2005-11-11 Branch respin (Firefox 1.5 RC3)

Fixes:

  • Fixed: 316025 - Crash at allaccess.com and MySpace (over 20% of all RC2 crashes).
  • Fixed: 315189 - Setting innerHTML to a string containing a <style> tag causes display problems (e.g. at Yahoo! Mail Beta) or crashes.
  • Fixed: Several OS/2 bugs.

No known major regressions.

Gecko 1.8 branch checkins between 2005-11-07 12:00 and 2005-11-11 16:05

Windows builds: Windows nightly, Windows hourly

Linux builds: Linux nightly, Linux hourly

Mac builds: Mac nightly, Mac hourly

86 Responses to “2005-11-11 Branch respin (Firefox 1.5 RC3)”

  1. peakshysteria Says:

    correction: exiting any tabs in any way (ctrl + w, middleclick tab or tab bar) makes the same error alert. Also ctrl + t doesnt make an insert point (which forces me to use ctrl + l or the mouse).

  2. peakshysteria Says:

    Hmmm, found out that it was the Ez sidebar extension which caused the errors. Tested it for a friend. No good on branch builds then…yet. No everything seems smooth again so far…

  3. nestastnik Says:

    Advertisement: Want to quickly run out of 1GB of memmory? Are you using Mozilla/5.0 (Windows; U; Windows NT 5.2; en-US; rv:1.8) Gecko/20051118 ? Don’t you have any money to get it free? Don’t hasitate, our company brought you this NEW exciting product!

    Just visit:
    http://www.2020media.com/nav.html?hosting-ASP_net.htm

    and let your Firefox eat eat eat eat …. and eat

  4. MechR Says:

    http://www.2020media.com/nav.html?hosting-ASP_net.htm

    Ahh, now that one I can confirm. High CPU, and memory escalates quickly.

  5. peakshysteria Says:

    hmm, nestastnik; http://www.magic-kinder.com/ still forces my CPU up in a constant of 100% (while playing those Flash games). Playing the games and browsing through the sites is no problem besides the peaking CPU thou, he.

    And yeah http://www.2020media.com/nav.html?hosting-ASP_net.htm breaks my FF. Still using Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.8) Gecko/20051119 Firefox/1.5 ID:2005111903 with Extensions (enabled: 13, disabled: 0):

    * ChatZilla 0.9.68.5.1
    * Content Holder 0.2.2005110501
    * Flashblock 1.3.3
    * Forecastfox 0.8.2.4
    * InfoLister 0.9c
    * Linkification 1.1.6
    * Nightly Tester Tools 0.7.9.10
    * Restart Firefox 0.3.1
    * Statusbar Clock 1.7.1
    * Tabbrowser Extensions 2.0.2005111702
    * Text Link 1.3.2005070402
    * Translate 0.6.0.8
    * Update Channel Selector 1.0.1

  6. CooLynX Says:

    This 2020media.com is a killer site for Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.8) Gecko/20051116 Firefox/1.5, too. :/

  7. nestastnik Says:

    peakshysteria, I don’t have flash installed for security reson, althought I know there exist some extension to block it. Can’t confirm, I will never install flash, I hate this technology, I assume because I don’t have kids yet heh =)

    and CPU is normal 0% usage on this site.

    Mozilla/5.0 (Windows; U; Windows NT 5.2; en-US; rv:1.8) Gecko/20051120 Firefox/1.5

    I’ll report of fixed update tommorow, today I still needed to unzip all the files again eh :-(

  8. peakshysteria Says:

    he, Flashblock takes care of those unwanted flash object and animations………but still leaves you with an option to play or view flash if you want. An awesome extension. But then again you dont use extensions nestastnik……

  9. Hank Roberts Says:

    Mozilla/5.0 (Macintosh; U; PPC Mac OS X Mach-O; en-US; rv:1.8) Gecko/20051119 Firefox/1.5

    Definitely eating up memory — and leaving it ‘inactive’ — these last few days’ builds all seem to have the same problem. After a while the response time starts to lag, and I see half the meg of RAM is inactive.

    Rebooting fixes the problem. Hey, like Windows … Waaaah.

  10. Igor N. Avtaev Says:

    Jesse, see the site: http://www.linuxshare.ru/tmp/ffox.

    Attention to Russian word in parentheses.

  11. Jesse Ruderman Says:

    Interesting. Frames don’t seem to matter: loading http://www.linuxshare.ru/tmp/ffox/index.cgi?item=%F1%CE%D5%D3 and waiting is sufficient to trigger the undesired behavior. I wonder whether this is closely related to https://bugzilla.mozilla.org/show_bug.cgi?id=261929 and https://bugzilla.mozilla.org/show_bug.cgi?id=284474 .

  12. Jesse Ruderman Says:

    “Янус” would become “/=CA” if you took the UTF-16 or UCS2 representation of “Янус”, threw away the high bytes, and treated the low bytes as UTF-8 or ASCII. Something in Gecko must be doing a “lossy UTF-16 to ASCII conversion” inappropriately, perhaps on a string it expected would already have been escaped. This is almost certainly a bug rather than a case where it’s unclear which charset should be used (as in bug 284474 and bug 261929).

  13. Jesse Ruderman Says:

    Loading http://www.linuxshare.ru/tmp/ffox/index.cgi?item=%F1%CE%D5%D3 doesn’t behave as you might expect in Firefox 1.0.7 either. In Firefox 1.0.7, it sends the refresh URL (at least the query part) as UTF-8. That’s arguably a reasonable thing to do, but it isn’t what your server expects and I don’t think it’s consistent with Firefox 1.0.7’s behavior for links.

  14. Jesse Ruderman Says:

    Igor, I filed https://bugzilla.mozilla.org/show_bug.cgi?id=317254 for your problem.

  15. Diego Pires Plentz Says:

    http://www.clarodigital.com.br breaks lots of times FF RC3 last build.

  16. nestastnik Says:

    Try this:

    Delete whole FF, prg files directory and 3 profiles directory(app data and local settings), and registry key(LM/talkback)
    UNZIP FF to new directory.
    Later.

    HELP- CHECK FOR UPDATES

    no updates found I mean, comeon!

    Mozilla/5.0 (Windows; U; Windows NT 5.2; en-US; rv:1.8) Gecko/20051120 Firefox/1.5

  17. peakshysteria Says:

    using Mozilla/5.0 (Windows; U; Windows NT 5.2; en-US; rv:1.8) Gecko/20051121 Firefox/1.5 ID:2005112104. Help –> check for updates doesnt work (same as from 20051120 to 20051121). Trying manuall install for 1122 vers.

  18. peakshysteria Says:

    hmm, now using Mozilla/5.0 (Windows; U; Windows NT 5.2; en-US; rv:1.8) Gecko/20051122 Firefox/1.5 ID:2005112211

    seems that the update channel selector automatically after a manual install sets itself to release channel update (instead of the usual nightly which is my usual choice). Not sure, but it looks like the update channel selector even after changing it from release to nightly goes searching for release update. Anyone else encountered something like this?

    Also two latest versions seems to load pages way slower than the earlier releases at first. But if pages are loaded once they load way faster the next time……..guess this is the cache kicking in….too bad it dont follow updates. Very irritating to begin all over again.

  19. Igor N. Avtaev Says:

    Update channel – nightly. After release 2005111405 of updatings does not find out.

    Mozilla/5.0 (Windows; U; Windows NT 5.0; ru-RU; rv:1.9a1) Gecko/20051114 Firefox/1.6a1 – Build ID: 2005111405

  20. CooLynX Says:

    Looks like it is time for K-meleon – no memory leaks, fast and light. =]

  21. peakshysteria Says:

    he, cant se the point of using K-meleon instead of official FF releases……..no point in comparing nightly branch (or trunk) builds with the latest stable official release. Nightly builds are unstable, thats the why we test……if its no fun….install an official release……

  22. tester Says:

    Kmeleon is OKish, but what about that URL http://www.clarodigital.com.br, good one for Crash testing, isn’t it? Already mentioned…

  23. peakshysteria Says:

    for those with autoupdate trouble…..this somewhat cryptict blog try to say something about cause….i think;

    http://coop.deadsquid.com/?p=530

  24. Hank Roberts Says:

    Problems displaying this Quicktime movie:
    http://goes.gsfc.nasa.gov/goeswest/pacific/colormov/0000_latest.mov

    (offset frame)
    Using Mozilla/5.0 (Macintosh; U; PPC Mac OS X Mach-O; en-US; rv:1.8) Gecko/20051124 Firefox/1.5

  25. peakshysteria Says:

    regarding autoupdate # Coop Says (http://coop.deadsquid.com/?p=530):
    November 24th, 2005 at 14:56

    Manual updates will be necessary until this is fixed, yes.

  26. Diego Pires Plentz Says:

    Sorry guys, http://www.clarodigital.com.br/ mvoed to http://www.claro.com.br/

  27. nestastnik Says:

    I hope Firefox won’t reach final state, e.g. 1.5 final if at least 110% of here mentioned bugs are corrected ;]

    OK so I am not banned from autoupdate ;), as I see others with similar troubles..

    I would DEFINITELY like clear private data button, I really miss it. Maybe its time to write my 1st extension :-)

  28. peakshysteria Says:

    eh nestastnik, do you really mean Ctrl + Shift + Del isnt enough? Hmmm, i’ve heard of you mouse-men…myself i’m a keybord-man hehe….Anyway check out X extension at: http://extensionroom.mozdev.org/clav/#x

    could maybe be something for you. Think i have heard of more, but this’ll do for now……to early in the morning…….

  29. nestastnik Says:

    thanks for the shortcut…

    I had this problem today, browsed a few sites, have 6-7 browser WINDOW opened, closing them and opening periodically each with 3 tabs etc…unless just one APPLICATION of FF was visible on the taskbar and I liked to end session. I closed the last one window and another windwos suddenly appeared out of the blue :-) with 2 tabs, so I closed it too and another one appeared from somewhere :-) I nearly forgot I had this windows opened, they weren’t visible on taskbar. And looking to Task Manager, win2k3 has freed much more memmory, more than in usuall cases, e.g. a little swapping began.

    Mozilla/5.0 (Windows; U; Windows NT 5.2; en-US; rv:1.8) Gecko/20051120 Firefox/1.5

  30. Daniel DeVaney Says:

    This build gave me memory leak that resulted in my system giving me a video card error, then when I restarted and opened Firefox it opened really skinny with all toolbars showing and all my bookmarks DELETED.
    The worst part is that even after uninstalling it, deleting the whole programs file for FF, and removing all remnants with CC CLeaner; I reinstalled it and it STILL does it. My whole damn system is screwed up!

  31. peakshysteria Says:

    For those complaining over high memory loss, check this out; http://primates.ximian.com/~federico/news-2005-11.html#moz-images

    Its a proposal for linux thou……Not interesting for myself since my FF mem. is ok. But DeVaney i’m getting back to you later tonight (since my OS also is acting up…)……….Been nice to now something about your themes and extensions an FF vers offcourse. And do you run 64Bit system (linux or Win)??

  32. Hank Roberts Says:

    Daniel DeVaney — Which “this build” were using when your system got screwed up?
    What OS/platform, and what build date?

  33. Hank Roberts Says:

    I just had an unexpected quit with Mozilla/5.0 (Macintosh; U; PPC Mac OS X Mach-O; en-US; rv:1.8) Gecko/20051126 Firefox/1.5

    Talkback was open — and didn’t record anything. It didn’t quit, just missed the event completely.

    I have an OSX crash log for the event — is there a way to manually put it into Talkback, that I’m missing here?

  34. Jesse Ruderman Says:

    You can’t manually add an OSX crash log to Talkback, but you can paste the first few lines of a stack trace here, or file a bug, or search Bugzilla or Talkback using the first line of the stack trace. If you end up filing a new bug, paste the first few lines of the stack trace into the bug report and upload the entire crash report as a text/plain attachment.

  35. Hank Roberts Says:

    I’m not sure if this is a “stack trace” —
    OS Version: 10.3.9 (Build 7W98)
    … Command: firefox-bin
    Path: /Users/hank/Applications/Firefox.app/Contents/MacOS/firefox-bin
    Version: 1.5 (1.5)
    PID: 22966
    Thread: 0

    Exception: EXC_BAD_ACCESS (0x0001)
    Codes: KERN_PROTECTION_FAILURE (0x0002) at 0x00000000

    Thread 0 Crashed:
    0 > 0x00000000 0 + 0
    1 libmozjs.dylib 0x06003544 JS_InitClass + 0x190
    2 libmozjs.dylib 0x06040004 js_InitObjectClass + 0x98
    3 libmozjs.dylib 0x06002464 JS_SetGlobalObject + 0x118
    4 libmozjs.dylib 0x06002a10 JS_ResolveStandardClass + 0x270
    5 org.mozilla.firefox 0x000866c4 nsXPConnect::InitClasses(JSContext*, JSObject*) + 0xf4
    eam*) + 0x68

    34 org.mozilla.firefox 0x00693f18 nsMacMessagePump::GetEvent(EventRecord&) + 0x74
    35 org.mozilla.firefox 0x00693e74 nsMacMessagePump::DoMessagePump() + 0x30
    36 org.mozilla.firefox 0x00300014 nsAppShell::Run() + 0x38
    37 org.mozilla.firefox 0x0039fe3c nsAppStartup::Run() + 0x3c
    38 org.mozilla.firefox 0x00014338 XRE_main + 0xed0
    39 org.mozilla.firefox 0x0000f6a8 start + 0x1b0
    40 org.mozilla.firefox 0x0000f528 start + 0x30
    (that’s the end of Thread 0) …

    Whole thing available to anyone who wants it, but I gather these usually aren’t helpful. If there’s any particular chunk that would indicate why Talkback failed to notice Firefox crashing, I’d guess that’s what’s more important than the crash itself.

  36. Hank Roberts Says:

    A thought —

    I often have both Firefox and Camino
    open at the same time — could this cause a problem with TalkBack working?

    (Why two browsers at once, you may ask? Camino is much faster on my G3/900mhz Powerbook — with it, I can look at several pages I like to check, while Firefox is working at loading a different site. But Camino doesn’t have AdBlock, Aardvark, or NoScript, so I don’t use it on pages I don’t already know I can stand. It has a less capable crapfilter (or one I don’t yet understand well enough to take proper advantabe of).)