Discussion:
SM crashes @ Hobbes
(too old to reply)
A.D. Fundum
2011-10-11 22:08:26 UTC
Permalink
While downloading all mentioned files using ...

http://hobbes.nmsu.edu/h-browse.php?dir=/pub/new&sort=date

... I just experienced my fifth SM crash in a row, right after a system
reboot because Rexx was producing SYS3175s and a WPS reset failed to
complete.


--
A.D. Fundum
2011-10-12 10:38:18 UTC
Permalink
Post by A.D. Fundum
While downloading all mentioned files using ...
http://hobbes.nmsu.edu/h-browse.php?dir=/pub/new&sort=date
... I just experienced my fifth SM crash in a row
Same today, the second time I used the "Show"-button of ...

http://www.lme.com/aluminium_graphs.asp

... it asked how to open the ASP file and crashed when I selected
"Cancel".

AFAICT this issue is "new", I've used SM frequently to download e.g.
Hobbes files and never had such an issue.


--
Steve Wendt
2011-10-13 03:32:03 UTC
Permalink
Post by A.D. Fundum
Post by A.D. Fundum
While downloading all mentioned files using ...
http://hobbes.nmsu.edu/h-browse.php?dir=/pub/new&sort=date
... I just experienced my fifth SM crash in a row
Same today, the second time I used the "Show"-button of ...
http://www.lme.com/aluminium_graphs.asp
... it asked how to open the ASP file and crashed when I selected
"Cancel".
Not sure what you expect, posting here instead of in the m.d.p.os2
newsgroup, and without the least bit of helpful information, like the
SeaMonkey version, whether there was an exceptq log, etc.
A.D. Fundum
2011-10-13 21:39:34 UTC
Permalink
Post by Steve Wendt
Post by A.D. Fundum
Post by A.D. Fundum
While downloading all mentioned files using ...
http://hobbes.nmsu.edu/h-browse.php?dir=/pub/new&sort=date
... I just experienced my fifth SM crash in a row
Same today, the second time I used the "Show"-button of ...
http://www.lme.com/aluminium_graphs.asp
... it asked how to open the ASP file and crashed when I
selected "Cancel".
Not sure what you expect
My combination of software shouldn't be that rare and e.g. Hobbes is a
simple website, so I'ld expect it to be a common problem or some
obscure local problem (no, no specific clues regarding that). Right
now I assume it was a local problem that occured at least twice.
Post by Steve Wendt
whether there was an exceptq log
I don't know, it just crashed (<Alt-F4>-style) without any pointer to
some file in some directory nor a POPUPLOG.OS2-entry. I just checked a
few SM-related directories, including the unexpected (I'm using OS/2,
not /etc/bin/-*ix! :-)) ones on the boot drive, and there's no new
file.


--
Dave Yeo
2011-10-13 06:06:14 UTC
Permalink
Post by A.D. Fundum
Post by A.D. Fundum
While downloading all mentioned files using ...
http://hobbes.nmsu.edu/h-browse.php?dir=/pub/new&sort=date
... I just experienced my fifth SM crash in a row
Same today, the second time I used the "Show"-button of ...
http://www.lme.com/aluminium_graphs.asp
... it asked how to open the ASP file and crashed when I selected
"Cancel".
AFAICT this issue is "new", I've used SM frequently to download e.g.
Hobbes files and never had such an issue.
Have you tried a different/new profile? Starting in safe-mode? Is there
a trap report in your SeaMonkey directory?
Dave
A.D. Fundum
2011-10-13 21:56:45 UTC
Permalink
Post by Dave Yeo
Have you tried a different/new profile?
No(t yet), I'll try it again as soon as it occurs again. FWIW, I've
used Hobbes (same page, different new files :-o) frequently with this
setup.
Post by Dave Yeo
Starting in safe-mode?
I'll figure that out too.
Post by Dave Yeo
Is there a trap report in your SeaMonkey directory?
Not in the same directory as seamonkey.exe.

Hobbes: in the end I was able to download all 5 new files and to view
1 detailed description. It crashed during a download, right after a
download, when I pressed the Back-button, and so on. Right after an
unrelated reboot (ORexx' SYS3175, WPS IPL failed to complete, fiull
reboot). The next time also was after a (re-) boot too, albeit I used
one or two compilers (ICC, IPFC) before I visited the website.

I'll assume it was some local problem, for one because I haven't seen
any reaction recognizing the situation. Please don't hasitate to let
me know if I have to try or check anything anyway.


--
Dave Yeo
2011-10-13 23:04:25 UTC
Permalink
Post by A.D. Fundum
Post by Dave Yeo
Have you tried a different/new profile?
No(t yet), I'll try it again as soon as it occurs again. FWIW, I've
used Hobbes (same page, different new files :-o) frequently with this
setup.
Always a good idea to try a new profile to eliminate profile corruption
or an extension problem.
Post by A.D. Fundum
Post by Dave Yeo
Starting in safe-mode?
I'll figure that out too.
SeaMonkey.exe -safe-mode or Click help --> restart with add-ons disabled.
Post by A.D. Fundum
Post by Dave Yeo
Is there a trap report in your SeaMonkey directory?
Not in the same directory as seamonkey.exe.
Search Hobbes for exceptq and install it.
[...]
Post by A.D. Fundum
I'll assume it was some local problem, for one because I haven't seen
any reaction recognizing the situation. Please don't hasitate to let
me know if I have to try or check anything anyway.
Yea, I don't recall anyone having a problem with Hobbes. Could be a
profile corruption problem which is why it is a good habit to make backups.
Dave
A.D. Fundum
2011-10-16 14:15:36 UTC
Permalink
Post by Dave Yeo
Post by A.D. Fundum
I'll assume it was some local problem, for one because I haven't
seen any reaction recognizing the situation.
Yea, I don't recall anyone having a problem with Hobbes. Could be
a profile corruption problem which is why it is a good habit to
make
Post by Dave Yeo
backups.
Thanks for the pointers. Forget Hobbes. The other day I had the same
problem with "about:blank" (loaded by the first set webpage) more than
once, requiring a reboot to get that working again.


--
A.D. Fundum
2011-10-16 23:54:00 UTC
Permalink
Could be a profile corruption problem which is why it is a
good habit to make backups.
FTR, and certainly not a RFC as such: having backups of everything
required to install SM (one of the few apps storing important parts on
a bootdrive by default) and plugins again, it's a minor pity that
"Managing stored password" doesn't allow adding entries.

Real example: a website which no longer supports storing the
username/password, or your account has expired. The list of stored
passwords then still can be used to find the right password. Another
minor PITA is that you'll have to log in to all websites. Some of
those websites may not be available for a while, it may show a few
annoying messages that you haven't changed your password last week,
your password doesn't match all 6 reqiurements, and so on.


--
Peter Brown
2011-10-17 00:13:37 UTC
Permalink
Hi
Post by A.D. Fundum
Could be a profile corruption problem which is why it is a
good habit to make backups.
FTR, and certainly not a RFC as such: having backups of everything
required to install SM (one of the few apps storing important parts on
a bootdrive by default) and plugins again, it's a minor pity that
"Managing stored password" doesn't allow adding entries.
Real example: a website which no longer supports storing the
username/password, or your account has expired. The list of stored
passwords then still can be used to find the right password. Another
minor PITA is that you'll have to log in to all websites. Some of
those websites may not be available for a while, it may show a few
annoying messages that you haven't changed your password last week,
your password doesn't match all 6 reqiurements, and so on.
--
You might want to have a look at the Saved Passwords Editor addon
https://addons.mozilla.org/en-US/seamonkey/addon/saved-password-editor/

Regards

Pete
A.D. Fundum
2011-10-20 10:15:23 UTC
Permalink
Post by Peter Brown
You might want to have a look at the Saved Passwords Editor
addon
https://addons.mozilla.org/en-US/seamonkey/addon/saved-password-editor
/

Excellent, I just added the XPI file to my collection!

It's not only used as a simple password database. It's also a way to
avoid letting some emailing Database Marketeer know that you're still
alive and interested, just because you visited their local newspaper's
website once a year to restore the password entry.


--
A.D. Fundum
2011-10-20 11:35:17 UTC
Permalink
A few minutes ago, I had the same problem while downloading the second
listed new file (LibDVDRead*) at Hobbes:

http://hobbes.nmsu.edu/h-browse.php?dir=/pub/new&sort=date.
Post by Dave Yeo
Have you tried a different/new profile?
Yes, earlier I deleted the old profile and created a new one.
Post by Dave Yeo
Starting in safe-mode?
I just did (without a reboot between attempts), same problem while
downloading the second listed new file (skipping #1, so the second one
then is SearchPlus*)
Post by Dave Yeo
Is there a trap report in your SeaMonkey directory?
No sign of any ExceptQ-activity, i.e. no sound nor file. It still
crashes as if <Alt-F4> was pressed.

So far the only consistency seems to be the second operation. An
earlier crashing about:blank was the second page loaded. It fails
while I download the second file at Hobbes. And a reboot may help,
about:blank loaded by my first visited webpage doesn't always crash.
Now this may matter indeed (eCS 1.2, PIII CPU):

Mozilla/5.0 (OS/2; Warp 4.5; rv:6.0) Gecko/20110821 Firefox/6.0
SeaMonkey/2.3


--
A.D. Fundum
2011-10-20 18:53:48 UTC
Permalink
Post by A.D. Fundum
Post by Dave Yeo
Starting in safe-mode?
I just did (without a reboot between attempts), same problem
while downloading the second listed new file (skipping #1,
so the second one then is SearchPlus*)
BTW, I used "while" intentionally. This crash happened during an
ADSL2-download of SearchPlus, which resulted in a SearchPlus*.part.1
file of about 1.65 MB (I was able to download it later, this file
itself certainly doesn't cause a crash as such).


--
A.D. Fundum
2011-10-24 10:17:03 UTC
Permalink
FTR: if the crashing condition applies (it has ctrashed, and typically
visiting a second page keeps causing another crash), restarting the
WPS also seems to works. And yesterday it crashed while downloading
(about 24 MB of) a (29 MB) file using FTP.


--
Steve Wendt
2011-10-25 02:45:46 UTC
Permalink
Post by A.D. Fundum
FTR: if the crashing condition applies (it has ctrashed, and typically
visiting a second page keeps causing another crash), restarting the
WPS also seems to works. And yesterday it crashed while downloading
(about 24 MB of) a (29 MB) file using FTP.
If you have RWS, have you tried disabling that?
Steve Wendt
2011-10-21 02:03:20 UTC
Permalink
Post by A.D. Fundum
Mozilla/5.0 (OS/2; Warp 4.5; rv:6.0) Gecko/20110821 Firefox/6.0
SeaMonkey/2.3
Any particular reason you don't use 2.3.3? Not that it should have any
effect on your problems...
A.D. Fundum
2011-10-21 11:10:46 UTC
Permalink
Post by Steve Wendt
Post by A.D. Fundum
Mozilla/5.0 (OS/2; Warp 4.5; rv:6.0) Gecko/20110821
Firefox/6.0 SeaMonkey/2.3
Any particular reason you don't use 2.3.3? Not that it should have
any effect on your problems...
No technical reason to not use 2.3.3.

If needed (and destroying reproducability doesn't matter) both
upgrading and removing all data files (Where, especially on a boot
drive? Both eCS and SeaMonkey are installed using defaults) is an
option, for one my setups are fully aimed at reinstalling instead of
restoring. Possible plugins e.a. in use (but not in safe mode): GBM,
SeaMonkey Language Pack (NL), Flash 7, and since a few days the
password manager.

My starting page is a web traffic measuring site, using cookies and
one of 'em 1x1 images on selected websites. When all cookies are
present, the first website loads "about:blank". BTW, I wasn't trying
to be facetious when I mentioned that it crashed right after
displaying the "about:blank" page. It really was that page. The page
(or downloaded file) itself isn't the problem, I think I now can
reload "about:blank" as often as I want. But as soon as it crashes,
the crash is reproducable.


--
A.D. Fundum
2012-01-14 01:31:43 UTC
Permalink
Post by A.D. Fundum
While downloading all mentioned files using ...
http://hobbes.nmsu.edu/h-browse.php?dir=/pub/new&sort=date
... I just experienced my fifth SM crash in a row, right after a
system reboot
Same on another system, but now first SM reported a memory related
error or warning. I was downloading the first file, the error/warning
occured when I was just starting to download the second file.

An about:memory reported about 29.7 MB in use, the system's RAM is 1
GB. No other apps running, only CMD.EXE to MD the download directory.

It happens all of the time, e.g. a Hobbes download only works
one-by-use. Sourceforge (why are we using that?!) downloads always
cause a crash. Plug-ins in use:

flash7a63RW.zip
npgbm145_os2.xpi
saved_password_editor-2_2_5-sm+tb+fx.xpi
seamonkey-2.5.nl.langpack.xpi
video_downloadhelper-4_9_8-fx+sm.xpi

No exceptq report, no POPUPLOG.OS2, but I hope the SeaMonkey memory
error/warning is a clue. The size of the first downloaded file doesn't
seem to matter.


--
A.D. Fundum
2012-01-14 02:00:47 UTC
Permalink
but now first SM reported a memory related error or warning.
That happened once. The usual crash happens when clicking on the
"Name" of the second or (occasionally) third file to download, even in
safe-mode (tested previously, deleting a profile too).

I cannot reproduce the SM error/warning message, albeit I may
recognize it. Perhaps it thinks it's running out of memory after a
file is downloaded, while all about:sessionrestore does is loading a
small Hobbes page and the first downloaded file isn't "huge" at all.

HTH.


--
Dave Yeo
2012-01-14 05:18:13 UTC
Permalink
Post by A.D. Fundum
Post by A.D. Fundum
While downloading all mentioned files using ...
http://hobbes.nmsu.edu/h-browse.php?dir=/pub/new&sort=date
... I just experienced my fifth SM crash in a row, right after a
system reboot
Same on another system, but now first SM reported a memory related
error or warning. I was downloading the first file, the error/warning
occured when I was just starting to download the second file.
Anabout:memory reported about 29.7 MB in use, the system's RAM is 1
GB. No other apps running, only CMD.EXE to MD the download directory.
It happens all of the time, e.g. a Hobbes download only works
one-by-use. Sourceforge (why are we using that?!) downloads always
Have you tried SET MOZ_NO_RWS=1 in your environment?
Dave
A.D. Fundum
2012-01-14 09:22:48 UTC
Permalink
Post by Dave Yeo
Have you tried SET MOZ_NO_RWS=1 in your environment?
Not yet. I added it to CONFIG,SYS, rebooted, went to the Hobbes page,
and downloaded the first file (> 20 MB). Unfortunately, while still
downloading the first file, starting the download of the second file
caused the crash.

14-01-12 10:16 0 0 a--- sdlmame0144_os2.zip
14-01-12 10:16 2.319.500 0 a--- sdlmame0144_os2.zip.part

So , as usual, it crashed after clicking on the next file, but before
it was saved (sdlmame now is the first file on the page).


--
A.D. Fundum
2012-01-14 10:15:56 UTC
Permalink
Post by Dave Yeo
Post by A.D. Fundum
http://hobbes.nmsu.edu/h-browse.php?dir=/pub/new&sort=date
Have you tried SET MOZ_NO_RWS=1 in your environment?
FTR: default eCS 1.2 install. SM 2.5, but I've seen the problem with
an earlier version. Default SM install (BTW, downloading more than 2
files with FF 9.0.1 for Windows works on the same machine).

Applied changes, FTTOMH: all add-ons I've mentioned earlier, verified
with about:plugins. And about:config, dom.max_script_run_time
increased (in this case to 20). I use a different "start page"
(http://www.webmeter.nl), but that's not important because an
about:sessionrestore after a crash skips that. I add about 40 entries
with the "password manager". Preferences, under Browser: "behaviour of
links", links that would open in a new window -> open in a new window.
Advanced, buffer: set to at least 96 MB (albeit the first crash with
the memory error/warning happened when SM was managing it, and the
size was the greyed-out default of 1024). Exceptq 7.1 installed.

The number of entries in the download manager is close to 0, I tend to
delete the entries after the downloads are finished. More than enough
disk space free on all drives, more than enough RAM free (with both a
desktop and a notebook machine).

Already tried earlier: deleting the profile, safe mode, upgrading SM.
IIRC it didn't happen with SM1.1.18, which I was using for a long time
before upgrading to SM 2.3.x.

I haven't tried it now, Hobbes is small and well-known, but ISTR
downloading sourceforge-files with their commercial delay never works.
As soon as the download would start (0 seconds left), it crashes too
(the other day I tried to download some Corel viewer tar.gz file of
less than 200 KB).


--
Steve Wendt
2012-01-14 19:09:44 UTC
Permalink
Post by A.D. Fundum
Exceptq 7.1 installed.
I haven't tried it now, Hobbes is small and well-known, but ISTR
downloading sourceforge-files with their commercial delay never works.
Do your Exceptq trap files look like the one here?
https://bugzilla.mozilla.org/show_bug.cgi?id=686186
A.D. Fundum
2012-01-15 00:07:13 UTC
Permalink
Post by Steve Wendt
Post by A.D. Fundum
Exceptq 7.1 installed.
I haven't tried it now, Hobbes is small and well-known, but
ISTR downloading sourceforge-files with their commercial
delay never works.
Do your Exceptq trap files look like the one here?
https://bugzilla.mozilla.org/show_bug.cgi?id=686186
This error, which is most common (at least as common as downloading
files), does not produce a TRP file. I've reported all (other) errors,
but the crashes in question all are Alt-F4-style crashes. The only new
message was the SM error/warning message, which I've seen just once
(in my life, not once per session).

Nevertheless I've looked at this bug and sourceforge, but mine is
related to downloading (files, at least). If I downloaded a file
earloer, and before the Save As-dialog of SM is shown for the next
file, it crashes without TRP (and/or Exceptq's sound) or POPUPLOG.OS2
entry. That's alo why I mentioned Hobbes, a simple functional website
with e.g. obvious font-related issues. In my case (on any machine so
far) the error/warning message is the only possible clue. I didn't
save it, because I assumed the error may have to do with the age of
the previous system.


--
A.D. Fundum
2012-01-19 03:37:46 UTC
Permalink
Post by A.D. Fundum
Post by A.D. Fundum
... I just experienced my fifth SM crash in a row, right after
a system reboot
Same on another system, but now first SM reported a memory
related error or warning.
In Dutch, after downloading the second file of 9.6 MB (sometimes
possible, the first file was 0.4 MB):

Downloadfout (title)

Er is niet genoeg geheugen om de door u verzochte handeling uit te
voeren.

Sluit enkele toepassingen en probeer het opnieuw.


Roughly translated:


Download error

There isn't enough memory to complete the requested operation.

Close some applications and try again.


Open windows: the Hobbes new files page (first file: Q*, second file
O*). The download manager (I already deleted the Q*.ZIP file line),
this error message.

The download manager window list this file, status "fail". If I delete
this entry and close the download manager, downloading the file again
causes the expected crash.

Again, there's no reason to trigger this memory-related warning. I
only visited webmeter.nl before I directly visited Hobbes...

HT, the second time I saw this error message, H.


--
Dave Yeo
2012-01-19 06:08:21 UTC
Permalink
Post by A.D. Fundum
Post by A.D. Fundum
Post by A.D. Fundum
... I just experienced my fifth SM crash in a row, right after
a system reboot
Same on another system, but now first SM reported a memory
related error or warning.
In Dutch, after downloading the second file of 9.6 MB (sometimes
Downloadfout (title)
Er is niet genoeg geheugen om de door u verzochte handeling uit te
voeren.
Sluit enkele toepassingen en probeer het opnieuw.
Download error
There isn't enough memory to complete the requested operation.
Close some applications and try again.
Open windows: the Hobbes new files page (first file: Q*, second file
O*). The download manager (I already deleted the Q*.ZIP file line),
this error message.
The download manager window list this file, status "fail". If I delete
this entry and close the download manager, downloading the file again
causes the expected crash.
Again, there's no reason to trigger this memory-related warning. I
only visited webmeter.nl before I directly visited Hobbes...
HT, the second time I saw this error message, H.
--
It's a weird error I haven't heard of anyone else getting so the
question is, what is different about your setup from everyone elses?
Dave
A.D. Fundum
2012-01-19 18:52:43 UTC
Permalink
Post by Dave Yeo
It's a weird error I haven't heard of anyone else getting so
the question is, what is different about your setup from
everyone elses?
Yesterday I downloaded about 14 Youtube videos (~1 GB) in a row using
a plugin, no problem. AFAICT it only happens when clicking on a file
to download (Hobbes), or when a download starts automatically. It's
reproducable on several different machines. I already mentioned what I
change after I install SM on an eCS 1.2 machine:

Plugins and install package, albeit it also failed in SM's safe mode:

seamonkey-2.5.en-US.os2.zip
seamonkey-2.5.nl.langpack.xpi
flash7a63RW.zip
npgbm145_os2.xpi
saved_password_editor-2_2_5-sm+tb+fx.xpi
video_downloadhelper-4_9_8-fx+sm.xpi

The install of plugins is verified using about:plugins. If needed, the
language "Nederlands" (Dutch) is selected to be used for SM's menus
(and the first choice for websites).

Edit -> Preferences -> Buffer -> 96 MB or more. I don't think this is
important, a first crash may happen if SM is still managing it and/or
the buffer was deleted recently.
Edit -> Preferences -> Browser -> Behaviour of links (sp?) -> Open
links that would open in a new window ... in a new window.

about:config: dom.max_script_run_time increased to 15-40

Saved passwords: about 40, most of those are added manually

Start page: http://www.webmeter.nl (uses cookies to measure website
visits), which loads about:blank after logging in

Exceptq installed on all machines (it works, I've send a few SM TRP
files).

Other than that: install default, so MOZ_PLUG is on the HPFS
bootpartition (< 2 GB). SM is installed on another HPFS partition (> 6
GB).

The files to be downloaded are saved to disk, there no other app
involved (in my case WarpZIP, most Hobbes files are *.ZIP).

One machine is an old IBM 300PL tower pc (Pentium III 500 MHZ, 768 MB
of RAM, LAN, ADSL2), another machine is a more recently re-installed
IBM ThinkPad T23 (Pentium III 1 GHz, 1 GB of RAM, MiniPCI WLAN). No
sign of e.g. swapping, the disks aren't full, no excessive number of
entries in the download manager, no other apps running. I think it
occurs since SM 2, but I skipped a few versions since 1.1.18. I assume
it will fail with any other eCS machine I'll grab, which all have the
same setup.

Typically the crash happens ASAP (to an user), i.e. as soon as you
press a link to download a Hobbes file. No save as-dialog. I've seen
the error message after the download twice now. Like yesterday, when
9.6 MB of 9.6 MB was downloaded and the status of this download was
"failed". I don't use it for email (yet), so I don't know if clicking
on a download link in an email also causes the error.


--
Dave Yeo
2012-01-22 02:54:14 UTC
Permalink
Post by A.D. Fundum
Typically the crash happens ASAP (to an user), i.e. as soon as you
press a link to download a Hobbes file. No save as-dialog. I've seen
the error message after the download twice now. Like yesterday, when
9.6 MB of 9.6 MB was downloaded and the status of this download was
"failed". I don't use it for email (yet), so I don't know if clicking
on a download link in an email also causes the error.
Seems you're not alone,
https://bugzilla.mozilla.org/show_bug.cgi?id=683148. Unluckily I have no
idea why this happens.
Dave
A.D. Fundum
2012-01-22 03:35:52 UTC
Permalink
Post by Dave Yeo
Seems you're not alone,
https://bugzilla.mozilla.org/show_bug.cgi?id=683148
Yes, ignoring usual minor error description differences. The only real
additional clue I had may be the memory error-message (assuming that
piece of code isn't triggered randomly).
Post by Dave Yeo
Unluckily I have no idea why this happens.
FTR and FWIW, it's reproducable and it always fails. Unfortunately the
only thing I cannot reproduce on command is the memory error-message,
which I've seen just twice.

Again FTR, it's sometimes possible to download more than one file
using just SM (i.e. no plugins like videodownloadhelper, no problem
using it's download several times). It's not "at a time", which may
suggest "at the same time". It also fails with sequential download,
so it's "in a session". I don't know what the status "UNCONFIRMED"
means, but I can confirm the error with all of my machines used for
multiple downloads. ISTR it also fails with FTP connections.

O, possible RFCs: maybe save EAs ASAP. If a download crashes, the URL
can be found in the EAs. And add an app-specific email address to TRP
files, so a "robot" can email 'em and you don't have to figure it out
each time (I know how it works w.r.t. SM, I've send a few).


--
Dave Yeo
2012-01-22 05:10:41 UTC
Permalink
Post by A.D. Fundum
Post by Dave Yeo
Seems you're not alone,
https://bugzilla.mozilla.org/show_bug.cgi?id=683148
Yes, ignoring usual minor error description differences. The only real
additional clue I had may be the memory error-message (assuming that
piece of code isn't triggered randomly).
Post by Dave Yeo
Unluckily I have no idea why this happens.
FTR and FWIW, it's reproducable and it always fails. Unfortunately the
only thing I cannot reproduce on command is the memory error-message,
which I've seen just twice.
I can download file after file with SeaMonkey without a crash so it is
only reproducible for you.
Post by A.D. Fundum
Again FTR, it's sometimes possible to download more than one file
using just SM (i.e. no plugins like videodownloadhelper, no problem
using it's download several times). It's not "at a time", which may
suggest "at the same time". It also fails with sequential download,
so it's "in a session". I don't know what the status "UNCONFIRMED"
means, but I can confirm the error with all of my machines used for
multiple downloads. ISTR it also fails with FTP connections.
O, possible RFCs: maybe save EAs ASAP. If a download crashes, the URL
can be found in the EAs. And add an app-specific email address to TRP
files, so a "robot" can email 'em and you don't have to figure it out
each time (I know how it works w.r.t. SM, I've send a few).
Unluckily at this point we're pretty well running on inertia. Seems I'm
the only one left and I'm not a programmer.
To make things worse, we haven't been able to build a debug version for
quite a while. Seems OMF just can't support the huge DLL if built as
debug which makes debugging that much harder.
One thing that may help is removing the 16 bit TCPIP dependencies which
I meant to do for Firefox 9 but didn't. I'll try that for Firefox 10
Dave
A.D. Fundum
2012-01-22 15:46:10 UTC
Permalink
Post by Dave Yeo
One thing that may help is removing the 16 bit TCPIP dependencies
which I meant to do for Firefox 9 but didn't. I'll try that for
Firefox 10

I'll report back (I don't always follow this newsgroup) when the
problem is solved. If anything, the recorded bug may be closed then.


--
A.D. Fundum
2012-01-27 23:41:07 UTC
Permalink
Post by A.D. Fundum
Same on another system, but now first SM reported a memory
related error or warning. I was downloading the first file, the
error/warning occured when I was just starting to download
the second file.
FTR: I'll email a PrtScr to the crash@-email address.

For the rest of the world, available for a short while:

Loading Image... (125 KB)


--
Dave Yeo
2012-01-28 06:41:12 UTC
Permalink
Post by A.D. Fundum
Same on another system, but now first SM reported a memory
related error or warning. I was downloading the first file, the
error/warning occured when I was just starting to download
the second file.
Was that you trp report sent on the 23rd with the same return address as
the one with subject "PNG file attached"?
If so it is one I can't remember seeing. Can I suggest you try reducing
your VIRTUALADDRESSLIMIT to 1536 and see if that helps.
Dave
A.D. Fundum
2012-01-28 21:06:30 UTC
Permalink
Post by Dave Yeo
Was that you trp report sent on the 23rd with the same return
address
Post by Dave Yeo
as the one with subject "PNG file attached"?
Guilty, albeit it's just "a" TRP file not directly (!) related to this
crashes..
Post by Dave Yeo
try reducing your VIRTUALADDRESSLIMIT to 1536 and see if that
helps.
It didn't. Adjusted (eCS 1.2's default: 2048), rebooted, SM launched,
it failed at the third file (EXIFTO*.ZIP, most common Alt F4-style
crash, without the memory error message).


--
Dave Yeo
2012-01-31 22:56:21 UTC
Permalink
Post by A.D. Fundum
Post by Dave Yeo
Was that you trp report sent on the 23rd with the same return
address
Post by Dave Yeo
as the one with subject "PNG file attached"?
Guilty, albeit it's just "a" TRP file not directly (!) related to this
crashes..
Post by Dave Yeo
try reducing your VIRTUALADDRESSLIMIT to 1536 and see if that
helps.
It didn't. Adjusted (eCS 1.2's default: 2048), rebooted, SM launched,
it failed at the third file (EXIFTO*.ZIP, most common Alt F4-style
crash, without the memory error message).
Try it SeaMonkey 2.7 available at netlabs. Also you need to download the
xqs files and unzip them into the program directory as I screwed up
packaging them.
There is also a bit of a focus problem where you can't use the URL bar
correctly, click the desktop then the URL bar to workaround this.
Dave

ftp://ftp.netlabs.org/incoming/mozilla/seamonkey-2.7.en-US.os2.zip
ftp://ftp.netlabs.org/incoming/mozilla/seamonkey-2.7.en-US.os2.xqs.zip
A.D. Fundum
2012-02-01 04:14:40 UTC
Permalink
Post by Dave Yeo
Try it SeaMonkey 2.7 available at netlabs.
With a 2.5 crash, it's more than one file... ;-)
Post by Dave Yeo
ftp://ftp.netlabs.org/incoming/mozilla/seamonkey-2.7.en-US.os2.zip
ftp://ftp.netlabs.org/incoming/mozilla/seamonkey-2.7.en-US.os2.xqs.zip

Tried it, unfortunately with the same problem. DEL *, reboot, UNZIP
*.ZIP twice (i.e. no 2.7 langpack yet), started SM 2.7, deleted the
cache, crash when attempting to download a second file at Hobbes. So
it seems the underlying problem isn't solved "accidently".


--

Loading...