Known issue with Firefox 4+ code

Talk about code development, features, specific bugzilla bugs, enhancements, patches, and other highly technical things.

Moderator: satrow

Forum rules
Please keep everything here strictly on-topic.
This board is meant for Pale Moon source code development related subjects only like code snippets, patches, specific referenced Bugzilla bugs, mercurial, etc.

This is not for tech support! Please do not post tech support questions in the "Development" board!
Please make sure not to use this board for support questions. Most "bug reports" do not belong in this board and should initially be posted in Community Support or other relevant support boards.

Please keep things on-topic as this forum will be used for reference for Pale Moon development. Expect topics that aren't relevant as such to be moved or deleted.
User avatar
Moonchild
Pale Moon guru
Pale Moon guru
Posts: 23925
Joined: 2011-08-28, 17:27
Location: 58°2'16"N 14°58'31"E
Contact:

Known issue with Firefox 4+ code

Unread post by Moonchild » 2011-11-02, 15:35

There's one major bug in the current Firefox code: cairo has some seriously broken code for animated images and it can saturate whatever core the lib is running on if there's more than a handful animated gifs or the like on the page, this can make things laggy.
Mozilla has been VERY lax about this problem (known since later v4 betas) -- and I don't have the know-how myself to fix it in the very complex graphics library that it is. It needs a cairo code guru (or more than one, as it turned out).

Mozilla has, as of today, marked the core bug for this problem bug #666446 as RESOLVED-FIXED, but the code implementation will not be included in the Firefox code base until Mozilla10 (which is scheduled to be released, apparently, end of January 2012 :x )
This highlights a major drawback with the current release carrousel that Mozilla uses: new fixes for major issues will be put on the Trunk and as such will always run behind 3 versions, unless they are specifically duplicated for beta and aurora (which most won't).

Thankfully, this rather invasive fix has been promised to be implemented in Aurora next week - which means I may be tempted to make an Aurora-based Pale Moon build for the next major release.
Last edited by Moonchild on 2012-03-09, 09:43, edited 1 time in total.
Reason: No longer sticky with the upcoming v11 release
City of Heroes public server: https://www.moonshard.org/ -- Vote for it on cohservers.com

"If you want to build a better world for yourself, you have to be willing to build one for everybody." -- Coyote Osborne
Image

User avatar
megaman
Moon Magic practitioner
Moon Magic practitioner
Posts: 2874
Joined: 2011-08-30, 16:40
Location: H-town

Re: Known issue with Firefox 4+ code

Unread post by megaman » 2011-11-02, 22:20

I am still wondering, will it be called Azure then?
Curiosity leads to a world of experiments.

User avatar
Moonchild
Pale Moon guru
Pale Moon guru
Posts: 23925
Joined: 2011-08-28, 17:27
Location: 58°2'16"N 14°58'31"E
Contact:

Re: Known issue with Firefox 4+ code

Unread post by Moonchild » 2011-11-03, 07:17

megaman wrote:I am still wondering, will it be called Azure then?
It has nothing to do with Azure; that's the cairo replacement they are working on.
City of Heroes public server: https://www.moonshard.org/ -- Vote for it on cohservers.com

"If you want to build a better world for yourself, you have to be willing to build one for everybody." -- Coyote Osborne
Image

User avatar
Moonchild
Pale Moon guru
Pale Moon guru
Posts: 23925
Joined: 2011-08-28, 17:27
Location: 58°2'16"N 14°58'31"E
Contact:

Re: Known issue with Firefox 4+ code

Unread post by Moonchild » 2011-11-03, 07:18

OK, so they backed out the solution yet again...
A bit more information -

We noticed a performance regression on Windows XP, as mentioned by Ed in
comment 410. As part of the investigation of this, bug 691775. After discussing
it on IRC for a bit, we came to the conclusion that the best course of action
is to take a slightly different approach in implementing the resolution, and
not cover quite as much ground with this rearchitecture as we had originally
planned.

Our original intention was to make images subscribe to notifications for
refresh with the refresh driver under the following conditions: 1) the image
was decoded and is animating (has more than 1 frame), or 2) the image is
decoding. As Kyle mentioned, this bug has enormous potential for regressions,
and could possibly set us back quite a bit if we're not careful. As such, we've
decided to be a bit more conservative, in case there's something else going on
none of us are seeing, and implement this in a way that solves the problem of
1) without extending our reach for 2).

This decision was reached because we believe it safeguards us against possible
really bad regressions, as yet unknown or invisible, in the future. I'll be
working on this as quickly as I can, so it will be in the release as soon as it
can be correctly and completely implemented.
City of Heroes public server: https://www.moonshard.org/ -- Vote for it on cohservers.com

"If you want to build a better world for yourself, you have to be willing to build one for everybody." -- Coyote Osborne
Image

User avatar
Moonchild
Pale Moon guru
Pale Moon guru
Posts: 23925
Joined: 2011-08-28, 17:27
Location: 58°2'16"N 14°58'31"E
Contact:

Re: Known issue with Firefox 4+ code

Unread post by Moonchild » 2012-01-20, 12:11

The string of bugs related to this issue have been finalized and marked solved, and should all be implemented in v11
City of Heroes public server: https://www.moonshard.org/ -- Vote for it on cohservers.com

"If you want to build a better world for yourself, you have to be willing to build one for everybody." -- Coyote Osborne
Image

Locked