Prozactive

JorgeA, If your system passed 24 hours of the Prime95 torture test, you can be extremely confident you don't have any hardware issues. Prime95 is a very popular tool in the overclocking community for testing system stability. In my experience, the test will fail within the first few iterations if you have any hardware problems. I've actually never run it the full recommended 24 hours but that would certainly give you more confidence. Good luck in finding the next Mersenne prime! Well it was worth a shot trying older versions of Spybot. I'm pretty much at a loss for other ideas to try. I'd be interested in hearing about your experience with SuperAntiSpyware if you decide to use it. I've DL'ed but not installed it yet. I'm currently having some strange problems with Avast! antivirus and may eventually have to bail out and install Norton AntiVirus by *gasp* Symantec as suggested in another thread if I can't resolve them. Like so many others, Avast! is discontinuing support for Win9x/ME shortly. If I do that, it'll be the fourth or fifth major AV change I've had to make within the last few years.

rilef has some good advice and brings up a point that I've also meant to discuss. You've mentioned that Spybot worked fine in the past with Norton Internet Security, SystemWorks, etc. installed, then you started getting the page fault error at some point later on. I recently reviewed my Windows error logs and noticed I did get a very similar Spybot page fault error to yours at some point in the past. I don't recall the specific circumstances of the crash but I suspect it was after one of the Spybot version updates (1.4.x to 1.5.x to 1.6.x). Apparently, unlike you, it only occurred once and hasn't recurred since. It would be worth following rilef's advice and reverting back to a previous version of Spybot to see if that makes a difference. I fully agree with georg - you are very determined. I can relate to that, but at some point it's probably time to throw in the towel from a time and work cost-benefit perspective. You've gone through considerable advanced technical troubleshooting, courtesy of georg, et al., and like you I've learned a lot from reading this thread. I personally suspected your problem was caused by incompatibilities with all your $ymantec system utilities but apparently not. I'm not a big fan of $ymantec at all, as I find their products extremely intrusive, overbearing, and bloated. To me they're the Microsoft (aka "The Evil Empire") of system utilities and it was interesting reading a few replies back that the two companies work closely together, which I didn't realize. Norton Utilities was a fantastic product way back when Peter Norton actually owned them (before he sold out to $ymantec). Also, thanks for the shoutout but I don't recall giving any specific anti-malware advice, at least not in this particular thread. I did suggest using AVZ (Antiviral Toolkit) recently in another thread discussing Win9x compatible anti-malware software. By the way, I have confirmed that Spybot definitely loads faster with the recent new definitions updates even though they're significantly larger in size. It now takes just under 3 minutes to load whereas before it was taking between 5 to 7 minutes.

Oh duh. I can't find the mp3 codec listed in WMP 9 but it seems to work fine.

I have confirmed that the latest version 5.7.6002.22145 of Jscript.dll from KB975542 installed by the recently updated SCR579X Unofficial MSE 5.7 update causes the Windows Media Player 9 error "An internal application error has occurred." that I reported. Reverting back to the previous version of Jscript.dll (5.7.0.16535) or reinstalling JScript 5.6 (version 5.6.0.8837) via SCR569X fixes the error. I also tried this command to reregister Jscript.dll (as suggested by several technical websites): regsvr32 c:\windows\system\jscript.dll but got the following error: LoadLibrary("c:\windows\system\jscript.dll") failed. GetLastError returns 0x0000001f I was able to reregister the previous version 5.7.0.16535 of Jscript.dll by the way. Since version 5.7.6002.22145 of Jscript.dll implements a critical security update, I will probably revert back to version 5.6.0.8837 which also implements the update. HTH

Thanks MDGx. Like I said, I have not completely verified that the latest version 5.7.6002.22145 of Jscript.dll causes the WMP 9 errors but I strongly suspect it. I'll report back as soon as I get some definitive results. I believe I have successfully updated wmp.dll to version 9.0.0.4508 from KB979402 by manually extracting it from the WinXP SP3 update file, replacing the existing version 9.0.0.4507 (from MP9URP9X), then running the following command to fix the WMP 9 error message saying it expected to find version 9.0.0.4507 of wmp.dll: C:\Windows\Inf\unregmp2.exe /UpdateWMP Similarly, I have also manually updated the MPEG Layer-3 codecs l3codeca.acm and l3codecx.ax from KB977816 (reference this MSFN thread) but I'm not quite sure how to test them yet. Thank you once again for all your hard work and support!

Thanks for the information. Is there a good way to test and verify that the updated files work properly? BTW, my existing files were also: l3codecx.ax 12-Dec-2002, 12:14:32 AM, 83,456 bytes, version 1,5,0,50 l3codeca.acm 03-Oct-2008, 2:42:10 AM, 290,816 bytes, version 1,9,0,0305 I believe the 03-Oct-2008 version of l3codeca.acm came from the latest version of MP9URP9X.

@MDGx: I'm still in the process of diagnosing the problem but it strongly appears that the latest version 5.7.6002.22145 of Jscript.dll from KB975542 causes an error in Windows Media Player 9 (after installing your revised SCR579X update). WMP 9 now gives the following error message: "An internal application error has occurred." Interestingly, WMP 9 worked fine with version 5.7.0.16535 of Jscript.dll after installing your previous version of SCR579X. I finally had to install JScript 5.6 via SCR569X to get WMP 9 to work again but I'm now getting some strange rectangular gray blocks in the video display for some reason. On a related note, could you please update MP9URP9X to incorporate the revised wmp.dll and MPEG Layer-3 codecs from the recent KB979402 and KB977816 updates, respectively? Thanks! @Fredledingue: Coincidentally I was researching that myself recently while trying to figure out how to manually install the latest KB978262 ActiveX Kill Bits security update. Apparently it's a utility provided by Microsoft in WinXP and later to create an installer for various system files. More info: http://en.wikipedia.org/wiki/IExpress

Prozactive, I thought I'd replied to you but it doesn't look like I did, sorry about that. A few weeks ago I discovered the Windows Memory Diagnostic tool. I used it and the RAM passed with flying colors. (The Spybot issue has been going on for almost two years.) My hard drive monitor reports a very healthy HDD. Is there any other hardware I should/can test with software? Thanks very much. --JorgeA JorgeA, Sorry I also missed your reply in this extremely long convoluted thread. That's good that your system RAM passed the WMD test. (That acronym always makes me laugh, BTW.) It rules out one possible cause of your errors. I don't believe you're having hardware-related problems, but one excellent overall test of your CPU/memory and other subsystems is the torture test in Prime95. If your system passes an extended run of that test, you can be quite confident everything is working properly. I have not thoroughly read and studied all of the long complex replies in this thread but you are definitely receiving a lot of excellent technically competent help. I'm sure you've already done this, but the general rule if something worked at one point then suddenly started giving errors, is to determine what system changes occurred in the interim then proceed with a methodical step-by-step clean boot troubleshooting procedure. I believe this is what you're doing now. On a somewhat different but related note, about 2 or 3 weeks ago Spybot apparently changed their definitions updates and now the program loads significantly faster. I haven't done a full system scan so I can't report if that is also faster. Good luck with your troubleshooting!

Good advice from georg, et al. I recommend you also rule out memory and other hardware issues as a possible cause of your problems by running various diagnostic tools like Memtest86, Windows Memory Diagnostic, etc. Spybot runs fine on my Win98 systems although quite slowly as others have mentioned. On my slowest system (mobile Celeron 700mHz, 192MB RAM), it usually takes about 5-7 mins to load and around 1.5 hrs to scan my 6GB Win98 partition. I've never encountered any page fault or other errors, although after the last definitions update Spybot locked up after initial loading and eventually stopped responding in Task Manager for some reason (on all of my systems). It subsequently worked fine after reloading again.

@duffy98: Versions 4.10.2230 and 4.10.2225 (updated) of ESDI_506.PDR are identical. This was recently discussed in this thread. @dencorso: Thank you for the info on the updated versions of KERNEL32.DLL and KRNL386.EXE. I didn't have those updates installed on my systems.

Thank you dencorso for your response. BTW I assume you meant 1 TB as the maximum HDD size limit. Since my initial post, I've come across more information that strongly suggests that 1 TiB is the maximum destination HDD size limit for Ghost 2003. It also appears that this is a hard disk not partition size limit, much like the 137 GB 48-bit LBA limit. You bring up a good point - I don't know what the maximum source HDD size limit would be. That would be interesting to test and experiment around with. Yes I also noted the erroneous info cited about Ghost 2003 vs. the more recent and updated Ghost 8.3 (as part of Ghost Solution Suite 1.1). However, I'm not familiar with the cdrlib parameter you mentioned. What exactly is that? EDIT: I just noticed your reference to Dan Goodell's post on Radified forums. You must've edited your initial reply as I don't recall seeing that. Thanks, and yes that's exactly where I got the technical information about 1 TiB being the likely size limit. (I'm also a member of that forum BTW and did post a similar inquiry.)

I can't say I fully understand all of your issues but I suggest you try the Native USB drivers (NUSB) for your flash drive(s) instead of the vendor-specific drivers you mentioned. They're available at MDGx's website: NUSB 3.3 About the drive letter issue(s) - you know you can specify a desired drive letter within Device Manager by editing the appropriate Disk drives entry's Settings tab. I've read about deleting ENUM and other registry settings and files to port over an existing Win98 installation to a completely different hardware platform but I've never tried it myself. I'm seriously considering doing it for my next Win98 system build as there's really nothing to lose. From your experience, are there any specific tricks and advice you could share to increase the likelihood of success?

I must say I'm shocked ("surprised" is too mild a word) to hear about *any* problems with Ghost. This is definitely the first time I've seen a problem reported. I've personally used Ghost 2003 for years and Ghost 6.0 for several years before that, successfully restoring dozens of images without any problems whatsoever. It's saved my bacon countless times, and I consider it one of the most valuable programs I own. BTW I always use fast compression and save the images to another HDD, and I always run an integrity check on the image afterwards just to be safe. I've never had an image fail its integrity check either. I have no experience directly burning an image to optical media, as I don't trust or like that method at all. I can imagine problems cropping up from doing so. I also have never used versions of Ghost from 9.0 onward. I understand they're based on PowerQuest's DriveImage and are completely different "under the hood" from previous versions. Still hoping to hear technical feedback from the MSFN community on Ghost 2003 HDD size limitations...

Recently there has been a lot of excellent technical discussions about the limitations of various utility programs (FDISK, FORMAT, SCANDISK, Norton Disk Doctor, etc.) vs. today's increasingly large hard drives. I was wondering if anyone knows anything about similar limitations of Norton Ghost 2003 (a DOS program)? Web searches and anecdotal reports indicate problems with HDDs in the 1 TB to 1.5 TB range with "divide overflow" errors often mentioned, but I have not seen anything definitive. I personally do not have a HDD in this size range to test and experiment around with, but recently I helped a friend make a Ghost image backup to his new 1 TB NTFS-formatted external HDD. I had a lot of difficulty getting various DOS USB drivers to recognize the large HDD (I believe I did get some "divide overflow" errors), but once I managed to do so, Ghost 2003 did not have any problems writing the image to the drive. Of course the new HDD was empty so the image was written well below any "hard size limit". It would be very useful to know the maximum size limit for Ghost 2003, and steps one could take (repartitioning, etc.) to compensate. Thanks for any information on this topic!

Thank you for your long and thorough reply, MDGx. I was not aware that your DST update had been revised. I installed it back in January when you first released it and it worked normally on another Win98 system but not my laptop for some reason. I was also not aware of the old Microsoft Win98 TIMEDATE.CPL hotfix you referenced. I thought I had installed all older Microsoft Win98 updates but obviously not. Anyway, I installed your revised DST update along with the Microsoft hotfix and tested it using your procedure - everything works normally now. Thanks again and I hope you get your PC repaired soon!

That is a good question and coincidentally is something I'm planning to check out more thoroughly soon. The short answer is no, WinXP does not give itself drive letter C:. That will remain with the primary Win98 partition. WinXP apparently gives itself the volume letter it's installed into, but that's the issue I want to investigate. It apparently depends on how the disk is partitioned. The hard disk in my laptop is partitioned into 3 volumes (2 FAT32 and one NTFS, in that order), and WinXP is installed in the (last) NTFS partition. When booting into WinXP, it assigns drive letter F: to its NTFS partition, behind C: and D: for the 2 FAT32 partitions and E: for the CD-ROM drive. The hard disk in one of my desktop systems is partitioned into 4 volumes (one FAT32, one NTFS, and 2 FAT32, in that order) and again WinXP is installed in the (2nd) NTFS partition. WinXP assigns drive letter D: to its partition which is causing some problems as I prefer a different drive letter. Apparently there is some way in WinXP to change assigned drive letters through its disk management function but I have not investigated this yet. I'm not very familiar with all the nuts and bolts of WinXP as I don't use it that much, strongly preferring Win98 (obviously).

I believe that is true. I dual-boot WinXP and Win98SE on several of my systems, with WinXP usually installed in an NTFS partition. rloew and the xt guy are correct - Win98 will not see the NTFS partition and will increment all subsequent drive letters up one letter. Of course WinXP will see all partitions (FAT and NTFS) and assign drive letters accordingly. There is no problem with registries. Win98 and WinXP maintain separate versions of their respective registries. HTH

@MDGx I installed the Unofficial December 2009 Q976098 Daylight Saving Time update but it did not automatically update my system time this morning after DST took effect. (?!?)

I agree. I would remove the HD, put it in an USB enclosure, then use another system with USB to image it using Norton Ghost or some other imaging program.

Thanks for the heads-up halohalo.

Thanks for the clarifications on version numbers and technical details, rloew and dencorso. I do recall reading a long discussion/debate on the original LLXX thread on version numbering but I didn't know how things finally sorted out. @Fredledingue: That's puzzling. I've only encountered MS-DOS compatibility mode on HDs connected to additional ATA/RAID controllers on which I haven't loaded drivers for yet. Once I load the proper drivers, that problem goes away. Is your I: drive on one of those type controllers? I've never experienced it on HDs connected to my main IDE controllers. There is the possibility that the modified ESDI_506.PDR isn't compatible with your IDE controllers for some reason. I'm assuming your HD is IDE/PATA. Thanks for the info and batch file code for installing Win98 on >137GB HDs. Just FYI, the referenced URL for "install-w98-large-drives-above-137gb-barrier" in your post doesn't work now that MSFN has changed their URL syntax. The correct URL should be: http://www.msfn.org/board/index.php?showtopic=113142 HTH

Thanks dencorso and lightning slinger. I did read most if not all of the loonnggg Enable48BitLBA LLXX thread back when I was researching the data corruption problem on my >137GB HDs and trying to find a solution. I finally DL'ed and installed an updated ESDI_506.PDR from that thread via MDGx's website I believe. Per advice in that thread, I wanted to find a utility that would write dummy data to the >137GB sectors on my HDs to thoroughly test out the updated ESDI_506.PDR before actually using it with real data, but I was unable to locate one. However, almost 2 years of use without any problems has given me confidence that it works properly. The updated version of ESDI_506.PDR that I installed back then had an indicated file version of 4.10.2225. I'm convinced that it's identical to the latest "updated" 4.10.2230 version installed by MDGx's Atadrv98.exe update. I've finally run a binary file comparison between the two versions and these were the only differences detected (as I already noted earlier via visual inspection): 00005D86: B1 B6 00005E4D: 32 33 00005E4E: 35 30 The last 2 differences are just the version number changing from 4.10.2225 to 4.10.2230. (Edit: Okay I just read your first linked thread dencorso and essentially you're confirming what I wrote here about the file versions. Thanks!)

What kind of problems do you experience, Fredledingue? These builds are *exactly* the same date/time (to the second) and size. I have not run a formal compare program but I did take a close look at both of them with an editor, and visually at least, I see no differences between them except in the description fields at the end. So I'm still very curious what changes and improvements were made in 4.10.2230.

Well I've installed the updated version 4.10.2230 of ESDI_506.PDR and it seems to be working fine so far without any problems as I expected. Just curious though.. do you know what was changed/improved in version 4.10.2230 from 4.10.2225? Both files have the exact same size and date (24,431 bytes, 20 July 2006, 12:38:32pm). Thanks!

I don't know what's causing your problem with Windows Explorer but I'm pretty sure you don't have a malware problem if Avast, Spybot, and AVZ scans are clean. AVZ is very thorough and does seem to flag various legitimate processes and files for various reasons. If you know you're the author of a script, then it's obviously not a problem. I have a similar and more aggravating problem with AntiVir in that it continues to detect several batch files I've created as FPs, and I can't seem to exclude them from its real-time scan. BTW, I came across another anti-malware program called Dr.Web CureIT! that claims to work in Win98. It gets good reviews but I have not personally tried it yet. You may want to try updating your version of Windows Explorer. Mine has been updated by NUSB33 (Native USB drivers) then again by Explor98, both of which are available at MDGx's website. HTH And thanks lightning slinger for the information on how to update SAS for Win98SE.