(please reply-to / follow-up to: dev-plann...@lists.mozilla.org /  
mozilla.dev.planning)

Hi everyone,

On June 17th, 2008, we shipped Firefox 3 / Gecko 1.9. As per the  
release roadmap (http://wiki.mozilla.org/ReleaseRoadmap) we're  
approaching the end of the six months for which Firefox 2 / Gecko  
1.8.1 support is planned. Support in this case means continuing to  
develop the security and stability patches for those code-bases, as  
well as issuing releases of Firefox and minor updates to the Firefox 2  
product.

Before setting a final date for this "end of life" activity (sounds  
drastic, doesn't it?) we wanted to get feedback on any requirements or  
issues that would prevent us from making the upcoming Firefox  
2.0.0.19 / Gecko 1.8.1.19 security and stability release our final  
support release on that product/branch.

The following issues have already been considered and covered:

* ensuring that we're able to issue a major update offer to the latest  
Firefox 3 release to all users (on platforms supported by Firefox 3)  
even after we officially end the support lifecycle for Firefox 2 , and  
thinking through the user experience there

* watching add-on compatibility numbers to ensure that the vast  
majority of Firefox 2 add-ons still under development have been made  
compatible with Firefox 3

* getting in touch with partners who bundle and distribute Firefox 2  
to make them aware

* considering implications for web services integrated into the  
product (ie: SafeBrowsing)

* ensuring that we don't offer major updates to people on unsupported  
platforms

* we have not begun triage for 1.8.1.19 in earnest, so please nominate  
any bugs that you feel should be included in this release, but realize  
that our development resources are finite, so decisions will need to  
be made :)

Presently 2/3rds of our users are using Firefox 3, with more than 50%  
accepting the first major upgrade offer back in late August. We're  
looking
through Hendrix and other sources to understand why people didn't want  
to upgrade and ensure that those bugs have been fixed - if you know of  
any that would result in the product not working on a user's system,  
please point it out. Please do not point out UI changes - those  
objections are well understood.

If you can think of other reasons why we shouldn't move towards end of  
life for Firefox 2 / Gecko 1.8.1, please let us know by replying (to dev-plann...@lists.mozilla.org
  or following-up to mozilla.dev.planning)

cheers,
mike

Mike Beltzner wrote on 29. Oct 2008:

Please see the discussion around John O'Duinn's post "Firefox2
desupport now only 3 months away" back in September here in this
newsgroup.

My biggest concern is that no other major mozilla.org Gecko consumer
(Thunderbird, Sunbird, SeaMonkey, Camino) has yet released a major
release based on Mozilla 1.9 code.

While one could argue that SeaMonkey, Sunbird and Camino are
niche-products and one mustn't wait for them to update, I'm
especially concerned about Thunderbird here.

Based on the current TB3 release planning, TB3 will be released 3-4
months after the Gecko 1.8 EOL, when TB2 is still the stable release.

Comments from my Chris Cooper and Mike Connor in the discussion
mentioned above removed my concerns at the time, but I can't find
anything in your post that addresses this.

See my original concerns:
http://groups.google.com/group/mozilla.dev.planning/msg/58159b24f4001abb

Mike's and Chris' answers:
http://groups.google.com/group/mozilla.dev.planning/msg/77b2e27865fb4719
http://groups.google.com/group/mozilla.dev.planning/msg/76fec4b026bf9d05

Cya
Simon

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

I'm recounting this second-hand, but I recall there being a problem with
printing support when one of the admins at my school tested FF 3.

On 29-Oct-08, at 9:26 AM, Simon Paquet wrote:

As with that previous discussion, right now our focus is understanding  
the impact on Firefox. Your points are quite valid, though, as Firefox  
is often tied to Gecko, and we should be careful when examining what  
actions we should take once Firefox 2 reaches end of life to ensure  
that we don't starve consumers still relying on Gecko 1.8.1.

Do we have similar targets for Seamonkey, Camino, Sunbird,  
InstantBird, Miro, etc?

Indeed - I should have mentioned that we'd registered the impact of  
decommissioning Gecko 1.8.1 infrastructure (in terms of test machines  
and emphasis on security issues) on other consumers. As Coop and  
mconnor stated, historically we've continued to support the platform  
even after the product has reached end of life (see Gecko 1.8) - it's  
just a matter of understanding what parts of the infrastructure will  
no longer be needed once Firefox 2 is out of the picture.

cheers,
mike

I am concerned with Gecko 1.8.1 getting EOLed when both SeaMonkey 2.0
and Thunderbird 3.0 not likely to be out until at least 2009/Q1. Under
optimum conditions Gecko 1.8.1 should not e EOLed until six months after
SeaMonkey 2.0+Thunderbird 3.0 are released, but I understand the
resource constraints on having to backport security and stability fixes.

Phil

--
Philip Chee <phi...@aleytys.pc.my>, <philip.c...@gmail.com>
http://flashblock.mozdev.org/ http://xsidebar.mozdev.org
Guard us from the she-wolf and the wolf, and guard us from the thief,
oh Night, and so be good for us to pass.
[ ]There's no future in time travel.
* TagZilla 0.066.6

[[ Setting follow-ups ]]

--
Philip Chee <phi...@aleytys.pc.my>, <philip.c...@gmail.com>
http://flashblock.mozdev.org/ http://xsidebar.mozdev.org
Guard us from the she-wolf and the wolf, and guard us from the thief,
oh Night, and so be good for us to pass.
[ ]The 486SX: Intel's test of your gullibility
* TagZilla 0.066.6

Yup, keep reading in the thread and you'll see that concern has indeed  
been registered! As has been discussed before, it's not like we're  
going to lock the CVS trees and walk away, and I don't think all  
development will stop ... but the active development of Firefox  
releases based on that platform and the milestone-based development  
cycles would probably stop.

So EOL doesn't mean "everyone stop coding and doing reviews", more  
that "we don't have a team doing weekly triage on blockers and  
milestone releases".

With that in mind - still concerned?

cheers,
mike

On 29-Oct-08, at 1:11 PM, Philip Chee wrote:

On Oct 29, 2008, at 10:20 AM, Mike Beltzner wrote:

The triage and release team that currently works on Firefox and  
Thunderbird 2.0.0.x releases will continue to triage requests for  
Thunderbird 2.0.0.x and maintain its releases until six months after  
the release of Thunderbird 3.

Note that this will mean that browser-specific security and stability  
bugs will likely be ignored/minused. We'll only be considering bugs  
that affect Thunderbird 2.0.0.x.

That being said, I know Gecko 1.8.0 is still maintained by Linux  
distros and we could hand over most of the branch to willing and  
capable parties as we did in that case.

-Sam

I've heard two issues over and over again:-

 - the places fsync problem and related urlbar performance problems
   (gone in 3.1, one hopes)
 - must have a particular extension

Regarding the second, you said you were "watching add-on compatibility
numbers to ensure that the vast majority of Firefox 2 add-ons *still
under development* have been made compatible with Firefox 3".  Well, in
the majority of cases where I've had that come up, the extension in
question was *not* still under development, but was clearly still
heavily used (just looking at the trail of comments on AMO begging for
an upgrade).

I think it might be worthwhile to devote some of our attention to
popular but unmaintained extensions.  We could, for instance, have
some sort of "adoption agency" on AMO which lists unmaintained
extensions and encourages third-party developers to take them over.
(The Debian package adoption procedure might be a model.)  It might
even be worth spending internal developer effort on updating extensions
where the fix looks to be easy.

zw

Won't this affect Seamonkey?

Ignoring I can understand (with the assumption that someone on the
Seamonkey team will be given the bits to do triage).  Minusing doesn't
seem right.

-Boris

On Wednesday 2008-10-29 10:33 -0700, Zack Weinberg wrote:

This would only be possible for open-source extensions.  (I don't
think AMO has metadata to indicate which extensions are open-source,
so figuring out which these are can involve a bit of digging.)

-David

--
L. David Baron                                 http://dbaron.org/
Mozilla Corporation                       http://www.mozilla.com/

As Simon has already mentioned, we're planning on a similar target for
SeaMonkey 2 as for Thunderbird 3.

Robert Kaiser

On 10/29/2008 4:48 AM, Mike Beltzner wrote [in part]:

Over the period of 7-20 October 2008, I logged the UA strings of 2097
visits to an eclectic mix of 20 of my Web pages.  Bots and crawlers
accounted for 39% of those visits.  Recognizable browsers accounted for
61%.

Of the 420 visits by some Gecko browser, 401 allowed me to identify the
Gecko version.  rv:1.9 appeared in 75% of those 401; rv:1.8 appeared in
25%.

While this might seem to strengthen the argument in favor of ending
support for Gecko 1.8.x -- 3/4 (not merely 2/3) using Gecko 1.9.x -- 25%
of the Gecko users (including SeaMonkey and Camino users, both of which
appeared in my log) should not be ignored.

--
David E. Ross
<http://www.rossde.com/>

Go to Mozdev at <http://www.mozdev.org/> for quick access to
extensions for Firefox, Thunderbird, SeaMonkey, and other
Mozilla-related applications.  You can access Mozdev much
more quickly than you can Mozilla Add-Ons.

Yes. I'm not concerned with active development. Just security and
stability backports. *And* fixing any regressions that break SeaMonkey
1.1.x.

I am reassured, mostly.

Phil

--
Philip Chee <phi...@aleytys.pc.my>, <philip.c...@gmail.com>
http://flashblock.mozdev.org/ http://xsidebar.mozdev.org
Guard us from the she-wolf and the wolf, and guard us from the thief,
oh Night, and so be good for us to pass.
[ ]If guns are outlawed, can we use swords?
* TagZilla 0.066.6

Oh *now* I'm *not* reassured. I suspect the Camino people won't be either.

Phil

--
Philip Chee <phi...@aleytys.pc.my>, <philip.c...@gmail.com>
http://flashblock.mozdev.org/ http://xsidebar.mozdev.org
Guard us from the she-wolf and the wolf, and guard us from the thief,
oh Night, and so be good for us to pass.
[ ]A 100% right of return both ways.
* TagZilla 0.066.6

Mozdev recently set up a formal takeover procedure for orphaned projects
(after an orphaned project turned out not to be orphaned after all, just
pinning for the fjords). Perhaps you could talk to David Boswell or some
of the other board memebers about details.

Phil

--
Philip Chee <phi...@aleytys.pc.my>, <philip.c...@gmail.com>
http://flashblock.mozdev.org/ http://xsidebar.mozdev.org
Guard us from the she-wolf and the wolf, and guard us from the thief,
oh Night, and so be good for us to pass.
[ ].signature: Permission denied.
* TagZilla 0.066.6

On 30-Oct-08, at 12:21 AM, Philip Chee wrote:

I'll wait until you get to the point where Sam clarifies and says that  
by "ignored" he meant "we won't be doing milestone release triage  
style stuff on it", not "we won't help with patches and reviews as  
always."

We've been through this before with the Gecko 1.8 end of life, and  
everyone survived. Don't panic. What we're primarily talking about is  
stopping the 6-8 week milestone release cycles.

cheers,
mike

So How does this jibe with the last word in this sentence?:

Also I don't see any clarifications by Sam in the newsgroup. Something
plugging up the mail/news gateway?

Phil

--
Philip Chee <phi...@aleytys.pc.my>, <philip.c...@gmail.com>
http://flashblock.mozdev.org/ http://xsidebar.mozdev.org
Guard us from the she-wolf and the wolf, and guard us from the thief,
oh Night, and so be good for us to pass.
[ ]It is a rather pleasant experience to be alone in a bank at night.
* TagZilla 0.066.6

So that means that SeaMonkey 1.x (and Camino) will intentionally be made
insecure and punished for never releasing their xpfe-based stuff on top
of Gecko 1.9?

Robert Kaiser

No, and straw men aren't appreciated. Please read on in the thread for where
I clarify Sam's words for Phil.

cheers,
mike

Thanks, I just wanted to get that clarified a bit more, as Sam's post
sounded very much like anything browser-specific being ignored and
dropped completely (which I felt was probably not the intention but the
sound of it).

Robert Kaiser

On 30-Oct-08, at 10:38 AM, Robert Kaiser wrote:

No worries. Sam's hatred for Camino is well documented. ;)

cheers,
mike

On Oct 29, 5:48 am, Mike Beltzner <beltz...@mozilla.com> wrote:

Mike,
Firefox 3 (even the latest release I tested) has serious printing
issues
that have prevented our organization from migrating.
Character spacing is way off on the printout, dpending on the font
used on a webpage (like courier) characters become even non-readable.

We have web based applications that generate invoices, packing lists,
etc.
That are printed via webbrowser to a PDF (Acrobat). Same problem with
direct
print to a laser printer however.
Firefox 2 printing works perfectly (so does IE ;-))   Firefox 3 is a
disaster.
So I would urge to not discontinue support for Firefox 2 until the
Firefox 3
printing issues are resolved.

thanks
Volker

I assume you reported this bug as soon as you discovered it, right?  Can
you please point me to the bug report?

-Boris

Right. I seem to remember I filed a bug saying a.m.o. should track this
metadata, but it didn't get accepted :-(

Gerv