FF 3.0.5 is OK; FF 3.1.x & 3.2.x keep crashing

I just discovered FF 3.0.5, FireFox-3.1b3pre.en-US.os2.zip and
Firefox-3.2a1pre.en-US.os2.zip and wanted to give them a try.

While 3.0.5 at first glance continues to work fine, the latter 2 keep
crashing here (even with new profile) shortly after the browser window
comes up with:

Killed by SIGILL
pid=0x022f ppid=0x0022 tid=0x0001 slot=0x0094 pri=0x0200 mc=0x0001
N:\MOZILLA-TEST\FF3.X\FIREFOX.EXE
LIBC063 ffffffff:ffffffff
cs:eip=005b:200c3e80      ss:esp=0053:0011cfb8      ebp=0011f570
 ds=0053      es=0053      fs=150b      gs=0000     efl=00212246
eax=0011d564 ebx=20911d00 ecx=0011d434 edx=00000000 edi=200bda5c
esi=200c3e80
Process dumping was disabled, use DUMPPROC / PROCDUMP to enable it.

To me that looks very similar to what I wrote on 2008-12-23 in
"SM 2.0a2/a3pre.en* keeps crashing when loading an URL"

System is W4 FP16+ on AMD K6-2/500
0
Wolfi
1/13/2009 8:02:52 AM
mozilla.dev.ports.os2 2337 articles. 0 followers. Post Follow

42 Replies
1295 Views

Similar Articles

[PageSpeed] 47
Get it on Google Play
Get it on Apple App Store

Am 13.01.09 02.02 schrieb Wolfi:
> I just discovered FF 3.0.5, FireFox-3.1b3pre.en-US.os2.zip and
> Firefox-3.2a1pre.en-US.os2.zip and wanted to give them a try.
> 
> While 3.0.5 at first glance continues to work fine, 

Well, looks I was blowing my horn too early :-(

FF 3.0.5 doesn't switch anymore to the language of an installed language
pack.
I already noticed the same issue with 3.0.4, but didn't yet investigate
it any further, other than to un- and re-install the langpack a few times.
With 3.05 it doesn't work with the old, nor with a new profile.

Doe anyone know where the selected language info is stored nowadays and
how on can manage to force FF to switch to the installed
3.0.5-langpack_de.xpi language?
0
Wolfi
1/13/2009 8:36:27 AM
3.2a1 runs fine here...

Try -safe-mode to disable plugins.

JMS

0
KOI8
1/13/2009 2:33:29 PM
On 01/13/09 12:02 am, Wolfi wrote:
> I just discovered FF 3.0.5, FireFox-3.1b3pre.en-US.os2.zip and
> Firefox-3.2a1pre.en-US.os2.zip and wanted to give them a try.
>
> While 3.0.5 at first glance continues to work fine, the latter 2 keep
> crashing here (even with new profile) shortly after the browser window
> comes up with:
>
> Killed by SIGILL
> pid=0x022f ppid=0x0022 tid=0x0001 slot=0x0094 pri=0x0200 mc=0x0001
> N:\MOZILLA-TEST\FF3.X\FIREFOX.EXE
> LIBC063 ffffffff:ffffffff
> cs:eip=005b:200c3e80      ss:esp=0053:0011cfb8      ebp=0011f570
>   ds=0053      es=0053      fs=150b      gs=0000     efl=00212246
> eax=0011d564 ebx=20911d00 ecx=0011d434 edx=00000000 edi=200bda5c
> esi=200c3e80
> Process dumping was disabled, use DUMPPROC / PROCDUMP to enable it.
>
> To me that looks very similar to what I wrote on 2008-12-23 in
> "SM 2.0a2/a3pre.en* keeps crashing when loading an URL"
>
> System is W4 FP16+ on AMD K6-2/500

Can you change the line
pref("javascript.options.jit.content",      true);
to false. The line will be in PROGRAM_DIRECTORY\greprefs\all.js around 
line 512 and see if it affects the sigill?
Dave
0
Dave
1/13/2009 3:51:40 PM
Am 13.01.09 09.51 schrieb Dave Yeo:
> On 01/13/09 12:02 am, Wolfi wrote:
>> I just discovered FF 3.0.5, FireFox-3.1b3pre.en-US.os2.zip and
>> Firefox-3.2a1pre.en-US.os2.zip and wanted to give them a try.
>>
>> While 3.0.5 at first glance continues to work fine, the latter 2 keep
>> crashing here (even with new profile) shortly after the browser window
>> comes up with:
>>
>> Killed by SIGILL
>> pid=0x022f ppid=0x0022 tid=0x0001 slot=0x0094 pri=0x0200 mc=0x0001
>> N:\MOZILLA-TEST\FF3.X\FIREFOX.EXE
>> LIBC063 ffffffff:ffffffff
>> cs:eip=005b:200c3e80      ss:esp=0053:0011cfb8      ebp=0011f570
>>   ds=0053      es=0053      fs=150b      gs=0000     efl=00212246
>> eax=0011d564 ebx=20911d00 ecx=0011d434 edx=00000000 edi=200bda5c
>> esi=200c3e80
>> Process dumping was disabled, use DUMPPROC / PROCDUMP to enable it.
>>
>> To me that looks very similar to what I wrote on 2008-12-23 in
>> "SM 2.0a2/a3pre.en* keeps crashing when loading an URL"
>>
>> System is W4 FP16+ on AMD K6-2/500
> 
> Can you change the line
> pref("javascript.options.jit.content",      true);
> to false. The line will be in PROGRAM_DIRECTORY\greprefs\all.js around
> line 512 and see if it affects the sigill?


Ooohh right, that's what Peter told me back then for SM 2.0a2+.
I completely forgot about that part :-[

And YES, :-) it makes all the difference in the world :-)

Now FF 3.1+ stays open.

Wolfi
0
Wolfi
1/13/2009 5:29:46 PM
Am 13.01.09 02.36 schrieb Wolfi:
> Am 13.01.09 02.02 schrieb Wolfi:
>> I just discovered FF 3.0.5,  and wanted to give them a try.
>>
>> While 3.0.5 at first glance continues to work fine, 
> 
> Well, looks I was blowing my horn too early :-(
> 
> FF 3.0.5 doesn't switch anymore to the language of an installed language
> pack.
> I already noticed the same issue with 3.0.4, but didn't yet investigate
> it any further, other than to un- and re-install the langpack a few times.
> With 3.05 it doesn't work with the old, nor with a new profile.
> 
> Doe anyone know where the selected language info is stored nowadays and
> how on can manage to force FF to switch to the installed
> 3.0.5-langpack_de.xpi language?

Well, now that FireFox-3.1b3pre and Firefox-3.2a1pre  also stay open,
thanks to  pref("javascript.options.jit.content", false);
in \PROGRAM_DIRECTORY\greprefs\all.js

those two also don't switch language after installing the respective
LangPack.

Anyone having better luck?
0
Wolfi
1/13/2009 6:00:26 PM
On 01/13/09 09:29 am, Wolfi wrote:
>>> Killed by SIGILL
>>> >>  pid=0x022f ppid=0x0022 tid=0x0001 slot=0x0094 pri=0x0200 mc=0x0001
>>> >>  N:\MOZILLA-TEST\FF3.X\FIREFOX.EXE
>>> >>  LIBC063 ffffffff:ffffffff
>>> >>  cs:eip=005b:200c3e80      ss:esp=0053:0011cfb8      ebp=0011f570
>>> >>     ds=0053      es=0053      fs=150b      gs=0000     efl=00212246
>>> >>  eax=0011d564 ebx=20911d00 ecx=0011d434 edx=00000000 edi=200bda5c
>>> >>  esi=200c3e80
>>> >>  Process dumping was disabled, use DUMPPROC / PROCDUMP to enable it.
>>> >>
>>> >>  To me that looks very similar to what I wrote on 2008-12-23 in
>>> >>  "SM 2.0a2/a3pre.en* keeps crashing when loading an URL"
>>> >>
>>> >>  System is W4 FP16+ on AMD K6-2/500
>> >
>> >  Can you change the line
>> >  pref("javascript.options.jit.content",      true);
>> >  to false. The line will be in PROGRAM_DIRECTORY\greprefs\all.js around
>> >  line 512 and see if it affects the sigill?
>
>
> Ooohh right, that's what Peter told me back then for SM 2.0a2+.
> I completely forgot about that part:-[
>
> And YES,:-)  it makes all the difference in the world:-)
>
> Now FF 3.1+ stays open.
>

I wonder if the JIT compiler is using instruction(s) that are not 
supported by the K6-2? Seems that you are the only one still running a 
K6 and also the only one getting a SIGILL (ANSI: Illegal instruction) 
with the GCC 3.3.5 compiled versions. I get a sigsegv (ANSI: 
Segmentation fault) when compiled with GCC 3.4.6.
Perhaps something to be added to the readme
Dave
0
Dave
1/13/2009 6:38:43 PM
Am 13.01.09 02.02 schrieb Wolfi:
> I just discovered FF 3.0.5, FireFox-3.1b3pre.en-US.os2.zip and 
> Firefox-3.2a1pre.en-US.os2.zip and wanted to give them a try.

There seems to be another problem (besides the one with language packs
not taking effect anymore) with the new Gecko 1.9 versions, I'm afraid :-(

They have a tendency to just crash here within some 24h or so.
This time, Gecko/2008121717 Firefox/3.0.5 was idling through the night,
when almost 12hrs later I found this in its error.log:

[ConvConfHandler] isPreferred contentType: application/pdf

Killed by SIGSEGV
pid=0x02ab ppid=0x0022 tid=0x0001 slot=0x00a7 pri=0x0200 mc=0x0001
N:\MOZILLA-TEST\FF3.X\FIREFOX.EXE
UCONV 0:000000eb
cs:eip=005b:1fee00eb      ss:esp=0053:0011eeec      ebp=0011f1c8
 ds=0053      es=0053      fs=150b      gs=0000     efl=00212246
eax=00000000 ebx=00000001 ecx=0000000b edx=00000104 edi=00130c8a
esi=0011ef80
Process dumping was disabled, use DUMPPROC / PROCDUMP to enable it.

All I had done previously in respect to PDFs, was some 14hrs earlier,
click onto 2 links, which turned out to be offering PDF documents and
Cancel them to be loaded. According to Addons -> Plugins I don't have
one related to PDFs loaded.

Unfortunately with those other "some once a day" crashes, I never did
find anything in error.log

Do others see similar things?

> System is W4 FP16+ on AMD K6-2/500
0
Wolfi
1/14/2009 8:43:09 PM
Wolfi wrote:

> 
> Do others see similar things?
> 

   Yes.  I abandoned the FF3 series some months back. 3.0.5 is the last 
version I've bothered with and it crashed often for me. Besides the 
printing issues and the horrid fonts, FF3 would often terminate while 
seemingly doing nothing. No error logs, of course, or I would have 
reported them. FF3 would almost always die in less than two days of 
being open and about 8 hours of interaction. Flash is not installed on 
my system. For reference, FF2 usually lasts until it chews up around 900 
meg of RAm (about four days.)

   If there is something I can do to help track this problem down, 
please instruct me.

--Mark
0
Mark
1/14/2009 10:03:49 PM
Am 14.01.09 14.43 schrieb Wolfi:
> Am 13.01.09 02.02 schrieb Wolfi:
>> I just discovered FF 3.0.5, FireFox-3.1b3pre.en-US.os2.zip and 
>> Firefox-3.2a1pre.en-US.os2.zip and wanted to give them a try.
> 
> There seems to be another problem (besides the one with language packs
> not taking effect anymore) with the new Gecko 1.9 versions, I'm afraid :-(
> 
> They have a tendency to just crash here within some 24h or so.
> This time, Gecko/2008121717 Firefox/3.0.5 was idling through the night,
> when almost 12hrs later I found this in its error.log:
> 
> [ConvConfHandler] isPreferred contentType: application/pdf
> 
> Killed by SIGSEGV
> pid=0x02ab ppid=0x0022 tid=0x0001 slot=0x00a7 pri=0x0200 mc=0x0001
> N:\MOZILLA-TEST\FF3.X\FIREFOX.EXE
> UCONV 0:000000eb
> cs:eip=005b:1fee00eb      ss:esp=0053:0011eeec      ebp=0011f1c8
>  ds=0053      es=0053      fs=150b      gs=0000     efl=00212246
> eax=00000000 ebx=00000001 ecx=0000000b edx=00000104 edi=00130c8a
> esi=0011ef80
> Process dumping was disabled, use DUMPPROC / PROCDUMP to enable it.
> 
> All I had done previously in respect to PDFs, was some 14hrs earlier,
> click onto 2 links, which turned out to be offering PDF documents and
> Cancel them to be loaded. According to Addons -> Plugins I don't have
> one related to PDFs loaded.
> 
> Unfortunately with those other "some once a day" crashes, I never did
> find anything in error.log
> 
>> System is W4 FP16+ on AMD K6-2/500

Well and like clockwork, another crash of FF 3.0.5 after some 20hrs or
so. This time I did find something in error.log:

Killed by SIGSEGV
pid=0x02fc ppid=0x0022 tid=0x0001 slot=0x00b2 pri=0x0200 mc=0x0001
N:\MOZILLA-TEST\FF3.X\FIREFOX.EXE
UCONV 0:000049aa
cs:eip=005b:1fee49aa      ss:esp=0053:0011ed4c      ebp=0011f068
 ds=0053      es=0053      fs=150b      gs=0000     efl=00212213
eax=022301e0 ebx=00000001 ecx=00000000 edx=00000000 edi=2eddd2c0
esi=0011ee20
Process dumping was disabled, use DUMPPROC / PROCDUMP to enable it.

Is this telling something to anyone?
0
Wolfi
1/16/2009 1:01:54 AM
On 1/15/2009 5:01 PM, Wolfi wrote:

> pid=0x02fc ppid=0x0022 tid=0x0001 slot=0x00b2 pri=0x0200 mc=0x0001
> N:\MOZILLA-TEST\FF3.X\FIREFOX.EXE
> UCONV 0:000049aa
> cs:eip=005b:1fee49aa      ss:esp=0053:0011ed4c      ebp=0011f068
>  ds=0053      es=0053      fs=150b      gs=0000     efl=00212213
> eax=022301e0 ebx=00000001 ecx=00000000 edx=00000000 edi=2eddd2c0
> esi=0011ee20
> 
> Is this telling something to anyone?

I think UCONV has something to do with UniCode.  Innotek was 
distributing a patched version at one point, because of memory leaks in 
the original IBM version.  Maybe if you post the file size, date, and 
time info, others can compare against yours.
0
Steve
1/16/2009 2:34:10 AM
Am 15.01.09 20.34 schrieb Steve Wendt:
> On 1/15/2009 5:01 PM, Wolfi wrote:
> 
>> pid=0x02fc ppid=0x0022 tid=0x0001 slot=0x00b2 pri=0x0200 mc=0x0001
>> N:\MOZILLA-TEST\FF3.X\FIREFOX.EXE
>> UCONV 0:000049aa
>> cs:eip=005b:1fee49aa      ss:esp=0053:0011ed4c      ebp=0011f068
>>  ds=0053      es=0053      fs=150b      gs=0000     efl=00212213
>> eax=022301e0 ebx=00000001 ecx=00000000 edx=00000000 edi=2eddd2c0
>> esi=0011ee20
>>
>> Is this telling something to anyone?
> 
> I think UCONV has something to do with UniCode.  Innotek was
> distributing a patched version at one point, because of memory leaks in
> the original IBM version.  Maybe if you post the file size, date, and
> time info, others can compare against yours.

 3.10.01  6.45         23.331     61 a---  uconv.dll

Build Level Display Facility Version 6.12.675 Sep 25 2001
(C) Copyright IBM Corporation 1993-2001
Signature:       @#IBM:14.083#@ Unicode Conversion Functions
Vendor:          IBM
0
Wolfi
1/16/2009 2:46:03 AM
On 01/15/09 06:46 pm, Wolfi wrote:

>> I think UCONV has something to do with UniCode.  Innotek was
>> distributing a patched version at one point, because of memory leaks in
>> the original IBM version.  Maybe if you post the file size, date, and
>> time info, others can compare against yours.
> 
>  3.10.01  6.45         23.331     61 a---  uconv.dll

That's the supposedly buggy one:
  7-17-02  10:31a     23331           0  uconv.dll
10-03-01   6:45a     23331          61  uconv.dll.save

Unfortunately, the bldlevel info (and the size, I see) are the same.

I'm not sure if it is available elsewhere, but you can try this:
http://www.os2bbs.com/os2news/uconv.zip
0
Steve
1/16/2009 4:08:13 AM
Hi Wolfi,

>  3.10.01  6.45         23.331     61 a---  uconv.dll
> Signature:       @#IBM:14.083#@ Unicode Conversion Functions
> Vendor:          IBM

4/09/02  17:30      18339           0  uconv_dll

Signature:       @#IBM:14.093#@ Unicode Conversion Functions
unreleased

Available at http://www.os2site.com/sw/graphics/fonts/index.html

uconv_dll.zip at the bottom of the page, and a slightly earlier
one that looks like the same one Steve is talking about.

Cheers
IBM
0
Blacklisted
1/16/2009 10:09:49 AM
Am 15.01.09 22.08 schrieb Steve Wendt:
> On 01/15/09 06:46 pm, Wolfi wrote:
> 
>>> I think UCONV has something to do with UniCode.  Innotek was
>>> distributing a patched version at one point, because of memory leaks in
>>> the original IBM version.  Maybe if you post the file size, date, and
>>> time info, others can compare against yours.
>>
>>  3.10.01  6.45         23.331     61 a---  uconv.dll
> 
> That's the supposedly buggy one:
>  7-17-02  10:31a     23331           0  uconv.dll
> 10-03-01   6:45a     23331          61  uconv.dll.save
> 
> Unfortunately, the bldlevel info (and the size, I see) are the same.
> 
> I'm not sure if it is available elsewhere, but you can try this:
> http://www.os2bbs.com/os2news/uconv.zip

After poking around a bit here locally I found the later one in XR_M017
and also in XR_C005_CP2-V4.52, so it appears to be the same one, that
you are suggesting:

 4.09.02 17.30         23.331     61 a---  UCONV.DLL
 3.10.01  6.45         23.331     61 a---  uconv.dll_14.083_buggy

Signature:       @#IBM:14.093#@ Unicode Conversion Functions
Vendor:          IBM
Revision:        14.93
File Version:    14.93
Description:     Unicode Conversion Functions - Copyright IBM


I replaced it and am now curious to see, if it makes any difference.
0
Wolfi
1/16/2009 6:58:57 PM
Mine is 14.083, same size, but from 29-04-2004.
Strange...

0
KOI8
1/16/2009 7:32:17 PM
On 16.01.09 02:01, Wolfi wrote:

> Well and like clockwork, another crash of FF 3.0.5 after some 20hrs or
> so. This time I did find something in error.log:
>
> Killed by SIGSEGV
> pid=0x02fc ppid=0x0022 tid=0x0001 slot=0x00b2 pri=0x0200 mc=0x0001
> N:\MOZILLA-TEST\FF3.X\FIREFOX.EXE
> UCONV 0:000049aa
> cs:eip=005b:1fee49aa      ss:esp=0053:0011ed4c      ebp=0011f068
>   ds=0053      es=0053      fs=150b      gs=0000     efl=00212213
> eax=022301e0 ebx=00000001 ecx=00000000 edx=00000000 edi=2eddd2c0
> esi=0011ee20
> Process dumping was disabled, use DUMPPROC / PROCDUMP to enable it.
>
> Is this telling something to anyone?

Do you have an old UCONV.DLL on your system? It normally is in \OS2\DLL
and you should have at least version 14.093 (as bldlevel tells me here).
-- 
Please     | Official Warpzilla Ports: http://www.mozilla.org/ports/os2/
reply in   |
newsgroup  |          Enhanced OS/2 builds: http://pmw-warpzilla.sf.net/
     Steve's Warpzilla Tips: http://www.os2bbs.com/os2news/Warpzilla.html
0
Peter
1/18/2009 1:34:28 PM
On 16.01.09 19:58, Wolfi wrote:

> After poking around a bit here locally I found the later one in XR_M017
> and also in XR_C005_CP2-V4.52, so it appears to be the same one, that
> you are suggesting:
>
>   4.09.02 17.30         23.331     61 a---  UCONV.DLL
>   3.10.01  6.45         23.331     61 a---  uconv.dll_14.083_buggy
>
> Signature:       @#IBM:14.093#@ Unicode Conversion Functions
> Vendor:          IBM
> Revision:        14.93
> File Version:    14.93
> Description:     Unicode Conversion Functions - Copyright IBM
>
>
> I replaced it and am now curious to see, if it makes any difference.

Oh, I see this message only now. This should help. (Mozilla itself
doesn't use the system Unicode functions but the included SQLite does.
The version in Firefox 3.1pre and 3.2pre already has a fix for this
problem, but the one in 3.0.x probably not.)
-- 
Please     | Official Warpzilla Ports: http://www.mozilla.org/ports/os2/
reply in   |
newsgroup  |          Enhanced OS/2 builds: http://pmw-warpzilla.sf.net/
     Steve's Warpzilla Tips: http://www.os2bbs.com/os2news/Warpzilla.html
0
Peter
1/18/2009 1:37:23 PM
On 14.01.09 23:03, Mark wrote:

> Yes. I abandoned the FF3 series some months back. 3.0.5 is the last
> version I've bothered with and it crashed often for me. Besides the
> printing issues and the horrid fonts, FF3 would often terminate while
> seemingly doing nothing.

Great that people appreciate the sleepless nights that I spent looking
at Mozilla source code or posting here answering the same stupid
question for the 20th time. ;-) (When reading such things I am somewhat
happy that I recently have so little time to work on Mozilla stuff.)

So, what "horrid fonts" are you talking about? Or did you not read
README.txt?

> No error logs, of course, or I would have reported them.

It does create them if you redirect the console output to some file.
Like this
    cd <firefox_installation_directory>
    firefox > firefox.log 2>&1
Sometimes, there is also information about a crash in \POPUPLOG.OS2.

> FF3 would almost always die in less than two days of
> being open and about 8 hours of interaction. Flash is not installed on
> my system.

That is already great as Flash is responsible for 90-95% of the crashes
on OS/2.

> If there is something I can do to help track this problem down, please
> instruct me.

I guess you have already seen the messages about UCONV.DLL in this
thread. Otherwise, logs (see above) _might_ tell us something.
-- 
Please     | Official Warpzilla Ports: http://www.mozilla.org/ports/os2/
reply in   |
newsgroup  |          Enhanced OS/2 builds: http://pmw-warpzilla.sf.net/
     Steve's Warpzilla Tips: http://www.os2bbs.com/os2news/Warpzilla.html
0
Peter
1/18/2009 1:47:52 PM
Peter Weilbacher <newsspam@weilbacher.org> wrote:

> Great that people appreciate the sleepless nights that I spent looking
> at Mozilla source code 

Make a paypal "donate" button on your homepage!!

0
KOI8
1/18/2009 9:05:05 PM
Mark <spam__magnet@dslextreme.com> wrote:

> FF3 would almost always die in less than two days of 
> being open and about 8 hours of interaction.

Did you turn off the javascript compiler?

0
KOI8
1/18/2009 9:06:01 PM
On 18.01.09 22:05, Jеns Mаrtin Schlаttеr wrote:
> Peter Weilbacher<newsspam@weilbacher.org>  wrote:
>
>> Great that people appreciate the sleepless nights that I spent looking
>> at Mozilla source code
>
> Make a paypal "donate" button on your homepage!!

No, don't want to do that. I have thought about that every now and then.
But first I don't like Paypal and second if I took money then I would
feel (even more) obliged to work on this stuff, even when I'm not in the
mood or have more important things to do.
-- 
Please     | Official Warpzilla Ports: http://www.mozilla.org/ports/os2/
reply in   |
newsgroup  |          Enhanced OS/2 builds: http://pmw-warpzilla.sf.net/
     Steve's Warpzilla Tips: http://www.os2bbs.com/os2news/Warpzilla.html
0
Peter
1/18/2009 9:45:26 PM
On 01/18/09 01:06 pm, Jеns Mаrtin Schlаttеr wrote:

>> FF3 would almost always die in less than two days of 
>> being open and about 8 hours of interaction.
> 
> Did you turn off the javascript compiler?

I believe the JIT stuff is new in version 3.1.
0
Steve
1/19/2009 2:08:11 AM
Am 15.01.09 22.08 schrieb Steve Wendt:
> On 01/15/09 06:46 pm, Wolfi wrote:
> 
>>> I think UCONV has something to do with UniCode.  Innotek was
>>> distributing a patched version at one point, because of memory leaks in
>>> the original IBM version.  Maybe if you post the file size, date, and
>>> time info, others can compare against yours.
>>
>>  3.10.01  6.45         23.331     61 a---  uconv.dll
> 
> That's the supposedly buggy one:
>  7-17-02  10:31a     23331           0  uconv.dll
> 10-03-01   6:45a     23331          61  uconv.dll.save
> 
> Unfortunately, the bldlevel info (and the size, I see) are the same.
> 
> I'm not sure if it is available elsewhere, but you can try this:
> http://www.os2bbs.com/os2news/uconv.zip

Thanks a lot, Steve :-)
The new DLL seems to have cured my problem.
Gecko/2008121717 Firefox/3.0.5 is now running more than 30hrs and for
the very first time without crashing  :-)

0
Wolfi
1/21/2009 7:40:35 PM
On 21.01.09 20:40, Wolfi wrote:
> Am 15.01.09 22.08 schrieb Steve Wendt:
>> On 01/15/09 06:46 pm, Wolfi wrote:
>>
>>>> I think UCONV has something to do with UniCode.  Innotek was
>>>> distributing a patched version at one point, because of memory leaks in
>>>> the original IBM version.  Maybe if you post the file size, date, and
>>>> time info, others can compare against yours.
>>>   3.10.01  6.45         23.331     61 a---  uconv.dll
>> That's the supposedly buggy one:
>>   7-17-02  10:31a     23331           0  uconv.dll
>> 10-03-01   6:45a     23331          61  uconv.dll.save
>>
>> Unfortunately, the bldlevel info (and the size, I see) are the same.
>>
>> I'm not sure if it is available elsewhere, but you can try this:
>> http://www.os2bbs.com/os2news/uconv.zip
>
> Thanks a lot, Steve :-)
> The new DLL seems to have cured my problem.
> Gecko/2008121717 Firefox/3.0.5 is now running more than 30hrs and for
> the very first time without crashing  :-)

Great.

Steve, the one in the ZIP package, is that now the original IBM version
or the Innotek patched one?
-- 
Please     | Official Warpzilla Ports: http://www.mozilla.org/ports/os2/
reply in   |
newsgroup  |          Enhanced OS/2 builds: http://pmw-warpzilla.sf.net/
     Steve's Warpzilla Tips: http://www.os2bbs.com/os2news/Warpzilla.html
0
Peter
1/21/2009 10:15:10 PM
On 1/21/2009 2:15 PM, Peter Weilbacher wrote:

>>> I'm not sure if it is available elsewhere, but you can try this:
>>> http://www.os2bbs.com/os2news/uconv.zip
>>
>> Gecko/2008121717 Firefox/3.0.5 is now running more than 30hrs and for
>> the very first time without crashing  :-)
> 
> Steve, the one in the ZIP package, is that now the original IBM version
> or the Innotek patched one?

That is the one from Innotek, but I don't think they patched it (I think 
it was an unofficial IBM fix).  I think the one that Ian pointed to on 
os2site.com is the same as the latest official IBM one.
0
Steve
1/21/2009 10:34:52 PM
Peter Weilbacher wrote:
> On 21.01.09 20:40, Wolfi wrote:
>> Am 15.01.09 22.08 schrieb Steve Wendt:
>>> On 01/15/09 06:46 pm, Wolfi wrote:
>>>
>>>>> I think UCONV has something to do with UniCode.  Innotek was
>>>>> distributing a patched version at one point, because of memory 
>>>>> leaks in
>>>>> the original IBM version.  Maybe if you post the file size, date, and
>>>>> time info, others can compare against yours.
>>>>   3.10.01  6.45         23.331     61 a---  uconv.dll
>>> That's the supposedly buggy one:
>>>   7-17-02  10:31a     23331           0  uconv.dll
>>> 10-03-01   6:45a     23331          61  uconv.dll.save
>>>
>>> Unfortunately, the bldlevel info (and the size, I see) are the same.
>>>
>>> I'm not sure if it is available elsewhere, but you can try this:
>>> http://www.os2bbs.com/os2news/uconv.zip
>>
>> Thanks a lot, Steve :-)
>> The new DLL seems to have cured my problem.
>> Gecko/2008121717 Firefox/3.0.5 is now running more than 30hrs and for
>> the very first time without crashing  :-)
> 
> Great.
> 
> Steve, the one in the ZIP package, is that now the original IBM version
> or the Innotek patched one?

The bldlevel for the uconv.dll above is 14.83 but the bldlevel for the 
one currently on my system is 14.93 so there was apparently some 
subsequent release in a fixpack.

-- 
Posted with OS/2 Warp 4.52
and Sea Monkey 1.5a
0
David
1/22/2009 12:12:29 AM
On 01/22/09 01:12, David T. Johnson wrote:

> The bldlevel for the uconv.dll above is 14.83 but the bldlevel for the
> one currently on my system is 14.93 so there was apparently some
> subsequent release in a fixpack.

Yes, that one comes with eCS, too.
-- 
Please                      Enhanced OS/2 builds of Mozilla applications
reply in                                    http://pmw-warpzilla.sf.net/
newsgroup
     Steve's Warpzilla Tips: http://www.os2bbs.com/os2news/Warpzilla.html
0
Peter
1/23/2009 9:24:27 AM
On 01/21/09 23:34, Steve Wendt wrote:
> On 1/21/2009 2:15 PM, Peter Weilbacher wrote:
>
>>>> I'm not sure if it is available elsewhere, but you can try this:
>>>> http://www.os2bbs.com/os2news/uconv.zip
>>>
>>> Gecko/2008121717 Firefox/3.0.5 is now running more than 30hrs and for
>>> the very first time without crashing :-)
>>
>> Steve, the one in the ZIP package, is that now the original IBM version
>> or the Innotek patched one?
>
> That is the one from Innotek, but I don't think they patched it (I think
> it was an unofficial IBM fix). I think the one that Ian pointed to on
> os2site.com is the same as the latest official IBM one.

That is annoying. I wanted to add something to README.txt about this but
if one cannot determine the stable version easily than I can forget
about that.
-- 
Please                      Enhanced OS/2 builds of Mozilla applications
reply in                                    http://pmw-warpzilla.sf.net/
newsgroup
     Steve's Warpzilla Tips: http://www.os2bbs.com/os2news/Warpzilla.html
0
Peter
1/23/2009 9:27:05 AM
Hi Peter,

> That is annoying. I wanted to add something to README.txt about this but
> if one cannot determine the stable version easily than I can forget
> about that.

uconv.dll Build Level 14.93 is the last stable IBM version.

I didnt realise it had been released as part of eComStation, also 
apparently it was in the last official (CP2 FP5) and unofficial OS/2 
v4.5 fixpack CP2 Fixpack6, if not also earlier FP's.

http://www.os2site.com/sw/graphics/fonts/uconv_dll.zip

I do all sorts of wierd stuff to my copy of OS/2 here, and its very 
stable, basically the same one I installed back in the early 90's, (an 
xcopy to a new HD every few years when I upgrade) with upgrades, 
fixpacks etc applied as they were released, including some unreleased 
testcase files, and I have no problems with Mozilla Seamonkey
(Offical versions or your version) and Sunbird BUT I do as a matter or 
cause install Noscript and Flashblock.

I also have eComStation loaded in a VPC under OS/2 and that doesn't have 
any problems, font or otherwise, running Firefox (your version). All 
systems use uconv.dll Build Level 14.93, also checked my server which is 
regularly up for months to years, that has the same uconv.dll BL14.93 on 
  it. No grief using VNC on that, which is font/WPS heavy.
0
Blacklisted
1/23/2009 11:53:42 AM
On 1/23/2009 1:27 AM, Peter Weilbacher wrote:

>> That is the one from Innotek, but I don't think they patched it (I think
>> it was an unofficial IBM fix). I think the one that Ian pointed to on
>> os2site.com is the same as the latest official IBM one.
> 
> That is annoying. I wanted to add something to README.txt about this but
> if one cannot determine the stable version easily than I can forget
> about that.

I suppose you could say something like "dated July 2002 or later, and/or 
with build level 14.093."  If a link to the 14.093 version were 
included, you could just mention that one.  I can put it on the 
Warpzilla Tips page, if that is desirable.
0
Steve
1/23/2009 6:05:27 PM
On Fri, 23 Jan 2009 18:05:27 UTC, Steve Wendt <spamsux@forgetit.org> 
wrote:

> I suppose you could say something like "dated July 2002 or later,

I wouldn't do that. The wizards at Mensys, seem to change dates, and 
times, when they happen to feel like it. Some of the ZIP/UNZIP 
versions also happen to pick local times, relative to UTC so you can't
tell, for sure, what it really is by date and time. Build level is 
really the only way to tell, as long as whatever it is, has the data 
in it.

-- 
From the eComStation 2.0 RC6a of Doug Bissett
dougb007 at telus dot net
(Please make the obvious changes, to e-mail me)

0
Doug
1/24/2009 3:24:37 AM
On 24.01.09 04:24, Doug Bissett wrote:
> On Fri, 23 Jan 2009 18:05:27 UTC, Steve Wendt<spamsux@forgetit.org>
> wrote:
>
>> I suppose you could say something like "dated July 2002 or later,
>
> I wouldn't do that. The wizards at Mensys, seem to change dates, and
> times, when they happen to feel like it. Some of the ZIP/UNZIP
> versions also happen to pick local times, relative to UTC so you can't
> tell, for sure, what it really is by date and time. Build level is
> really the only way to tell, as long as whatever it is, has the data
> in it.

If Mensys has recently messed with the file dates then it should still
be look a lot newer than July 2002, whatever local time zone was used.
The requirement in the README for the FF 3.0.6 release now reads
      + Unicode Conversion (UCONV.DLL), build level 14.093 (levels of
        14.083 might work, too, if the file date is July 2002 or later)
That should be better than what was there before.
-- 
Please     | Official Warpzilla Ports: http://www.mozilla.org/ports/os2/
reply in   |
newsgroup  |          Enhanced OS/2 builds: http://pmw-warpzilla.sf.net/
     Steve's Warpzilla Tips: http://www.os2bbs.com/os2news/Warpzilla.html
0
Peter
2/2/2009 8:33:34 PM
On 24.01.09 04:24, Doug Bissett wrote:

> The wizards at Mensys, seem to change dates, and times,

Usually, files of a distribution were touched. IBM did that as well.

> when they happen to feel like it.

No, distributing touched files is common usage. Usually it eases
support. But for the eCS CDs, I would also prefer to keep the original
dates. One reason is that (unfortunately) not all files have a
build level string added.

-- 
Andreas Schnellbacher
0
Andreas
2/3/2009 8:52:16 PM
On 01/13/09 19:38, Dave Yeo wrote:

> I wonder if the JIT compiler is using instruction(s) that are not
> supported by the K6-2? Seems that you are the only one still running a
> K6 and also the only one getting a SIGILL (ANSI: Illegal instruction)
> with the GCC 3.3.5 compiled versions. I get a sigsegv (ANSI:
> Segmentation fault) when compiled with GCC 3.4.6.
> Perhaps something to be added to the readme

While I added something at the top of the nightly-build-ReadMe files
since then, your guess apparently turned out to be correct. See
    https://bugzilla.mozilla.org/show_bug.cgi?id=477471#c3
and
    https://bugzilla.mozilla.org/show_bug.cgi?id=473552
As the latter was fixed on mozilla-central yesterday, the next Firefox
3.2pre nightly (once I get my new machine to a state to be able to build
one) will probably be fixed for Wolfi.
-- 
Please                      Enhanced OS/2 builds of Mozilla applications
reply in                                    http://pmw-warpzilla.sf.net/
newsgroup
     Steve's Warpzilla Tips: http://www.os2bbs.com/os2news/Warpzilla.html
0
Peter
2/19/2009 1:18:35 PM
Am 19.02.09 07.18 schrieb Peter Weilbacher:
> On 01/13/09 19:38, Dave Yeo wrote:
> 
>> I wonder if the JIT compiler is using instruction(s) that are not
>> supported by the K6-2? Seems that you are the only one still running a
>> K6 and also the only one getting a SIGILL (ANSI: Illegal instruction)
>> with the GCC 3.3.5 compiled versions. I get a sigsegv (ANSI:
>> Segmentation fault) when compiled with GCC 3.4.6.
>> Perhaps something to be added to the readme
> 
> While I added something at the top of the nightly-build-ReadMe files
> since then, your guess apparently turned out to be correct. See
>    https://bugzilla.mozilla.org/show_bug.cgi?id=477471#c3
> and
>    https://bugzilla.mozilla.org/show_bug.cgi?id=473552
> As the latter was fixed on mozilla-central yesterday, the next Firefox
> 3.2pre nightly (once I get my new machine to a state to be able to build
> one) will probably be fixed for Wolfi.

Thanks a lot for the follow-up, Peter.

Looks like also some Linux folks are still using such older �Ps :-)

This problem now being fixed surely will make my piece of vintage
silicon quite happy ;-)


Wolfi
-- 
            K6-2+/500, OS/2 Warp4 + FP16 + Kernel 14.104a
         SeaMonkey 1.1.14 (de-AT: rv:1.8.1.19) Gecko/20081212

0
Wolfi
2/20/2009 6:14:28 PM
On 13.01.09 18:29, Wolfi wrote:
> Am 13.01.09 09.51 schrieb Dave Yeo:
>> Can you change the line
>> pref("javascript.options.jit.content",      true);
>> to false. The line will be in PROGRAM_DIRECTORY\greprefs\all.js around
>> line 512 and see if it affects the sigill?
>
>
> Ooohh right, that's what Peter told me back then for SM 2.0a2+.
> I completely forgot about that part :-[
>
> And YES, :-) it makes all the difference in the world :-)
>
> Now FF 3.1+ stays open.

Wolfi, could you try the current nightly build and turn jit on again?
Firefox or SeaMonkey from
    http://ftp.mozilla.org/pub/mozilla.org/firefox/nightly/contrib/latest-mozilla-1.9.1/
    http://ftp.mozilla.org/pub/mozilla.org/seamonkey/nightly/contrib/latest-comm-central/
should both have some fixes that might also help your problem.
-- 
Please     | Official Warpzilla Ports: http://www.mozilla.org/ports/os2/
reply in   |
newsgroup  |          Enhanced OS/2 builds: http://pmw-warpzilla.sf.net/
     Steve's Warpzilla Tips: http://www.os2bbs.com/os2news/Warpzilla.html
0
Peter
4/23/2009 9:37:05 PM
Am 23.04.09 16.37 schrieb Peter Weilbacher:
> On 13.01.09 18:29, Wolfi wrote:
>> Am 13.01.09 09.51 schrieb Dave Yeo:
>>> Can you change the line
>>> pref("javascript.options.jit.content",      true);
>>> to false. The line will be in PROGRAM_DIRECTORY\greprefs\all.js around
>>> line 512 and see if it affects the sigill?
>>
>>
>> Ooohh right, that's what Peter told me back then for SM 2.0a2+.
>> I completely forgot about that part :-[
>>
>> And YES, :-) it makes all the difference in the world :-)
>>
>> Now FF 3.1+ stays open.
> 
> Wolfi, could you try the current nightly build and turn jit on again?
> Firefox or SeaMonkey from
>   
> http://ftp.mozilla.org/pub/mozilla.org/firefox/nightly/contrib/latest-mozilla-1.9.1/

I'm just trying this one, but can't get the
Firefox-3.5.B4-Langpack_de.xpi to install. Shiretoko claims that
"Deutsch (DE) Language Pack 3.5b4 could not be installed because it is
not compatible with Shiretoko 3.5b4pre."
which is kind of ridiculous.

>From its install.rdf:
        <em:id>{ec8030f7-c20a-464f-9b0e-13a3a9e97384}</em:id>
        <em:minVersion>3.5b4</em:minVersion>
        <em:maxVersion>3.5b4</em:maxVersion>

Can someone give me a pointer, how the min- or maxVersion needs to be
modified in order to satisfy Shiretoko 3.5b4pre.


Shiretoko basically seems to work now for me as well on my K6-2, but
having a tendency to crash easily, when the session manager tries to
re-open the previously active tabs (crashes in both, off- and online mode):

Killed by SIGSEGV
pid=0x01fd ppid=0x0028 tid=0x0001 slot=0x00ad pri=0x0200 mc=0x0001
N:\MOZILLA-TEST\FF3.5X\FIREFOX.EXE
XUL 0:00b07c11
cs:eip=005b:10f17c11      ss:esp=0053:00119e80      ebp=00119e8c
 ds=0053      es=0053      fs=150b      gs=0000     efl=00212246
eax=240b60d0 ebx=00119f54 ecx=0000000d edx=00000006 edi=00000000
esi=23e42dd8
Process dumping was disabled, use DUMPPROC / PROCDUMP to enable it.

Killed by SIGILL
pid=0x020b ppid=0x0028 tid=0x0001 slot=0x004e pri=0x0200 mc=0x0001
N:\MOZILLA-TEST\FF3.5X\FIREFOX.EXE
LIBC063 ffffffff:ffffffff
cs:eip=005b:23a30e4c      ss:esp=0053:0011c1c0      ebp=0011c1d8
 ds=0053      es=0053      fs=150b      gs=0000     efl=00212246
eax=00000004 ebx=00000002 ecx=0011c1f0 edx=20000004 edi=0011cb94
esi=00000004
Process dumping was disabled, use DUMPPROC / PROCDUMP to enable it.

Killed by SIGILL
pid=0x0212 ppid=0x0028 tid=0x0001 slot=0x00a1 pri=0x0200 mc=0x0001
N:\MOZILLA-TEST\FF3.5X\FIREFOX.EXE
LIBC063 ffffffff:ffffffff
cs:eip=005b:23414b45      ss:esp=0053:0011c660      ebp=0011c688
 ds=0053      es=0053      fs=150b      gs=0000     efl=00212246
eax=00000009 ebx=00000009 ecx=237ffc68 edx=00000009 edi=0011d0ec
esi=0011c6a0
Process dumping was disabled, use DUMPPROC / PROCDUMP to enable it.


> http://ftp.mozilla.org/pub/mozilla.org/seamonkey/nightly/contrib/latest-comm-central/
> 
> should both have some fixes that might also help your problem.

I still need to d/l and test SM.

Wolfi
-- 
            K6-2+/500, OS/2 Warp4 + FP16 + Kernel 14.104a
         SeaMonkey 1.1.16 (de-AT: rv:1.8.1.19) Gecko/20090412

0
Wolfi
4/25/2009 12:19:56 PM
On Apr 25, 2:19=A0pm, Wolfi <publicalfa...@yahoo.fr> wrote:
>
> I'm just trying this one, but can't get the
> Firefox-3.5.B4-Langpack_de.xpi to install. Shiretoko claims that
> "Deutsch (DE) Language Pack 3.5b4 could not be installed because it is
> not compatible with Shiretoko 3.5b4pre."
> which is kind of ridiculous.
>
> >From its install.rdf:
>
> =A0 =A0 =A0 =A0 <em:id>{ec8030f7-c20a-464f-9b0e-13a3a9e97384}</em:id>
> =A0 =A0 =A0 =A0 <em:minVersion>3.5b4</em:minVersion>
> =A0 =A0 =A0 =A0 <em:maxVersion>3.5b4</em:maxVersion>
>
> Can someone give me a pointer, how the min- or maxVersion needs to be
> modified in order to satisfy Shiretoko 3.5b4pre.
You have to set the maxVersion to 3.5b4pre
Walter
0
Walter
4/25/2009 4:10:50 PM
Am 25.04.09 11.10 schrieb Walter Meinl:
> On Apr 25, 2:19 pm, Wolfi <publicalfa...@yahoo.fr> wrote:
>> I'm just trying this one, but can't get the
>> Firefox-3.5.B4-Langpack_de.xpi to install. Shiretoko claims that
>> "Deutsch (DE) Language Pack 3.5b4 could not be installed because it is
>> not compatible with Shiretoko 3.5b4pre."
>> which is kind of ridiculous.
>>
>> >From its install.rdf:
>>
>>         <em:id>{ec8030f7-c20a-464f-9b0e-13a3a9e97384}</em:id>
>>         <em:minVersion>3.5b4</em:minVersion>
>>         <em:maxVersion>3.5b4</em:maxVersion>
>>
>> Can someone give me a pointer, how the min- or maxVersion needs to be
>> modified in order to satisfy Shiretoko 3.5b4pre.
> You have to set the maxVersion to 3.5b4pre

Unfortunately that doesn't do the trick :-(
I'm still getting the same refusing message.

Previously I had tried it with maxVersion>3.5b4*, assuming that it would
trump all that "pre" stuff, but that failed as well.

There isn't any other place that has to be modified as well, is there?

Wolfi
0
Wolfi
4/25/2009 8:11:03 PM
On Apr 25, 10:11=A0pm, Wolfi <publicalfa...@yahoo.fr> wrote:
> Am 25.04.09 11.10 schrieb Walter Meinl:
>
>
>
> > On Apr 25, 2:19 pm, Wolfi <publicalfa...@yahoo.fr> wrote:
> >> I'm just trying this one, but can't get the
> >> Firefox-3.5.B4-Langpack_de.xpi to install. Shiretoko claims that
> >> "Deutsch (DE) Language Pack 3.5b4 could not be installed because it is
> >> not compatible with Shiretoko 3.5b4pre."
> >> which is kind of ridiculous.
>
> >> >From its install.rdf:
>
> >> =A0 =A0 =A0 =A0 <em:id>{ec8030f7-c20a-464f-9b0e-13a3a9e97384}</em:id>
> >> =A0 =A0 =A0 =A0 <em:minVersion>3.5b4</em:minVersion>
> >> =A0 =A0 =A0 =A0 <em:maxVersion>3.5b4</em:maxVersion>
>
> >> Can someone give me a pointer, how the min- or maxVersion needs to be
> >> modified in order to satisfy Shiretoko 3.5b4pre.
> > You have to set the maxVersion to 3.5b4pre
>
> Unfortunately that doesn't do the trick :-(
> I'm still getting the same refusing message.
>
> Previously I had tried it with maxVersion>3.5b4*, assuming that it would
> trump all that "pre" stuff, but that failed as well.
>
> There isn't any other place that has to be modified as well, is there?
>
> Wolfi

Maybe you try to install a langpack for 3.5b4pre
ftp://ftp.mozilla.org/pub/firefox/nightly/latest-mozilla-1.9.1-l10n/firefox=
-3.5b4pre.de.langpack.xpi.
(the install.rdf sets minVersion and maxVersion to 3.5b4pre)
Walter
0
Walter
4/25/2009 10:35:36 PM
Am 25.04.09 17.35 schrieb Walter Meinl:
> On Apr 25, 10:11 pm, Wolfi <publicalfa...@yahoo.fr> wrote:
>> Am 25.04.09 11.10 schrieb Walter Meinl:
>>
>>
>>
>>> On Apr 25, 2:19 pm, Wolfi <publicalfa...@yahoo.fr> wrote:
>>>> I'm just trying this one, but can't get the
>>>> Firefox-3.5.B4-Langpack_de.xpi to install. Shiretoko claims that
>>>> "Deutsch (DE) Language Pack 3.5b4 could not be installed because it is
>>>> not compatible with Shiretoko 3.5b4pre."
>>>> which is kind of ridiculous.
>>>> >From its install.rdf:
>>>>         <em:id>{ec8030f7-c20a-464f-9b0e-13a3a9e97384}</em:id>
>>>>         <em:minVersion>3.5b4</em:minVersion>
>>>>         <em:maxVersion>3.5b4</em:maxVersion>
>>>> Can someone give me a pointer, how the min- or maxVersion needs to be
>>>> modified in order to satisfy Shiretoko 3.5b4pre.
>>> You have to set the maxVersion to 3.5b4pre
>> Unfortunately that doesn't do the trick :-(
>> I'm still getting the same refusing message.
>>
>> Previously I had tried it with maxVersion>3.5b4*, assuming that it would
>> trump all that "pre" stuff, but that failed as well.
>>
>> There isn't any other place that has to be modified as well, is there?
>>
>> Wolfi
> 
> Maybe you try to install a langpack for 3.5b4pre
> ftp://ftp.mozilla.org/pub/firefox/nightly/latest-mozilla-1.9.1-l10n/firefox-3.5b4pre.de.langpack.xpi.
> (the install.rdf sets minVersion and maxVersion to 3.5b4pre)
> Walter

Aha, looks like I took a wrong turn then ;-)
Thanks for the link. This one worked fine now.

Mmh, but I still don't get it to switch from the default English to
German, when using "-UI-locale de-DE" as parameter.
Or did something change here in the syntax?

Trying to find the matching langpack I was ending up and grabbing the
one from
ftp://ftp.mozilla.org/pub/mozilla.org/firefox/nightly/3.5b4-candidates/build1/win32/xpi/
after first looking in 3.5b-something directory, having only
3.6b-something files in it.

To be honest, I got lost since quite a little while now, in the Mozilla
jungle of code-names, currents, mozilla-centrals, tracemonkeys,
*.candidates, latest trunks and mozilla-1.x.y

There have been times, where it was easier to follow-up for the ordinary
every-once-in-a-while user doing some testing.

Wolfi
0
Wolfi
4/26/2009 12:18:26 AM
Am 25.04.09 17.35 schrieb Walter Meinl:
> On Apr 25, 10:11 pm, Wolfi <publicalfa...@yahoo.fr> wrote:
>> Am 25.04.09 11.10 schrieb Walter Meinl:
>>
>>
>>
>>> On Apr 25, 2:19 pm, Wolfi <publicalfa...@yahoo.fr> wrote:
>>>> I'm just trying this one, but can't get the
>>>> Firefox-3.5.B4-Langpack_de.xpi to install. Shiretoko claims that
>>>> "Deutsch (DE) Language Pack 3.5b4 could not be installed because it is
>>>> not compatible with Shiretoko 3.5b4pre."
>>>> which is kind of ridiculous.
>>>> >From its install.rdf:
>>>>         <em:id>{ec8030f7-c20a-464f-9b0e-13a3a9e97384}</em:id>
>>>>         <em:minVersion>3.5b4</em:minVersion>
>>>>         <em:maxVersion>3.5b4</em:maxVersion>
>>>> Can someone give me a pointer, how the min- or maxVersion needs to be
>>>> modified in order to satisfy Shiretoko 3.5b4pre.
>>> You have to set the maxVersion to 3.5b4pre
>> Unfortunately that doesn't do the trick :-(
>> I'm still getting the same refusing message.
>>
>> Previously I had tried it with maxVersion>3.5b4*, assuming that it would
>> trump all that "pre" stuff, but that failed as well.
>>
>> There isn't any other place that has to be modified as well, is there?
>>
>> Wolfi
> 
> Maybe you try to install a langpack for 3.5b4pre
> ftp://ftp.mozilla.org/pub/firefox/nightly/latest-mozilla-1.9.1-l10n/firefox-3.5b4pre.de.langpack.xpi.
> (the install.rdf sets minVersion and maxVersion to 3.5b4pre)
> Walter

Aha, looks like I took a wrong turn then ;-)
Thanks for the link. This one worked fine now.

Mmh, but I still don't get it to switch from the default English to
German, when using "-UI-Locale de-DE" as parameter.
Or did something change here in the syntax?

Oooops, found it. There was an invalid hyphen creeping in. With the
correct "-UILocale de-DE" it's now working.

Trying to find the matching langpack I was ending up and grabbing the
one from
ftp://ftp.mozilla.org/pub/mozilla.org/firefox/nightly/3.5b4-candidates/build1/win32/xpi/
after first looking in 3.5b-something directory, having only
3.6b-something files in it.

To be honest, I got lost since quite a little while now, in the Mozilla
jungle of code-names, currents, mozilla-centrals, tracemonkeys,
*.candidates, latest trunks and mozilla-1.x.y

There have been times, where it was easier to follow-up for the ordinary
every-once-in-a-while user doing some testing.

Wolfi
0
Wolfi
4/26/2009 12:26:30 AM
Reply:

Similar Artilces:

enigmail-1.0.1 for TB-3.0.x and enigmail-1.1.2 for TB-3.1.x available
Hi, enigmail-1.0.1 (TB-3.0.x) and enigmail-1.1.2 (TB-3.1.x) are available at http://enigmail.mozdev.org. Note for Seamonkey-2.0.x you can download and install enigmail-1.0.1. Have fun, Walter ...

FF 3.0.3 & 3.0.5 sluggish
Hi all, On OpenSUSE 10.3 with FF 3.0.3 and 3.0.5, as well as Windows Vista Business 32 bit with FF 3.0.3 and 3.0.5, FF sort of reacts only after 2 - 3 seconds when scrolling or entering fields. The delay increases proportional to the size of the page and amount of data on it. We also noticed that FF 3.0.3 on Vista Business 32 bit completely stops when winrar decompresses something, but all other programmes run normally. Any suggestions? TIA :-) Al ...

Delayed action in FF 3.0.3 & 3.0.5 Linux & Windows
Hi all, New here. I scanned through the mails I got the last month for this problem. I have heard of it somewhere though. I also used www.clusty.com to search for "delay" or "freeze" and "firefox 3", but couldn't find what describes the problem. Problem: Systems - On OpenSUSE 10.3 with FF 3.0.3 and 3.0.5, as well as Windows Vista Business 32 bit with FF 3.0.3 and 3.0.5. Description - When entering something in fields or just scrolling down and up, FF sort of reacts only after 2 - 3 seconds. The delay increases proportional to the size of the pag...

SM 2.0/2.1 vs FF 3.0/3.1?
I'm just wondering what the SM release plans are with FF considering a "quick" follow-on to FF 3.0, currently being refered to as FF 3.1. Will SM 2.0 coincide with such a FF 3.1? (~ DEC 2008 ...) Or will SM 2.0 ship after FF 3.0, followed by SM 2.1 after FF 3.1? Thank you, Eddie Maddox Eddie-MacG3 wrote: > I'm just wondering what the SM release plans are > with FF considering a "quick" follow-on to FF 3.0, > currently being refered to as FF 3.1. > > Will SM 2.0 coincide with such a FF 3.1? > (~ DEC 2008 ...) > > Or w...

Crash in FF 3.0.5 (also 3.0.6 and 3.1b1)
Start link "wichtiger Service-Hinweis..." from http://www.heise.de/preisvergleich/a352430.html crashes Firefox (javascript enabled). Turn off jacascript and the direct link works. Right mouse on link and "Open link in new tab" works, also copy link and open in new tab. Any hints? Thanks! Walter Walter Lonsdorf wrote: > Start link "wichtiger Service-Hinweis..." from > http://www.heise.de/preisvergleich/a352430.html > crashes Firefox (javascript enabled). > Turn off jacascript and the direct link works. > Right mouse on link and &...

FF-3.5.8 SM-2.0.3 TB-3.0.2 uploaded to netlabs ftp server
Hi, I've just finished building Firefox-3.5.8, Seamonkey-2.0.3 and Thunderbird-3.0.2 (with lightning). These builds can be downloaded from ftp.//ftp.netlabs.org/incoming/mozilla The sources were pulled from http://hg.mozilla.org/releases/mozilla-1.9.1/ (tag FIREFOX_3_5_8_RELEASE) and http://hg.mozilla.org/releases/comm-1.9.1/ (tags SEAMONKEY_2_0_3_RELEASE and THUNDERBIRD_3_0_2_RELEASE), respectively. I used gcc-3.3.5 for all of these builds. The usual fixes (bug451278 and bug453705) for building with gcc-3.3.5 were applied. All builds contain now the fix for the missing handlers...

Security Advisory for Bugzilla 3.0.3, 3.1.3, 2.22.3, and 2.20.5
--Sig_/7+QS=YT68me2o8pI2lL1LPd Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable Summary =3D=3D=3D=3D=3D=3D=3D Bugzilla is a Web-based bug-tracking system, used by a large number of software projects. This advisory covers three security issues that have recently been fixed in the Bugzilla code: * Users without the "canconfirm" privilege could enter a bug as NEW or ASSIGNED by using the XML-RPC interface. * When viewing several bugs at once, there was a Cross-Site Scripting hole. * The inbound email interface allowed you t...

FF 3.5.3 crashes my X server
There's been a lot of this happening over the past few years. I've got an old fedora 6 system: X Window System Version 7.1.1 xorg-x11-server-Xorg-1.1.1-47.10.fc6 NV: driver for NVIDIA chipsets: RIVA 128, RIVA TNT, RIVA TNT2, Up to 3.0.14 all is working fine. At 3.5.3 the www.theage.com.au page reliably crashes the X server. It happens with no plugins and no add-ons enabled. Backtrace: 0: /usr/bin/Xorg(xf86SigHandler+0x81) [0x80bc151] 1: [0x268420] 2: /usr/lib/xorg/modules/libfb.so(fbCopyAreammx+0x18f) [0x18a34f] 3: /usr/lib/xorg/modules/libfb.so(fbCompositeCopyAr...

possible ipluginw problems with firefox-3.5.1-media, solution copy over ipluginw from no-media or ff-3.0.x
Only now I got aware by a posting in de.os2.org that there are problems with java with the new firefox-3.5.1-media-gcc44 build. As I've written in the readme ipluginw.dll built with gcc-4.4.0 doesn't work at all, therefore I recompiled it with gcc-3.4.6. This is working perfectly on my machine but won't do on others. The reason is that rebuilding of ipluginw.dll with gcc-3.4.6 pulls in a dependency on gcc346.dll, which is present on my machine but certainly not on most machines of users. As a workaround you could either download gcc346.dll from Paul Smedleys Homepage or use...

FF-4.01/3.6.17/3.5.19; SM-2.0.14; TB-3.1.10 released
Hi, security and stability releases are available Firefox-4.0.1 ftp://ftp.mozilla.org/pub/firefox/releases/4.0.1/contrib/firefox-4.0.1.en-US.os2.zip No new OS/2 fixes (but some of the patches for 4.0 are now officially included in the sources) Firefox-3.6.17 ftp://ftp.mozilla.org/pub/firefox/releases/3.6.17/contrib/firefox-3.6.17.en-US.os2.zip Firefox-3.5.19 ftp://ftp.mozilla.org/pub/firefox/releases/3.5.19/contrib/firefox-3.5.19.en-US.os2.zip SeaMonkey-2.0.14 ftp://ftp.mozilla.org/pub/seamonkey/releases/2.0.14/contrib/seamonkey-2.0.14.en-US.os2.zip Thunderbird-3.1.10 ftp://ftp.mo...

From 2.1.2 to 3.1 or 2.1.2 to 3.0.13 to 3.1
I'm checking back in after being away for a couple of months and I missed the 3.1 release. I can't seem to find any info on what's new in 3.1 and also what the upgrade path should be. So, is the recommended path for upgrading from v 2.1.2 to go directly to 3.1 or do I need to go to 3.0.13 first? Are there any big issues in 3.1 that make it a "gotta have it" version?I did a couple of 2.1.2 to 3.0.12 upgrades but haven't upgraded to or tested 3.1 yet. I have one 2.1.2 site that is commercial and I need it to be down as little time as possible. It also has a lot of third party modules...

FF 3.0.1 has same problem as FF 3.0
Name: Ken Workman Product: Firefox Summary: FF 3.0.1 has same problem as FF 3.0 Comments: There are many things I like about FF 3.0.1. But there is one vexing problem that was introduced with version 3.0 - duplicate entries in the Desktop Menu Bar. Currently my Menu Bar reads as follows: Firefox File Edit View History Bookmarks Tools Window Help File Edit View History Bookmarks Tools Window Help. Disabling all Add-ons does not help. Nor does starting the the Safe Mode or creating new Login Account. Config: 1.25GHz G4 Power Mac (MDD), OS 10.4.11 Browser Details: Mozill...

[ANN] Release of Bugzilla 3.0.1, 3.1.1, 2.22.3, and 2.20.5
--Sig_nEH4Ypdt2vWEVOYyO6=XaoK Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: quoted-printable Today the Bugzilla project has four releases! 3.0.1 is the first bug-fix release for the 3.0 series. 3.0 was very stable, but 3.0.1 adds a lot of little polish fixes that greatly improve the experience of using Bugzilla. 3.1.1 is our first development release toward Bugzilla 3.2. It has a huge number of new features, but is EXTREMELY UNSTABLE. It has not been tested, and should not be used in a production environment. It may fail in critical ways, or destro...

Testopia 1.3 (Bugzilla 3.0) and Testopia 1.2.2 (Bugzilla 2.22.x) Released
Testopia 1.3 has been released.=20 This Testopia is compatible with Bugzilla 3.0 and mod_perl and provides Postgres support. For those of you still running Bugzilla 2.22.x, I am also releasing 1.2.2 which is a (major) bug fix release. They are available from the project home page at http://www.mozilla.org/projects/testopia Thanks for all the feedback. Keep it coming. Greg Hendricks ...

Web resources about - FF 3.0.5 is OK; FF 3.1.x & 3.2.x keep crashing - mozilla.dev.ports.os2

Wedding crashing - Wikipedia, the free encyclopedia
... the Busted song, see Crashed the Wedding . For something that is offered at the event, such as free catered food or alcoholic beverages . Crashing ...

Search Twitter - twitter crashing
Log in Sign up Search Refresh Nine News Australia @ 9NewsAUS 33m . @U2 ’s Bono will require surgery after crashing his bicycle in New York City’s ...

Driver casually exits truck through windshield after crashing - YouTube
Like a Boss ! Авария на южной объездной г.Владимир Чувак вышел через лобовуху

Trail bike rider dies after crashing into tree
A 24-year-old trail bike rider has died after he crashed into a tree south-west of Sydney.

Satellite Pictures Of The Empty Chinese Cities Where Home Prices Are Crashing
The long-predicted crash has arrived with a vengeance in China’s original ghost city. Click here to see satellite pictures > Home prices have ...

How the deliberate crashing of Germanwings Flight 4U9525 unfolded
As they climbed towards their cruising altitude of 38,000 feet, the conversation between Captain Patrick Sonderheimer and his co-pilot Andreas ...

A water skier is critical after crashing into trees at 90km/h on Hawkesbury River
A WATER skier is in a critical condition after crashing into trees at 90km along the banks of the Hawkesbury River this morning.

Man dies after crashing into roadworks
A man has been killed after crashing his van into roadworks equipment in central Queensland.

Apple working on resolving crashing app issue
Apple working on resolving crashing app issue

Google disables support for third-party keyboards in Chrome due to ‘crashing issues’
... Apps and Google Chrome at 9to5Mac . What do you think? Discuss "Google disables support for third-party keyboards in Chrome due to ‘crashing ...

Resources last updated: 12/9/2015 1:19:56 AM