http://hacks.mozilla.org/2009/06/beautiful-fonts-with-font-face/

The Arabic example should download the OpenType font so it works with Harfbuzz. 
Instead, the ligatures still do not appear correctly.

Fixed with patch to nsChildView.mm inactiveWindowAcceptsMouseEvent to call 
SetFrontProcessWithOptions with kSetFrontProcessFrontWindowOnly. This forces it 
to always activate regardless of whether Gecko would have done it for us or not.

* I read everything above and have demonstrated this bug only occurs on
10.4Fx by testing against this official version of Firefox 4 (not
applicable for startup failure) - specify: 4.0b11

* This is a startup crash or failure to start (Y/N): n

* What steps are necessary to reproduce the bug? 

try to increase font size via cmd+, font remains unchanged, bonk sound emitted, 
cmd- reduces font size as expected, and both work as expected on the intel 
ffx4.0b11


* Describe your processor, computer, operating system and any special
things about your environment.

1.25ghz g4,  aL powerbook, 10.4.11,
Mozilla/5.0 (Macintosh; PPC Mac OS X 10.4; rv:2.0b11pre) Gecko/20110202 
Firefox/4.0b11pre TenFourFox/7450

This is a Classilla-style hack, but it solves our problem (mostly). I got this 
idea from bug 645289 (see 
http://hg.mozilla.org/releases/mozilla-2.0/rev/62e68e402bfd ). If we can 
determine we have any plugin visible, then we can hack ::ScrollTo in 
layout/generic/nsGfxScrollFrame.cpp to always scroll smooth. This uses much 
smaller repaints and significantly reduces artifacting. This will make users 
happier on 4.0 until the axe comes down in 5.0.

Unable to reproduce in the official beta, but it is also using a different 
rendering method, so it is unclear where the bug originates. Applies to both 
Flash and QuickTime, so this is generic to plugins.

Tested with the latest Firefox b11 build.

WebM videos are not playing smoothly. It plays for 3-4 seconds, stops for a 
second and then plays again (like jerks).

Tested with Opera 10.62 and the same video plays smoothly.

In both the cases CPU usage is around 80%-90%.

* I read everything above and have demonstrated this bug only occurs on
10.4Fx by testing against this official version of Firefox 4 (not
applicable for startup failure) - specify:

* This is a startup crash or failure to start (Y/N): N

* What steps are necessary to reproduce the bug? These must be reasonably
reliable.

Enable HTML5 in youtube. 
Play any WebM video (Ex: http://www.youtube.com/watch?v=W15f6xM18wQ)

* Describe your processor, computer, operating system and any special
things about your environment.

iBook 12" Model.
Processor: G4 1.33 GHz
RAM: 1 GB
OS: Mac OSX 10.4.11

* If this is a startup crash or failure to start, have you tried restarting
with a clean profile? Does this fix the problem?



* For crashes or startup failure (optional): paste in any information from
Crash Reporter, or if you are able to run 10.4Fx in gdb, paste in the
backtrace. You can often do it like this from within Terminal.app:

cd appname.app/Contents/MacOS
gdb firefox-bin
run
*crash the app
bt full

* Include any useful additional information, steps you have tried, etc.:

Thanks for your help!

Reviewing my changes to widget, we should backport M604455 to our codebase and 
get the startup win. Our version of the TSM manager is based on 1.9.2.

https://bugzilla.mozilla.org/show_bug.cgi?id=604455

* Does NOT occur in Firefox 4 for Snow Leopard / Intel

* It could be b/c of Rosetta, because I tried tenfourfox on my intel mac (I do 
not have a PPC mac to test on (::cry::)).

* Here is a video of it running on snow leopard Intel: 
http://www.youtube.com/watch?v=OIiz5ih2IBo

* Steps to reproduce:
1) go to http://www.grantgalitz.org/gameboy/
2) load in any gameboy color ROM (.gb or .gbc file) by either loading it via 
base64 copy paste, ajax, or the file api. These are accessed by the emulator's 
file button, of which a menu appears and these options are under the open-as 
sub-menu.
3) Should crash when the file has been decoded and the emulation starts.
4) Yes, this is a GameBoy Color emulator written entirely in JavaScript without 
the use of plugins (You can check it at 
https://github.com/grantgalitz/GameBoy-Online ).

Crash Log:

http://pastebin.com/xejJ8Vfs


* Include any useful additional information, steps you have tried, etc.:

Works fine in Firefox 4 beta 11 For snow leopard (As well as in the firefox 
nightlies)

* I read everything above and have demonstrated this bug only occurs on
10.4Fx by testing against this official version of Firefox 4 (not
applicable for startup crashes) - specify:

* This is a startup crash (Y/N): N

* What steps are necessary to reproduce the bug? These must be reasonably
reliable. Unable to launch the G4 version on a G4 7410 CPU.


* Describe your processor, computer, operating system and any special
things about your environment. G3 Pismo upgraded to G4 7410 CPU @ 533mhz. 1Gb 
RAM 10.5.8.


* If this is a startup crash, have you tried restarting with a clean
profile? Does this fix the problem? n/a



* For crashes (optional): paste in any information from Crash Reporter, or
if you are able to run 10.4Fx in gdb, paste in the backtrace. You can often
do it like this from within Terminal.app:

cd appname.app/Contents/MacOS
gdb firefox-bin
run
*crash the app
bt full


Thanks for your help!

* I read everything above and have demonstrated this bug only occurs on
10.4Fx by testing against this official version of Firefox 4 (not
applicable for startup failure) - SPECIFY VERSION YOU TESTED AGAINST: 4.0 Linux

* Layout bugs MUST be tested against a system WITHOUT hardware
acceleration. Go to Help, Troubleshooting Information to see if your test
system is accelerated. If your test system IS accelerated, you must test on
ANOTHER system that isn't. If this is NOT a layout or display bug, you can
skip this step.

* This is a startup crash or failure to start (Y/N): N

* This is a fault of JavaScript acceleration (Y/N): Y (Use the steps above to
find out. Do NOT report if it is not.)

* What steps are necessary to reproduce the bug? These must be reasonably
reliable.

Help -> Restart with Add-ins disabled
OR
on hushmail.com, click to delete all pending messages

In both cases, the javascript pop-up for confirmation will have no text 
anywhere, but will have the button outlines. Clicking the buttons does nothing. 
If this is done on a web page, the only way to continue is to remove the tab or 
window; if this is the Restart then the only way to continue is to kill 
tenfourfox.

* Describe your processor, computer, operating system and any special
things about your environment.

MacBook Pro G4 7450 processor, OS-X Leopard


* If this is a startup crash or failure to start, have you tried restarting
with a clean profile? Does this fix the problem?



* For crashes or startup failure (optional): paste in any information from
Crash Reporter, or if you are able to run 10.4Fx in gdb, paste in the
backtrace. You can often do it like this from within Terminal.app:

cd appname.app/Contents/MacOS
gdb firefox-bin
run
*crash the app
bt full

* Include any useful additional information, steps you have tried, etc.:

Changing the jit content tracing to off allows the prompt on hushmail to work 
fine.

Thanks for your help!

We need to decide on our final user agent string.

Current thinking is to tag TenFourFox/PPC to the end of a standard Fx4 user 
agent string, in the way we tag Classilla/CFM to Classilla's.

Our old code handled Cmd-~, but not all the other non-menu shortcuts.

Our new code handles all the other non-menu shortcuts, but not Cmd-~.

Regression from issue 22. The simplest way and the one least likely to add 
problems of its own is to add back our code to nsMenuBarX.mm, but only for this 
specific key combination (and any others that also don't work). We need to fix 
this for final.

A)  Help Menu:  

1) "TenFourFox Help" item goes to Mozilla Firefox Help.

2) "Submit Feedback..." item goes to TenFourFox and then redirects to a Firefox 
feedback page

B)  Bookmarks Menu:

"Bookmark This Page" item brings up a window which does not resize properly if 
the title of the page being bookmarked is very long.

* This is a startup crash or failure to start (Y/N):

No

* This is a fault of JavaScript acceleration (Y/N): (Use the steps above to 
find out. Do NOT report if it is not.)

No

* What steps are necessary to reproduce the bug? These must be reasonably 
reliable.

See above

* Describe your processor, computer, operating system and any special
things about your environment.

Mac G4 PPC, OS 10.4.11

* I read everything above and have demonstrated this bug only occurs on
10.4Fx by testing against this official version of Firefox 4 (not
applicable for startup failure) - specify:

* This is a startup crash or failure to start (Y/N): N

* What steps are necessary to reproduce the bug? These must be reasonably 
reliable.

Start tenfourfox on a non english system

* Describe your processor, computer, operating system and any special
things about your environment.

IMac Power PC G4 1,25Ghz

No localization is included into the package. I Try to replace en-en.jar by the 
official fr.jar in chrome directory but didn't work. I manage to change search 
engines, and dictionaries... 

How could we localize tenfourfox ? 

Unique to internal beta 8. The new tab modal dialogues suck and are slow. If 
the performance is not fixed by code freeze, we need to turn that pref off for 
shipping as it is just unacceptably bad even on a G5.

Bugs to watch:
https://bugzilla.mozilla.org/show_bug.cgi?id=613714
https://bugzilla.mozilla.org/show_bug.cgi?id=613876

- If the pref is disabling plugins, then navigator.plugins should also be null, 
and about:plugins should be null too, and so should the Add-ons tab.

- If the pref is disabling plugins, we should not offer the user the chance to 
find or download any.

* I read everything above and have demonstrated this bug only occurs on 10.4Fx 
by testing against this official version of Firefox 4 (not applicable for 
startup failure) - specify:

* This is a startup crash or failure to start (Y/N): N

* What steps are necessary to reproduce the bug? These must be reasonably 
reliable.

1) Set javascript.options.tracejit.content TRUE

2) Go to www.lordofultima.com and play the game. It's free.

3) Once the game loads, click the "Trade" button on the left side of the screen 
(it's next to City Info, Recruitment, and Commands).

4) The Trade window opens, but any attempt to click on a tab at the top of the 
window causes a crash.

* Describe your processor, computer, operating system and any special things 
about your environment.

Power Mac G4 Quicksilver (2002) with OWC Mercury Extreme G4 1.4GHz (7450) 
processor. 1.5GB RAM. Mac OS X 10.5.8.

* If this is a startup crash or failure to start, have you tried restarting 
with a clean profile? Does this fix the problem?

N/A

* For crashes or startup failure (optional): paste in any information from 
Crash Reporter, or if you are able to run 10.4Fx in gdb, paste in the 
backtrace. You can often do it like this from within Terminal.app:

Program received signal EXC_BAD_ACCESS, Could not access memory.
Reason: KERN_INVALID_ADDRESS at address: 0xffff0004
0xffff0004 in ?? ()
(gdb) bt full
#0  0xffff0004 in ?? ()
No symbol table info available.
Cannot access memory at address 0xffff0004
Cannot access memory at address 0xffff0004
Cannot access memory at address 0xffff0004
Cannot access memory at address 0xffff0004
Cannot access memory at address 0xffff0004
Cannot access memory at address 0xffff0004
#1  0xffff0007 in ?? ()
No symbol table info available.
Cannot access memory at address 0xffff0004
Cannot access memory at address 0xffff0004
Cannot access memory at address 0xffff0004
Cannot access memory at address 0xffff0004
Cannot access memory at address 0xffff0004
Cannot access memory at address 0xffff0004
Cannot access memory at address 0xffff0004
Cannot access memory at address 0xffff0004
#2  0xffff0007 in ?? ()
No symbol table info available.
Cannot access memory at address 0xffff0004
Previous frame identical to this frame (gdb could not unwind past this frame)

* Include any useful additional information, steps you have tried, etc.:

This only seems to happen when javascript.options.tracejit.content is set to 
TRUE. If it's set to FALSE, the game (and the Trade window) works properly.

This also only appears to be an issue with TenFourFox 4 final release. Beta 13 
worked fine regardless of the javascript.options.tracejit.content setting.

Thanks for your help!

* This is a startup crash or failure to start (Y/N):

No

* This is a fault of JavaScript acceleration (Y/N): (Use the steps above to 
find out. Do NOT report if it is not.)

No

* What steps are necessary to reproduce the bug? These must be reasonably 
reliable.

When RealPlayer Downloader is invoked (for example, by playing a YouTube video 
in the browser) it calls for Firefox even though 104Fox is the default browser. 
 The video plays normally in the 104Fox browser window, but Firefox attempts to 
start, and then an error message pops up saying that only one copy of Firefox 
can be running at a time.  Clicking "OK" to this message dismisses it and the 
video continues to play normally.

* Describe your processor, computer, operating system and any special
things about your environment.

Mac G4 PPC, MacOS 10.4.11.  Firefox 3.6.14 still present on the system drive 
but is not the default app for any purpose.

* If this is a startup crash or failure to start, have you tried restarting 
with a clean profile? Does this fix the problem?

N/A

* Include any useful additional information, steps you have tried, etc.:

Unloading RealPlayer Downloader and RealPlayer Downloader Agent eliminates the 
problem.  I realize this is not a 104Fox bug as such, but it may indicate that 
104Fox is not fully recognized by the OS as the equivalent of Firefox.

Using TenFourFox 4.0b13pre on 1gHx 10.4.11 eMac.

Bottom scroll bar doen't appear (along with assorted icons as Foxcast and Stop 
Ad, etc). Only when Customize Toolbar is activated does bottom scroll bar 
temporarily appear.

Otherwise brillant and much appreciated work!!

James Greenidge

We have gcc-4.0.1 on Tiger, so there's no good reason we can't have PGO in our 
builds except that we need to set up the infrastructure. I'd prefer this to be 
semiautomated to a baseline test suite, maybe optimized over my common sites as 
a starter. Things to look at:

https://developer.mozilla.org/en/building_with_profile-guided_optimization
https://bugzilla.mozilla.org/show_bug.cgi?id=513508
https://bugzilla.mozilla.org/show_bug.cgi?id=439087

!!!!STOP!!!!  !!!!READ THIS FIRST!!!!

First, make sure you are using the most current version: visit
www.tenfourfox.com to verify.

During the public testing phase, only LIMITED bug reports will be accepted.
This is to reduce the worklist down to the most important items. IF YOU
DON'T READ AND FOLLOW THE BELOW INSTRUCTIONS (and remove this blurb from
your report), YOUR BUG REPORT WILL BE CLOSED!

While 10.4Fx is in source-parity stage (see the Wiki page on
SupportLevels), we build almost directly from Mozilla hg except for those
parts required for compatibility. That means it is imperative that you
verify THE BUG IS IN 10.4Fx SPECIFICALLY. It is quite possible that the bug
is in Firefox, and we don't track those bugs -- Mozilla does.

Except for STARTUP FAILURE, you MUST TEST your bug in Firefox 4 on one
of the tier 1 platforms (Windows XP+, OS X 10.5+ i386 or Linux). If your
bug occurs there too, DO NOT SUBMIT IT TO US: submit it to Mozilla. Bonus
points for testing on Mac specifically, although comparisons with non-Mac
tier 1 builds will be accepted as bug reports.

If your bug occurs ONLY on 10.4Fx, continue with this report.

STARTUP CRASHES are always accepted, especially until we get our own crash
report infrastructure running. Make sure you fill out all the fields below.

-- 8< CUT EVERYTHING UP TO AND INCLUDING THIS LINE FROM YOUR REPORT 8< --

* I read everything above and have demonstrated this bug only occurs on
10.4Fx by testing against this official version of Firefox 4 (not
applicable for startup failure) - specify:

* This is a startup crash or failure to start (Y/N):

* What steps are necessary to reproduce the bug? These must be reasonably
reliable.


* Describe your processor, computer, operating system and any special
things about your environment.


* If this is a startup crash or failure to start, have you tried restarting
with a clean profile? Does this fix the problem?



* For crashes or startup failure (optional): paste in any information from
Crash Reporter, or if you are able to run 10.4Fx in gdb, paste in the
backtrace. You can often do it like this from within Terminal.app:

cd appname.app/Contents/MacOS
gdb firefox-bin
run
*crash the app
bt full

* Include any useful additional information, steps you have tried, etc.:

Thanks for your help!

!!!!STOP!!!!  !!!!READ THIS FIRST!!!!

First, make sure you are using the most current version: visit
www.tenfourfox.com to verify.

During the public testing phase, only LIMITED bug reports will be accepted.
This is to reduce the worklist down to the most important items. IF YOU
DON'T READ AND FOLLOW THE BELOW INSTRUCTIONS (and remove this blurb from
your report), YOUR BUG REPORT WILL BE CLOSED!

While 10.4Fx is in source-parity stage (see the Wiki page on
SupportLevels), we build almost directly from Mozilla hg except for those
parts required for compatibility. That means it is imperative that you
verify THE BUG IS IN 10.4Fx SPECIFICALLY. It is quite possible that the bug
is in Firefox, and we don't track those bugs -- Mozilla does.

Except for STARTUP FAILURE, you MUST TEST your bug in Firefox 4 on one
of the tier 1 platforms (Windows XP+, OS X 10.5+ i386 or Linux). If your
bug occurs there too, DO NOT SUBMIT IT TO US: submit it to Mozilla. Bonus
points for testing on Mac specifically, although comparisons with non-Mac
tier 1 builds will be accepted as bug reports.

If your bug occurs ONLY on 10.4Fx, continue with this report.

STARTUP CRASHES are always accepted, especially until we get our own crash
report infrastructure running. Make sure you fill out all the fields below.

-- 8< CUT EVERYTHING UP TO AND INCLUDING THIS LINE FROM YOUR REPORT 8< --

* I read everything above and have demonstrated this bug only occurs on
10.4Fx by testing against this official version of Firefox 4 (not
applicable for startup failure) - specify:

* This is a startup crash or failure to start (Y/N): N

* What steps are necessary to reproduce the bug? These must be reasonably
reliable.

Go to http://www.macintoshgarden.org/apps/c
type in "German" into the search field, press "Enter"
a result appears, the beachball appears, the app crashes


* Describe your processor, computer, operating system and any special
things about your environment.

Powerbook G4 1,67 GHz, 1,5 GB RAM, 10.5.8,


* If this is a startup crash or failure to start, have you tried restarting
with a clean profile? Does this fix the problem?



* For crashes or startup failure (optional): paste in any information from
Crash Reporter, or if you are able to run 10.4Fx in gdb, paste in the
backtrace. You can often do it like this from within Terminal.app:

cd appname.app/Contents/MacOS
gdb firefox-bin
run
*crash the app
bt full

* Include any useful additional information, steps you have tried, etc.:



Thanks for your help!

The black used for bad OpenGL surfaces seems to leak into the compositor such 
that if there is no defined background, then the black is used and appears as 
the opacity overlay. This is probably Mozilla's bug but we should fix this if 
they do not. A simple first pass would be to set the black colour to white, but 
it probably should simply be set to whatever the default background colour is. 
This needs to be fixed before final release.

* I read everything above and have demonstrated this bug only occurs on
10.4Fx by testing against this official version of Firefox 4 (not
applicable for startup failure) - SPECIFY VERSION YOU TESTED AGAINST:

* Layout bugs MUST be tested against a system WITHOUT hardware
acceleration. Go to Help, Troubleshooting Information to see if your test
system is accelerated. If your test system IS accelerated, you must make
sure it is OFF, OR test on ANOTHER system that isn't. If this is NOT a
layout or display bug, you can skip this step.

* This is a startup crash or failure to start (N):

* This is a fault of JavaScript acceleration (Y): (Use the steps above to
find out. Do NOT report if it is not.)

* What steps are necessary to reproduce the bug? These must be reasonably
reliable.


* Describe your processor, computer, operating system and any special
things about your environment.


* If this is a startup crash or failure to start, have you tried restarting
with a clean profile? Does this fix the problem?



* For crashes or startup failure (optional): paste in any information from
Crash Reporter, or if you are able to run 10.4Fx in gdb, paste in the
backtrace. You can often do it like this from within Terminal.app:

cd appname.app/Contents/MacOS
gdb firefox-bin
run
*crash the app
bt full

* Include any useful additional information, steps you have tried, etc.:

Thanks for your help!


The preferences pane is broken!  I can open the panel but not close it.  Also, 
I cannot change the preferences in the open panel.  eMac G-4 7450 processor.  
Leopard 10.5.8

At some point we need to enable tests and make whatever changes are needed to 
allow the tests to pass on 10.4.

This is a Mozilla issue, but it's worth tracking because we might revert cairo 
if we get enough complaints.

https://bugzilla.mozilla.org/show_bug.cgi?id=595671

Between beta 12 and final, the Ray Trace test is now failing.

It does work on G5, so this is a pathway difference. Issue 37 is the only thing 
that could have caused something like that. Marking high, but waiting to see 
fallout before I kick out a separate version.

Bottom line, we needed to certify this better; it was not a minor change in the 
end.

I have been using the "about:config" variable "font.single-face-list" to make 
"American Typewriter Condensed" my default font. Will that be supported in 
tenforfox?

* I read everything above and have demonstrated this bug only occurs on
10.4Fx by testing against this official version of Firefox 4 (not
applicable for startup failure) - TenFourFox7450

* This is a startup crash or failure to start (Y/N): N

* This is a fault of JavaScript acceleration (Y/N): N

* What steps are necessary to reproduce the bug? 

Go to a website that need a java applet to allow Like http://www.keepvid.com 
TFF craches.

It works when no autorisation is needed like :
http://java.sun.com/applets/jdk/1.4/demo/applets/WireFrame/example4.html 


        Mozilla/5.0 (Macintosh; PPC Mac OS X 10.4; rv:2.0) Gecko/20110324 Firefox/4.0 TenFourFox/7450


Crash report joined.

We have submitted our nanojit to Firefox as

https://bugzilla.mozilla.org/show_bug.cgi?id=624164

After careful consideration, we will ship the nanojit in beta 9, but disabled 
by default. What we need to do to get it default:

- Make it faster in as many circumstances as possible.

- Find the source of the crash that causes it to occasionally die as the 
browser is shutting down.

- Test on G3, where we do not yet have test coverage.

Our about box needs to be a custom job instead of the Minefield default. We 
should have:

- what CPU
- build ID

At the same time we need SLUGGO!

Plan for b9.

Confirmed specific to 10.4Fx, though I have no idea why. STRs cause failure in 
10.4Fx b9, but not in Firefox b9.

1. Type http:// such that the awesome bar appears.
2. Type gibberish such that it does not.

The awesome bar is now "dead" for that window, although hitting TAB for 
suggestions still works, implying it is simply not being displayed.

This is not a blocker, but we need to repair this.

* I read everything above and have demonstrated this bug only occurs on
10.4Fx by testing against this official version of Firefox 4 (not
applicable for startup failure) - specify: 4.0b11

* This is a startup crash or failure to start (Y/N): n

* What steps are necessary to reproduce the bug? 

try to increase font size via cmd+, font remains unchanged, bonk sound emitted, 
cmd- reduces font size as expected, and both work as expected on the intel 
ffx4.0b11


* Describe your processor, computer, operating system and any special
things about your environment.

1.25ghz g4,  aL powerbook, 10.4.11,
Mozilla/5.0 (Macintosh; PPC Mac OS X 10.4; rv:2.0b11pre) Gecko/20110202 
Firefox/4.0b11pre TenFourFox/7450

In issue 26 a user brought up poor WebM speed. This is undoubtedly due to issue 
7, but we should look into writing a PPC-specific decoder path along the lines 
of the ARM and x86 code that already exists to at least speed up the decoding 
step even if rendering still sucks.

-- 8< CUT EVERYTHING UP TO AND INCLUDING THIS LINE FROM YOUR REPORT 8< --

* I read everything above and have demonstrated this bug only occurs on
10.4Fx by testing against this official version of Firefox 4 (not
applicable for startup failure) - specify:

TenFourFx Beta 9

* This is a startup crash or failure to start (Y/N):

No

* What steps are necessary to reproduce the bug? These must be reasonably 
reliable.

Daily use. As far as I know:
"Cmd + a number" to switch to the corresponding tab doesn't work.
"Cmd + + ", press it once, firefox zooms for four times.
"Cmd + Enter" in the location bar doesn't auto-complete "www.*.com"

* Describe your processor, computer, operating system and any special
things about your environment.

G5/PowerPC 970 1.8 GHz, iMac, 10.5.8

* If this is a startup crash or failure to start, have you tried restarting 
with a clean profile? Does this fix the problem?

Clean profile doesn't fix it.

Thanks for your help!

Thanks for developing TenFourFx.

* I read everything above and have demonstrated this bug only occurs on
10.4Fx by testing against this official version of Firefox 4 (not
applicable for startup failure) - specify:
www.vikatan.com does not display tamil characters correctly. FX 3.6 displays it 
correctly. Encoding is the same in both browsers.

* This is a startup crash or failure to start (Y/N): N

* What steps are necessary to reproduce the bug? These must be reasonably 
reliable.
Just visiting www.vikatan.com in both Fx 3.6 and 10.4Fx can show the 
difference. 10.4Fx display tamil. Only certain characters are messed up.

* Describe your processor, computer, operating system and any special
things about your environment.
1.33 GHz PowerPC G4, OS X 10.4.11

* If this is a startup crash or failure to start, have you tried restarting
with a clean profile? Does this fix the problem?
-NA-


* For crashes or startup failure (optional): paste in any information from
Crash Reporter, or if you are able to run 10.4Fx in gdb, paste in the
backtrace. You can often do it like this from within Terminal.app:

cd appname.app/Contents/MacOS
gdb firefox-bin
run
*crash the app
bt full

* Include any useful additional information, steps you have tried, etc.:

Thanks for your help!

Users on the G4/G5 build state that its performance is degraded on 7400/7410, 
so we will generate a separate 7400/7410 build for beta 8.

The title bar colour doesn't match without the tab bar.

https://bugzilla.mozilla.org/show_bug.cgi?id=618770
https://bugzilla.mozilla.org/show_bug.cgi?id=616232

We particularly need to watch for the call in M616232 c7 -- we should just stub 
that out and WONTFIX it for 10.4.

The most recent internal builds for 10.4Fx have an occasional flash of the 
window titlebar line. I'm not sure why. It may be troublesome to fix while this 
code is broken.

We need to set up update infrastructure. Not decided yet if I will do this with 
Sparkle or the built-in Mozilla updates (pointed to floodgap).

Mozilla is forcing the issue early:

http://groups.google.com/group/mozilla.dev.planning/browse_thread/thread/de5f1db
462d2661e

This necessarily includes a whole new build system, to wit:
https://bugzilla.mozilla.org/show_bug.cgi?id=584474

I think I can hack IPC into building, but this is definitely a big hit and will 
slow us down. Also, I hope our debug libxul doesn't get too large.

This will get promoted to critical after final release as we probably can't 
progress to Fx5 without it.

Find those operations that _are_ faster on G5 with the nanojit. Dromaeo sans 
SunSpider is a win even for G5, so we know they exist. Spun off issue 20.

nanojit PPC does not work for Firefox.

https://bugzilla.mozilla.org/show_bug.cgi?id=458077

We also need to be mindful of the various changes Mozilla is doing to 2.0, 
although it seems that some attempt is being made to keep the PPC side up to 
date for right now.

I like Gopher, and this is my project. Rather than two separate codebases, 
however, I will simply implement Mozilla bug 572000 (my patch) if Mozilla does 
not so that people can install OverbiteFF and I can maintain one codebase for 
Gopher support. To be resolved before final release.

Program received signal EXC_BAD_ACCESS, Could not access memory.
Reason: KERN_INVALID_ADDRESS at address: 0xffffffff
0xfffeff20 in objc_msgSend_rtp ()
(gdb) bt
#0  0xfffeff20 in objc_msgSend_rtp ()
#1  0x937a6524 in -[NSApplication sendEvent:] ()
#2  0x9379d960 in -[NSApplication run] ()
#3  0x017132d8 in nsAppShell::Run ()
#4  0x13538f20 in nsAppStartup::Run ()
#5  0x0003e628 in XRE_main ()
#6  0x00003a40 in main ()
(gdb) q

The selector is indeed sendEvent: so this doesn't help us. What it looks like 
is that the menupopup gets destroyed and we don't realize it, try to pass it a 
message, and boom. This should be fixed for the next release.

Followup bug to issue 31. This is very infrequently hit but may be a 
performance win in certain cases. Right now it causes an abort if hit to avoid 
the crash reported in issue 31.

At the layout level, modeled after Classilla's code.

I left the somewhat haphazard window ordering in simply because it was 
sometimes useful, but now that I've worked with it for awhile, it's not 
predictable enough to be a "feature." We should force all clicks in open 
windows to become main windows. We should be careful, as this may make certain 
popups and so on unclickable.

After several days of uptime, random slowdowns will occur. These are not severe 
but they sure are annoying and they cause intermittent beachballs.

These bugs

https://bugzilla.mozilla.org/show_bug.cgi?id=579653
https://bugzilla.mozilla.org/show_bug.cgi?id=614687

imply it's Mozilla's fault and implicates the cycle collector which is indeed 
the most likely cause. This bug

https://bugzilla.mozilla.org/show_bug.cgi?id=565217

will probably solve it, or at least hide it.

In Firefox 3.6, I'm able to close a tab by pressing Cmd (Apple key) and 
clicking on it. This is very useful, since I run the browser on a PowerBook G4 
with a single-key touchpad, and considerably speeds up tab management. This 
functionality seems to be gone from Tenfourfox :( 

As of b11, nanojit is not shippable for chrome:

- When no location, "undefined" sometimes appears in the location bar.
- Unexplained Components bugs with addons.
- Crash still when closing the app.

The problem is, it's unclear what is Mozilla's bug and what is ours, and these 
could be symptoms of several different actual problems gemisched together.