OS2 World Community Forum
OS/2, eCS & ArcaOS - Technical => Applications => Topic started by: R.M. Klippstein on February 07, 2019, 01:41:22 am
-
I just finally got around to updatng ANPM's listed necessary updates and re-booted and now Firefox 45.9 throws the error "Couldn't load XUL.DLL (OS2 error 87). Any ideas as to what is needed to make Firefox work once again?
Thanks for any responses
klipp
-
Hi klipp
I just did a Google for OS2 error 87 and the most interesting result is here https://github.com/bitwiseworks/mozilla-os2/issues/255
It may be worth checking which libcx package is installed. I seem to recall Seamonkey had problems with libcx which were resolved by installing the netlabs-exp version, 0.64-1.
I seem to recall adding netlabs-exp by the simple method of installing it from the Available list.
Regards
Pete
-
Hi Klipp, did you check popuplog.os2? Which version of Firefox?
I doubt that you need to enable the exp repository if it was previously running.
-
I do not know if this will work for your mistake. But I'm running firefox and SM in / safe-mode
I have also removed the exceptq package from the anpm.
And it went very well.
I am not using neither highmem or the mozturbo.
-
Hi Dave
The reason I mention libcx is that seamonkey did not work here with 0.6.3-1 which seems to be the current release version but works fine with 0.6.4.1 which is only available from netlabs-exp.
May be worth klipp checking...
Regards
Pete
-
Hi Pete, that's because I'd upgraded before building the most recent releases and didn't think of downgrading, so linked against the most recent that had new exports. This may also be true for my last Firefox build but not for earlier builds or Bitwise builds. In Klipp's case, his problems started after doing a yum upgrade, which shouldn't have downgraded libcx. Also why I asked which version/build.
Bitwise has been adding functionality to libcx for a while, this includes functions that were previously unsupported on OS/2, configure finds them and uses different code paths.
It'll get more interesting as Bitwise has decided to fork libc, as libcn, which I'd assume will also rapidly evolve.
Hi Roberto, the exceptq DLL is dynamically loaded, so usually not used and more and more projects use it as the functionality is now in libcx. While things will work fine without it, crashes will be much less informative.
As for safe-mode, you should be able to make everything permanent, add-ons, plugins and such all turned off.
It's usually add-ons that create problems, especially if you have lots and there's a conflict somewhere, though as Ivan found, some plugins also cause problems
-
Hey Pete,Good to hear from you, Its been quite a while since you last bailed me out! Turns out I didn't have Netlabs-exp installed, so I installed it,- but no cigar. It upgraded libcx to 0.4.1, Should I maybe try 0.3.1?
Dave, Looking at popuplog.os2 it shows Firefox at the bottom of the list, but I don't see anything that gives an indication of version/build. What should I be looking for?
This is a pretty old version of 45.9, and I have no idea if I can re-install a newer version over it or not.
Thanks for the help guys!
Oh, One last thought, ANPM can't upgrade heimdal-libs package, its got bad problems, could that affect Firefox 45.9 maybe?
-
The question is, what is the last entry or entries if at the same time, in popuplog related to Firefox?
What are the dates of the binaries? Wouldn't hurt to update to the latest. Heimdal libs shouldn't matter
-
Hi Dave
Does Firefox 45.9 use any of the libs that have been replaced by libs with the "-legacy" bit tagged onto the original name? eg libvpx and libvpx-legacy or fontconfig and fontconfig-legacy
I ask as updating libs screwed up OpenOffice for me a while back until I discovered 1 of the libs now needed had the "-legacy" bit added.
Just a thought...
Regards
Pete
-
No, not yet. Some older versions of FF depend on the legacy packages,38 needs libvpx-legacy and older needs the legacy fontconfig.
Generally the legacy packages have older versions of DLLs that aren't binary compatible as the API has changed and the DLLs themselves are usually versioned to co-exist. foo1.dll and foo2.dll with foo-legacy containing foo1.dll
-
I'm still stuck without the ability to run Firefox 45.9, I cant get it working with the latest updates to ANPM installed, and I can't downgrade libc and libcx. Is there any way of getting a copy of the .RPM version? I see now that supposedly it is only available to beta testers.
Thanks klipp
-
You shouldn't need to downgrade libc and libcx. The RPM version is available through Arca Noae's RPM repository, you should have got instructions on how to access I believe.
You still haven't posted your popuplog errors so it hard to say what the problem is.
Have you tried to start in safe mode? Incompatible plugins are a known issue, for example the Java iced tea ones.
firefox.exe --safe-mode
-
I'm still stuck without the ability to run Firefox 45.9, I cant get it working with the latest updates to ANPM installed, and I can't downgrade libc and libcx. Is there any way of getting a copy of the .RPM version? I see now that supposedly it is only available to beta testers.
Thanks klipp
The YUM installer
https://ecsoft2.org/rpmyum-os2
and here for a good explanation and how to install components
https://ecsoft2.org/firefox
-
Hi Dave, Here are the last Firefox entries in popup.log:
02-12-2019 10:22:57 SYS3175 PID 0053 TID 0001 Slot 0079
K:\USR\BIN\FIREFOX.EXE
c0000005
11aca9cb
P1=00000001 P2=0000000c P3=XXXXXXXX P4=XXXXXXXX
EAX=00000000 EBX=16e95b5c ECX=ffffffff EDX=00000000
ESI=16e95ae0 EDI=00000000
DS=0053 DSACC=d0f3 DSLIM=5fffffff
ES=0053 ESACC=d0f3 ESLIM=5fffffff
FS=150b FSACC=00f3 FSLIM=00000030
GS=0000 GSACC=**** GSLIM=********
CS:EIP=005b:164e14f5 CSACC=d0df CSLIM=5fffffff
SS:ESP=0053:0013ff6c SSACC=d0f3 SSLIM=5fffffff
EBP=0013ff78 FLG=00010297
PSPRINT.DRV 0001:0001a9cb
------------------------------------------------------------
02-12-2019 11:49:26 SYS3175 PID 0157 TID 0001 Slot 0079
K:\USR\BIN\FIREFOX.EXE
c0000005
11aca9cb
P1=00000001 P2=0000000c P3=XXXXXXXX P4=XXXXXXXX
EAX=00000000 EBX=16d65b5c ECX=ffffffff EDX=00000000
ESI=16d65ae0 EDI=00000000
DS=0053 DSACC=d0f3 DSLIM=5fffffff
ES=0053 ESACC=d0f3 ESLIM=5fffffff
FS=150b FSACC=00f3 FSLIM=00000030
GS=0000 GSACC=**** GSLIM=********
CS:EIP=005b:164514f5 CSACC=d0df CSLIM=5fffffff
SS:ESP=0053:0013ff64 SSACC=d0f3 SSLIM=5fffffff
EBP=0013ff70 FLG=00010297
PSPRINT.DRV 0001:0001a9cb
Safe mode gives the same error and I've added or installed no plugins. I'm really dumb on browseres,
David G Thanks for the help I've got RPM installed and rpm'ed firefox-highmem-45.9.0-1.oc00.i686.rpm and it looks like it installed but can't see where to go to see if it runs! Maybe I should delete usr/bin/firefox.exe? and run it again?
-
Hi Klipp, now that is a weird crash that I haven't seen. Let me understand your situation.
This is the latest Firefox from Github? Installed into @UNIXROOT?
It was running fine until you updated with ANPM and you just had the regular Netlabs repository enabled when you updated?
What version of OS/2 (eCS or ArcaOS) are you running?
What printer are you using and how is it being used? Through cups or a regular printer object?
As for the firefox-highmem package, that just marks the DLLs to load into high memory. I take it you did that after the crashes started?
Have you tried one of my Firefox builds?
-
Hey Dave, It'll probably be a couple of days before I can get back to you. Gotta go to the hospital (heart Problems) for some tests!
Thanks for the help.
klipp
-
Hope it goes well for you
-
I had the psprint problem on one of my systems too. It was with the latest Seamonkey. In the end I uninstalled some printers and got it working.
Before I tried different ps drivers too without success. Copied them over from a working system did not work too. Think I used the pr...exe comming with pmfax(?) to delete the queues and drivers until SM was happy. I guess my printer installations where messed up badly. Though I had > 6 printers installed which I didn't even owned at that time with different queues and.... So cleaning up this mess was not a bad idea anyway.
-
Hi Dave, I'm finally back, they put three more "stents" in my heart! Aagghhh! Yesterday I got back to trying to make Firefox 45.9 once again work, no luck. I gave up went to bed. This morning I got up and went back to work and all 4 machines were waiting for me with Air-Boot waiting for my command (Lost electricity during the night!) I booted up and now Firefox 45.9 works fine?¿?¿ Can't explaine it, Guess best blame it on Mexican electricity here. The voltage here does periodically vary from 105 volts to 130 volts.
Thanks for your assistance once again.
Now I'm back to trying to make DropBox work once again, it worked a couple of years ago but now its back to the "no token found" error.
klipp
-
ArcaOS 5.03
OS/2 kernel
45.9.0 FFPentium
I have had several traps while doing all this testing. I have attached two trap messages.
-
Hi David, the first and third trps are familiar. I think that Mozilla keeps a list of fonts, perhaps including attributes and other stuff in the pldhashtable and now and again it gets corrupted,perhaps by something weird in a font, perhaps for some other reason. I don't have the skills to say much more then that.
They seem to happen here perhaps once a month on average, they'll be a couple within days and then no more for a couple of months. I've seen it in SeaMonkey and Thunderbird. Something to live with I guess.
The middle one, 0048_23.trp, is one I've never seen. Our TCIP stack seems to have had a problem looking up a host name, possibly netflix or something netflix loads. Not sure if caused by FF passing corrupted data or what. Hopefully it is a one time thing.
Are you using a hosts file to block ads? If so, you might want to review it for any weird characters.
-
Hi David, the first and third trps are familiar. I think that Mozilla keeps a list of fonts, perhaps including attributes and other stuff in the pldhashtable and now and again it gets corrupted,perhaps by something weird in a font, perhaps for some other reason. I don't have the skills to say much more then that.
They seem to happen here perhaps once a month on average, they'll be a couple within days and then no more for a couple of months. I've seen it in SeaMonkey and Thunderbird. Something to live with I guess.
Dave, I know on one of my partitions I use the patch
http://hobbes.nmsu.edu/download/pub/os2/system/patches/PMFIX.zip
FIX1: SYS3171 (Stack Overflow) on PMSHELL.EXE and others. FIX2: USING PM_SystemFonts->DefaultFont CHANGES ALL FORM SIZES
I cannot remember if I patched both ArcaOS partitions are not. I will try to check tomorrow. On the partition with the OS/4 kernel I am not seeing the traps. FFsandybridge with the O4 kernel is running everything I throw at it, especially since I corrected the FFturbo. It is on the other ArcaOS partition I am seeing the traps. I may not have patched its PMShell.exe. Will find out tomorrow.
-
Well Mozilla apps don't use any fonts that aren't true-type or PS so probably not related to the FIX2. Mozilla does pull the font list from OS2.INI so perhaps the patch has something to do with it.
Interesting that you're getting no crashes with the OS4 kernel.
As for ffturbo, its function is mostly to work around a bug in the OS/2 kernel that is fixed in the OS4 kernel (not releasing high memory when closing the app) though it will still help prevent fragmentation where you end up with only small memory holes. I'd assume the OS4 kernel has other bug fixes that may prevent those trps. Wish I was more knowledgeable :)
-
Dave, I copied the patched PMMerge.dll fom my OS/4 partition to my OS/2 partition. On this partition, I am using the OS/2 kernel with the FFPentium. This is just preliminary, but the trapping has disappeared. This is early, however, it did not take long to trap with the original PMMerge.dll.
-
The middle one, 0048_23.trp, is one I've never seen. Our TCIP stack seems to have had a problem looking up a host name, possibly netflix or something netflix loads. Not sure if caused by FF passing corrupted data or what. Hopefully it is a one time thing.
Are you using a hosts file to block ads? If so, you might want to review it for any weird characters.
No more XUL.dll traps. However, I had more traps with the tcpip32.dll. Investigating I found that the tcpip32.dll are different for the ArcaOS partitions.
ArcaOS 5.03 dll is dated 2004 and the ArcaOS 5.02 is dated 2011.
It is the ArcaOS 5.03 dll that is crashing. I have attached the traps and pic of the dlls.
Can someone with ArcaOS 5.03 check the date on their dll? I don't recall changing this dll.
-
Looks like somehow you have the original tcpip32.dll, which was not SMP safe. The 2011 one is patched for SMP.
Try replacing the DLL. Or update ACPI, which should update all the SMP unsafe stuff.
-
Looks like somehow you have the original tcpip32.dll, which was not SMP safe. The 2011 one is patched for SMP.
Try replacing the DLL. Or update ACPI, which should update all the SMP unsafe stuff.
Thanks Dave. Replaced it and now to continue to use and monitor the OS/2 partition.
-
I'd assume the OS4 kernel has other bug fixes that may prevent those trps. Wish I was more knowledgeable :)
OS/4 has quite a lot of different bug fixes and rewritten parts. It's very difficult to say which bug fixes have prevented those traps.
-
Looks like somehow you have the original tcpip32.dll, which was not SMP safe. The 2011 one is patched for SMP.
Try replacing the DLL. Or update ACPI, which should update all the SMP unsafe stuff.
I have been extensively testing the FFPentium with the OS/2 kernel. Since I have patched the PMMerge.dll and corrected the tcpip32.dll, this system appears to be trap free. I will continue to test and monitor since you and others have stated that it may go days or weeks before a trap appears.
The other partition with the OS/4 kernel and the FFSandybridge with the AVX libs continues to operate trap free. Again, this partition is also using the patched PMMerge.dll.
This is also using the FFturbo.exe on both partitions also.
Try as I might, I cannot get the FFPentium brower to trap. I have gone to every site that has trapped it before and others numerous times, repeatedly shut it down and restarted, and the thing just won't trap.
David
-
I spoke too soon Dave. I had a trap, but it did not crash FF. It just would not play any more youtube video until I closed it and restarted it. I attached the trap. It was the XUL.DLL This was with the OS/2 kernel with the FFPentium.
Now to see if the OS/4 kernel crashes with the FFSandybridge.
-
It's an OOM (out of memory) trap. Notice the CPU instructions,
15753F5D INT 3 (cc)
15753F5E >NOP (90)
Mozilla put in a breakpoint (INT 3) for certain out of memory scenarios.
Hmm, actually, this is different from the usual OOM handler, this is a watchdog thread that'll issue a breakpoint if the browser hangs during shutdown. Seems that it still might be a memory leak or such. You might still have a file, ShutdownDuration.json or ShutdownDuration.json.tmp in your profile with info about what actually happened.
My guess is that video playback stopped due to out of memory condition and the browser hung during shutdown for similar reason and the watchdog thread killed it.
BTW, you should install the debug packages for NSPR4, libc and libcx
-
I spoke too soon Dave. I had a trap, but it did not crash FF. It just would not play any more youtube video until I closed it and restarted it. I attached the trap. It was the XUL.DLL This was with the OS/2 kernel with the FFPentium.
Now to see if the OS/4 kernel crashes with the FFSandybridge.
Dave,
I see this trap from time to time on my comp. It is described in https://github.com/bitwiseworks/mozilla-os2/issues/243
-
I must say that all the 49.5.0 FF versions are all pretty stable. After OS2User asked me to watch CPU loads while watching videos, I can definitely state that the OS/4 kernel does a much better job when playing video. I am now back to the OS/4 partition and will stay there to see what kind of trap if any I can produce with the OS/4 kernel, the FFSandyBridge, and the AVX dlls. It has yet to trap, but that is what I said about the other until it finally trapped. I am finally enjoying using the internet under ArcaOS. Only have to go to Windows to watch Netflix or Amazon Prime Video.
There is one thing I have noticed about FF. When watching video, the sound cannot be controlled except by the operating system volume controller. The sound volume cannot be adjusted on the video itself.
-
While using my OS/2 partition and FFPentium, I had one more trap which this time closed the browser. Have attached the trap.
-
While using the OS/2 kernel, the FFPentium 45.9.0, and noAVX libs, I had a strange trap.
-
It's much the same as usual, pldhashtable and fonts.
-
Though this is an old thread I'm using it since it addresses the same situation I have.
I have both Firefox 45.9.0 and Seamonkey 2.42.9esr running fine on my ThinkPad T-540p. I'm running FF under @unixroot; FF is also installed on a separate partition and directory. Likewise, SM is installed on a separate partition and directory.
The data for <xul.dll> is:
@unixroot
29,398,067 05/21/18
Programs\firefox
37,270,233 03/21/20
Telecomm\seamonkey
41,620,640 08/17/19
Despite the discrepancies all three programs open fine.
However, on my desktop I can't get any of them to run. (Both the ThinkPad and Desktop are at the ArcaOS 5.0.7 level) The data for <xul.dll> on it is:
@unixroot
29,398,067 05/21/18
Programs\firefox
37,270,233 03/21/20
Programs\seamonkey
39,651,277 03/20/20
Telecomm\firefox
41,023,852 11/11/20
Telecomm\seamonkey
41,741,025 06/12/22
Entering <-safe-mode> in the Parameters didn't help.
The applicable <popuplog.os2> entries are:
01-21-2023 18:34:29 SYS3175 PID 0098 TID 0001 Slot 0020
O:\TELECOMM\SEAMONKEY\SEAMONKEY.EXE
c0000005
0ae9a9cb
P1=00000001 P2=00000002 P3=XXXXXXXX P4=XXXXXXXX
EAX=00000000 EBX=11664c90 ECX=0c9ce020 EDX=00000000
ESI=11664bf8 EDI=00000000
DS=0053 DSACC=d0f3 DSLIM=5fffffff
ES=0053 ESACC=d0f3 ESLIM=5fffffff
FS=150b FSACC=00f3 FSLIM=00000030
GS=0000 GSACC=**** GSLIM=********
CS:EIP=005b:0fc6cf1b CSACC=d0df CSLIM=5fffffff
SS:ESP=0053:001eff64 SSACC=d0f3 SSLIM=5fffffff
EBP=001eff70 FLG=00010297
PSPRINT.DRV 0001:0001a9cb
------------------------------------------------------------
02-08-2023 15:21:23 SYS3175 PID 0089 TID 0001 Slot 000c
J:\@UNIXROOT\USR\LIB\FIREFOX-45.9.0\FIREFOX.EXE
c0000005
0ba407c7
P1=00000001 P2=00000002 P3=XXXXXXXX P4=XXXXXXXX
EAX=00000000 EBX=10d7e384 ECX=0e430aa8 EDX=00000000
ESI=10d7e2f4 EDI=00000000
DS=0053 DSACC=d0f3 DSLIM=5fffffff
ES=0053 ESACC=d0f3 ESLIM=5fffffff
FS=150b FSACC=00f3 FSLIM=00000030
GS=0000 GSACC=**** GSLIM=********
CS:EIP=005b:104dc427 CSACC=d0df CSLIM=5fffffff
SS:ESP=0053:0012ff5c SSACC=d0f3 SSLIM=5fffffff
EBP=0012ff68 FLG=00010293
PSPRINT.DRV 0002:000107c7
------------------------------------------------------------
02-08-2023 19:20:30 SYS3175 PID 0104 TID 0001 Slot 00dc
O:\TELECOMM\FIREFOX\FIREFOX.EXE
c0000005
0ba407c7
P1=00000001 P2=00000002 P3=XXXXXXXX P4=XXXXXXXX
EAX=00000000 EBX=10c75550 ECX=0bcbe9f0 EDX=00000000
ESI=10c754c0 EDI=00000000
DS=0053 DSACC=d0f3 DSLIM=5fffffff
ES=0053 ESACC=d0f3 ESLIM=5fffffff
FS=150b FSACC=00f3 FSLIM=00000030
GS=0000 GSACC=**** GSLIM=********
CS:EIP=005b:0ed54cbb CSACC=d0df CSLIM=5fffffff
SS:ESP=0053:0013ff64 SSACC=d0f3 SSLIM=5fffffff
EBP=0013ff70 FLG=00010293
PSPRINT.DRV 0002:000107c7
I'm fairly certain I have all the dependencies installed. The last ones I noted in the RPM Requirements for the <@unixroot> install I attempted to install through <yum install> and te following was the output:
yum install libgcc1 libstdc++6
[J:\]yum install libgcc1 libstdc++6
Setting up Install Process
Package libgcc1-4.9.2.1-3.oc00.pentium4 is obsoleted by libgcc-9.2.0-5.oc00.i686
which is already installed
Package libstdc++6-4.9.2.1-3.oc00.pentium4 is obsoleted by libstdc++-9.2.0-5.oc0
0.i686 which is already installed
Nothing to do
I searched the boot drive and <psprint.drv> was installed in both <OS2\dll\psprint> and <sys\install\prndrv\pmdd_6> and are identical, both being 827,719 and dated 08/20/18. Printers on the Desktop are:
a Dot Matrix installed on LPT1,
a HP DeskJet installed on USB,
the PMPDF generator, and
a Postscript printer configured for LPT2.
Right now I have no recollection why I installed the last one. I probably can delete it if this might be the cause.
Any ideas?
-
That's a new one to me. Have you tried a new profile? --ProfileManager in the Parameters. Also whether it crashes before the profile manager comes up.
Have you installed any printers in FF/SM? If so perhaps try remming out psprint.drv, try starting FF/SM and remove all printer entries. Then unrem psprint.drv and see if they still crash.
Another possibility might be something wrong with your ini files, have they been cleaned lately?
-
Inserting
-ProfileManager
in the Parameters still causes a crash and throws out the same "Couldn't load XUL.DLL (OS2 error 87)" error.
I may have installed some printers in FF 38.8.0 but not 45.9.0 or Seamonkey since I could never get them to run. In fact I don't recall any menu for printer installation under any of them.
I see no reference to <psprint.drv> in the <config.sys>. I did try to rename it, to see if this helped, but failed since it said it was being used by another application. Accordingly I deleted the printer object for the PS Printer I had created - since I probably had created it erroneously thinking PMPDF may have required it, or for creating files for Distiller - and rebooted. Still the same problem.
As far as my INI files I have only relied on IFX examining and cleaning them. I did, long, long ago, purchase a license for UniTools and have it installed on the desktop. Do you think I should give this a whirl?
-
Well, if it is crashing before displaying the profile manager, that's a sign that it has nothing to do with your profile and the problem is deeper.
To remove something that is open, do unlock psprint.drv. The unlock comand will load the file into memory so you can move it out of the way and then reboot.
Wouldn't hurt to clean your ini files, I like checkini, https://hobbes.nmsu.edu/download/pub/os2/util/config/WPTools_3-3.zip (https://hobbes.nmsu.edu/download/pub/os2/util/config/WPTools_3-3.zip). Back up your desktop first by right clicking the desktop and choosing restart, then click archive desktop.
You could also try reinstalling cups, yum reinstall cups, IIRC, or use ANPM, right click on cups and choose reinstall. Probably won't help but easy to do.
BTW, error 87 is "ERROR_INVALID_PARAMETER"
Try seamonkey -help and see if it runs long enough to leave a vio window displaying the help, might need the window list to find it.
Also try running seamonkey 2>&1 | tee seamonkey.log and see if anything appears in seamonkey.log, best on a cmd line in the SeaMonkey directory.
Otherwise, am out of ideas, have to think on it.
-
OK, will give these a try today, or at east some of them between other commitments. I'll try your recommendation of using WPTools before UniMaint; I suffered a Senior Moment yesterday when I dubbed UniMaint as UniTools.
I did last night before retiring to bed boot the desktop into Linux and used DFSee to create an image backup of 5.0.7. Possible messing up these INI efforts then won't be dangerous.
The residing of PMPRINT.DRV in the INI files may well be the problem; the old creation of the PS Printer Object likely inserted it. I do not presently perceive any danger in removing it as I don't conceive any need for it; the only potentially-associated application would seem to be PMPDF and it doesn't seem to use or need it.
I'll report later. Thanks for all your effort.
-
If those don't help, I guess we'll have to try to check for broken DLL's in the dependencies.
Another test would be to try the screensavers Cairo clock, with the date turned on, as that would test the fontconfig and such and whether they're broken. Desktop properties is where you find the screensaver settings.
-
Since the exception is happening in the PSPRINT driver, I suggest you delete all PM printers (or at least all of those using PSPRINT). If that fixes it, you can then reinstall printer(s) one by one and make sure Firefox still runs.
(By delete, I mean fully delete the printer queue and the driver instance, which you can do using PM Printer Manager, or the old FaxWorks PRNDRV.EXE.)
-
I found an issue where Lewis mentioned the same problem, https://github.com/bitwiseworks/mozilla-os2/issues/255 (https://github.com/bitwiseworks/mozilla-os2/issues/255) which was from importing a PPD into psprint.drv, which as he put it, wrecked havoc with the native print system. Also confirms that F 45.9 tries to initialize the printing system upon program load
So follow Alex's advice to remove all printers
-
SUCCESS!!! I now have FF 45.9.0 and SM 2.42.9 opeating on the Desktop.
I ran <check.ini> yesterday and, while it seemed to report FF w/out an ID - though this could have referred to 38.8.0 installed w/ ArcaOS 5.0.0 - the log didn't show anything. (It seemed it was doing something w/ the INI files since the Archive operation was activated during the process - I having it set to automatically archive for any change greater than 2.5% and every 15 days - but neither FF or SM still functioned afterward. (However since I was interrupted in following the whole process by attending to a church member who was here fixing a faulty circuit breaker, I may have missed something.)
Before trying a clean w/ UniMaint, I today moved PSPRINT.DRV to another partition, and then rebooted. On running both FF and SM the Profile Manager window for each opened and I created them. I then instructed them to open and they did what they were supposed to do.
I've noted the intervening three messages and will print them for future reference and possible future use. I had last week deleted the PS Printer Object I had created but this didn't help. Since FF and SM now are operating I haven't endeavored to pursue the printer delete suggestion; I'm not sure it would have had any effect since the printers consist of: a dot matrix connected to LPT1; an HP DeskJet w/ a USB connection; and the PMPDF generator. (I also do have the PMFax driver installed but this doesn't seem to cause a problem. (It's a shame configuration of PMFax w/ new hardware is so difficult since it was a fine answering machine, I now having to use an application under XP on a different computer for this purpose.))
The situation was definitely strange since on this ThinkPad T-540p PSPRINT.DRV continues to reside in <os2\dill\psprint\> and it never stopped FF or SM from opening. I suppose it must have been that PS Printer Object I created long ago that inserted something weird in the INI files.
Thanks Much to all for their help.
-
As an addendum to yesterday's message, I know that my solution was not particularly elegant. But it accomplished what I immediately required.
When I have more time I plan to use the Registry Editor and UniMaint to locate the PSPRINT.DRV references and explore how any anomalies might be corrected. The USB connected printer might require CUPS and thus might need the driver; right now, though, I'm not using it on the desktop computer and can defer resolution.
At least I can save the current partition and files, to revert to them, if the efforts and movement back of PSPRINT.DRV again causes problems.