Paquet, dev-pl...@lists.mozilla.org
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

John's original message was pretty clear that we're still trying to
figure out what to do with Thunderbird 2.

When Thunderbird 1.5 was EOLed, we supported the Thunderbird 1.5 builds
long after support for Firefox 1.5 had ended. Indeed, we also ended up
supporting Firefox 1.5 for far longer than originally intended due to
last minute requests from a few parties. This why we're giving lots of
loud, advance, public warning this time.

I'm not going to officially put us (the MoCo release team) on the hook
for anything, but if history is any precedent, Thunderbird support will
*likely* continue beyond the EOL for Firefox 2. Precisely how long will
depend on the schedule for Thunderbird 3 and how aggressive Mozilla
Messaging wants to be with major updates.

cheers,
- --
coop

Simon Paquet wrote:
> Well, personally I think that MoCo has an obligation to provide longer
> platform support for its former product Thunderbird, given that MoCo
> was primarily responsible for the complete lack of further development
> of Thunderbird for about a year, which was only fixed recently with
> the formation of Mozilla Messaging.
>
> I'm not advocating that the lifecycle of the 1.8 branch is prolonged
> until TB3 has been out for six months, but ending its lifecycle in
> December, when TB3 will probably be released only 1-2 months later
> sounds like a pretty bad decision for me from a general mozilla.org
> point of view to me.
>
> That can easily kill a lot of goodwill in the market for Thunderbird.
> Given that TB is supposed to generate revenues in the future, I really
> wouldn't recommend to walk this route.

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.3 (Darwin)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFI2kLhqdY6vdJz4M8RAsPkAJ9fU1ArTrIKBpjWybxsWL75weTGywCeMN1j
HEuwXKRvyjOMrjmXcO9qejI=
=ZM+V
-----END PGP SIGNATURE-----

Mark Finkle

unread,
Sep 24, 2008, 10:13:05 AM9/24/08
to

Mike,

What I heard at the summit was an interest to have security fixes
applied to the code, when possible. Even if those patches came from
external people, backporting trunk/branch patches.

Given that we didn't release any official XULRunner for 1.8.1, I think
this is more a "go forward" concern.

Michael Connor

unread,
Sep 25, 2008, 7:37:20 PM9/25/08
to Simon Paquet, dev-pl...@lists.mozilla.org

On 24-Sep-08, at 4:16 AM, Simon Paquet wrote:

> Michael Connor wrote on 23. Sep 2008:
>
>>> The XULRunner community also voiced the same concern at the
>>> Mozilla Summit. There are a few deployed XULRunner apps that
>>> would benefit from security releases after December, until
>>> they get completely moved to XULRunner 1.9
>>
>> Is this an issue related to communication? Or are we just
>> hitting "six months isn't enough time for platform consumers"
>> here? We can at least piggyback on Thunderbird-required work,
>> but we should really consider a different plan for dropping
>> support for platform versions than for Firefox.
>>
>> (That should not be interpreted as "MoCo will continue to maintain
>> XULRunner beyond the corresponding Firefox end of life date" as MoCo
>> is a product-focused organization, not a platform provider.
>
> Well, personally I think that MoCo has an obligation to provide longer
> platform support for its former product Thunderbird, given that MoCo
> was primarily responsible for the complete lack of further development
> of Thunderbird for about a year, which was only fixed recently with
> the formation of Mozilla Messaging.

Mozilla (in some form) will provide support for Thunderbird based on
the official lifecycle policy, like we did for 1.0 and 1.5.

> I'm not advocating that the lifecycle of the 1.8 branch is prolonged
> until TB3 has been out for six months, but ending its lifecycle in
> December, when TB3 will probably be released only 1-2 months later
> sounds like a pretty bad decision for me from a general mozilla.org
> point of view to me.

You are asserting things that aren't based on what I said.
Thunderbird has the same product lifecycle policy as Firefox (six
months after next release). However, we have generally focused on
issues that actually affect Thunderbird when backporting platform
fixes, which is a smaller set than issues that affect Firefox, or some
other platform consumers. "Because we're supporting Thunderbird" is
not a reason to continue to support the whole platform for all
consumers.

> That can easily kill a lot of goodwill in the market for Thunderbird.
> Given that TB is supposed to generate revenues in the future, I really
> wouldn't recommend to walk this route.

I'm glad we're in agreement here.

-- Mike

Reply all
Reply to author
Forward
0 new messages