The Official Win32 20100905 Trunk build is out.

[Morpog]

Does it work when you hover over them? The browser should switch through them live?

Yep it works.

[owen.c93]

FML, ok thanks guys.

[Recall]

Is the aero-peek tab previews working for anyone?

WFM with 6 tabs.

It is still pretty useless when the "peeks" are still suffering from poor dithering...

https://bugzilla.mozilla.org/show_bug.cgi?id=524468

[geeknik]

Is the aero-peek tab previews working for anyone?

WFM with 6 tabs.

Does it work when you hover over them? The browser should switch through them live?

Works for me, 10 tabs. Browser switches through them just fine. =)

[mattyv]

Re: https://bugzilla.mozilla.org/show_bug.cgi?id=572488 that landed in the last nightly, anyone finding active tabs not very distinguishable from inactive ones? I've tried a number of system themes, using the Firefox default with no personas.

Note: applies to Linux builds only.

[Omega X]

Re: https://bugzilla.mozilla.org/show_bug.cgi?id=572488 that landed in the last nightly, anyone finding active tabs not very distinguishable from inactive ones? I've tried a number of system themes, using the Firefox default with no personas.

Note: applies to Linux builds only.

It looks like the way Windows tabs were before they adjusted the background contrast.

[pal-moz]

Re: https://bugzilla.mozilla.org/show_bug.cgi?id=572488 that landed in the last nightly, anyone finding active tabs not very distinguishable from inactive ones? I've tried a number of system themes, using the Firefox default with no personas.

Note: applies to Linux builds only.

https://bugzilla.mozilla.org/show_bug.cgi?id=593650

[_Alexander]

There is some visual glitches with tab animations and accelerated layers, bug # ?

EDIT:
Also,
6th is Labor Day thus I am not sure whether or not this effects anything. Hopefully not the ETA for performance regression fix...

[KWierso]

From #gfx on IRC:

[00:27] <joe> hold on to your butts, guys
[00:27] <joe> we're green on birch
[00:27] <joe> which means i'm pushing to m-c
[00:28] * tn cheers.

Edit: D3D just got turned on by default again.


EDIT:EDIT: But now this:

[00:41] <joe> fffu-
[00:43] <joe> Bas: ping
[00:45] <joe> roc: also ping
[00:45] <roc> hi
[00:45] <roc> did you see my comment?
[00:45] <joe> yes
[00:46] <joe> thus i am panicking a little
[00:46] <joe> because i pushed literally seconds before you made that comment
[00:46] <roc> I meant to post it earlier on my way out hours ago, but I got a mid-air collision
[00:46] <joe> crap
[00:46] <joe> so, in my tests using the clipping example, we remain transparent
[00:46] <joe> are there specific things you had in mind?
[00:47] <joe> (by 'the clipping example' i mean layout/reftests/border-radius/clipping-3.html)

So who knows if it'll stick this time...

[mattyv]

Thanks Omega, pal-moz. I really should sign up to Bugzilla one of these days...

[michaell522]

6th is Labor Day thus I am not sure whether or not this effects anything. Hopefully not the ETA for performance regression fix...

Well, Mozilla is spread all over the world as you probably know, and the developers (even the employed ones) sometimes work at weekends or other odd times. But for those in the US and Canada, I guess the activity on Monday will be equivalent to a weekend day, maybe less.

I'm still not entirely clear what "the" performance fix people are asking about is - the layers bug linked earlier is being worked on by Robert O'Callahan in New Zealand, and Monday isn't a holiday there.. but if it needs review from someone who's not working that might hold it up. But the discussion in that bug doesn't seem to reflect the range and severity of the performance issues people are mentioning. I guess when that and the couple of other fixes coming soon are done, it will be easier to target remaining issues...

[a;skdjfajf;ak]

Geez, each 'new and improved' update is making the browser slower and slower not faster. FishIE is now 1/2 what it was, 13fps, after the turn-on of D3d.

Build based on cset: http://hg.mozilla.org/mozilla-central/rev/37f65e179e42 OK 24fps - no worse than yesterday
Build based on cset: http://hg.mozilla.org/mozilla-central/rev/b57dda2ee56d Bad, now fps is 13

So.. not directly result of D3D.. but something else in one or roc's patches is making it worse..

I can't file.. can someone file please.

[myzar74]

Geez, each 'new and improved' update is making the browser slower and slower not faster. FishIE is now 1/2 what it was, 13fps, after the turn-on of D3d.

Build based on cset: http://hg.mozilla.org/mozilla-central/rev/37f65e179e42 OK 24fps - no worse than yesterday
Build based on cset: http://hg.mozilla.org/mozilla-central/rev/b57dda2ee56d Bad, now fps is 13

So.. not directly result of D3D.. but something else in one or roc's patches is making it worse..

I can't file.. can someone file please.

Roc said in the performance bug that the D3D slow down with D2D will be fixed in https://bugzilla.mozilla.org/show_bug.cgi?id=593268

[flying fox]

When Accelerated Layers are enabled, the page became corrupted and diagonal after I scroll it (it's worse when smooth scrolling is on). Using the latest hourly on Windows XP, Core2 @1.86, Onboard display adapter, Intel G41 Express Chipset.
http://i54.tinypic.com/rucw77.jpg

[michaell522]

Geez, each 'new and improved' update is making the browser slower and slower not faster.

To some extent, this is happening, because the feature freeze is beta 6, and the freeze for that is next week. So if anything is going to make it into Firefox 4, it has to happen now. From that point onwards, things should only get better

Edit: (Except of course there will be a release branch at some point, and then things on trunk might get worse again when they start doing work for Firefox 5 or whatever the next version might be numbered...)