jds

Thanks, Tomas, I'll check that out (just downloading now) Thanks for the extra details, loblo. I did google WPF, and the half dozen hits I opened all said about that Wordperfect stuff. That's why I asked. As for the two dotNet installers I have, "dotnetfx3setup.exe" and "dotnetfx35setup.exe", neither of them have a "wcu" directory, etc. Hence my confusion. I'll download "dotnetfx35SP1.exe" and check it out. Thanks again, guys. Joe. PS. I've now downloaded 3.5SP1 from MS under the file name "dotnetfx35.exe". It's huge, about 10 times bigger than the 2.0 framework, and almost 100 times bigger than the two 3.0 and 3.5 installer packages I'd previously downloaded. In retrospect, it's now obvious that the files I had previously downloaded from MS were just downloaders, not the real dotNet packages.

Maybe the Norton "DiskEdit" utility from the Norton Emergency Disk 2001 (from Symantec)? Search for "ned_2001". For normal formatting needs, I often use FreeDOS. Joe.

Well, I started looking at this by extracting the dotNet 3 installer contents with 7-Zip. That resulted in a reasonably large bunch of files to sift through. So I deleted all the localized versions of the various files, which was about half of them, then all the files that had "setup" in their name or within their Version information, similarly any download related files, then anything else that didn't look useful. In the end, I had no files left. So it's not clear to me at all where are these "missing assemblies" that you mention within the dotNet 3 installer. If it's a file called 'system.core.dll', well this wasn't present in the stuff extracted by 7-Zip. BTW, I also don't know what is a "WPF" file, I presume it's not a Wordperfect Form? I don't think I would be alone in needing an extra little clue or two to get started with this dotNet 3.X stuff. TIA, Joe.

Thanks rainyd, it seems to work geat! Joe. PS. New build 1565 also working well : http://snapshot.opera.com/12.02-1565_windows.html

Yeah, that rings a bell. Maybe the SP should make the necessary registry setting to fix this? Else I think there was a patch from MS that did the same thing. Joe.

Hey, that's great news, loblo. Are you able to give a little more details on the process? Joe.

Uxtheme causes .NET problems on Win2K: Yeah, I saw some of these issues mentioned in regards the alternative 'uxtheme.dll' files from some of the unofficial W2000 updates, which is why I looked further afield. The only dotNet application I use (that I'm aware of) is the command line "odfconverter" utility. This seems to be functioning normally after installing the Wine-derived 'uxtheme.dll' file (and 'libwine.dll') in the 'system' directory. Also, the "netfx_setupverifier" utility says my dotNet (2.0) installation is still OK. Joe.

FWIW, this thing lists as "TCP/IP driver" in the version information, whereas by contrast, 'ndis.sys', which is about half the size, lists as "NDIS 3.0 wrapper driver". Joe.

I came across this interesting and potentially useful extension, a substitute for 'uxtheme.dll' which was introduced in XP. It's actually built from Wine sources, intended to fill the same hole in W2000, but according to Dependency Walker, it's fine for W98 too : http://kzar.co.uk/blog/view/fixing-spotify-uxtheme.dll-error-on-windows-2000 This is also a potentially useful approach to filling other holes in W98's set of DLL's. Joe.

Me too. That explains why jumper didn't see the Dependency Walker problem ... Joe.

Hey, I like your spirit! Walter Oney's book and code examples, which I referred to in my 'wdmstub' thread is probably your best starting point. Since UDF is already supported by MS drivers, and third-party NTFS solutions also exist (don't know myself how good they are), perhaps it would be easier/faster to just target EXFAT initially. Joe.

Well, that's OK if you're developing your own package, but if you need to run someone else's package that calls 'reg.exe', you can't just substitute 'reg2inf.exe' for it, you have to supply the required version of 'reg.exe'. BTW, finding which (9X or NT) version is required (the command line and output syntax differ) can either be by "trial and error", an "educated guess" or by more cunning means (eg. writing a dummy 'reg.exe' that records its command line parameters). I've so far had only one situation where I've needed 'reg.exe' and it was the NT version I needed, the 2k build I've described seems to run fine on W98. Joe.

Wow!!! What a fantastic resource, thank you so much jaclaz! Two thing I found noteworthy : Firstly, the page http://www.minuszerodegrees.net/failure/failure%20-%203lt.htm suggests there are two possible polarities for three-legged tantalum capacitors. Unless some evidence indicates otherwise, I believe this is hypothetical and only the negative-positive-negative polarity actually exists. Secondly, page 2-108 of the scanned IBM document "IBM_5150_Technical_Reference_6025005_AUG81.pdf" (yep, from 1981) shows the following signals on the floppy diskette interface (among others, of course) : Motor Enable A, Drive Select B, Drive Select A, Motor Enable B Joe.

Err ... shouldn't you have used Norton Disk Editor instead? I think there's a version of this within the Norton Emergency Disk ("ned_2001.exe", available from Symantec's FTP server). Does that mean more sectors in the outer cylinders & less in the inner cylinders? Joe.

Thanks for that link to John Elloitt's page, lots of useful details there. I notice it says ... ... which is the first explanation I've seen about why IBM contorted the Shugart interface (OK, I haven't exactly been looking), and is perhaps plausible, as I recall the power supply in the 5150 was quite underpowered by today's standards, something like 63W (if I recall). @Laser98IX : Have you tried either option suggested in my first posting (#18)? Joe.

Interesting. My 5150 was '83 or '84 vintage, 4164 based, and came with a contorted (twisted) floppy cable configuration. To this I added a Tallgrass Technologies HD interface card and external HD (12MB, as I recall). Perhaps yours was the earlier 4116 based version of the 5150? AFAIK, there were at least two versions of the 5150 motherboard, and no doubt, different versions too of floppy disk controller card. So assuming we are both right, IBM must have transitioned from the standard Shugart cabling to their contorted version during the life of the 5150. Interesting historical trivia, not sure how useful it is to the problem at hand. Joe.

Option 1. Try finding and re-installing the drivers for your SBC's chipset. Option 2. May be risky, but delete the floppy and its controller in Device Manager and rescan for "new" hardware. I don't think so. It was IBM that introduced the hair-brained (hare-brained?) "twisted floppy cable" scheme to the world. OTOH, I had a clone by a company called Eagle and that used a straight floppy disk cable, so it supported up to 4 floppy drives, yet the case could only fit two! BTW, I also used a 5150 at the time, and that definitely used a twisted floppy disk cable. Joe.

You both raise interesting points. I don't know how different are the 9X vs. NT registry structures, I thought the latter was a superset, in which case the 2K version of REG should be OK in most situations. As for using the 98 version of REG, alas, the only application where I've needed REG, it was the 2K version that was needed, as both the command line syntax and the output format of these two versions are incompatible, as demonstrated here : reg_98 query hkcu\software\7-zip\extraction\extractmode REG_DWORD extractmode 1 reg_2k query hkcu\software\7-zip\extraction /v extractmode ! REG.EXE VERSION 2.0 HKEY_CURRENT_USER\software\7-zip\extraction extractmode REG_DWORD 0x1 Joe.

Thanks Dave! I tracked down a copy of Driver Cleaner Pro 1.5 and followed the instructions in the ReadMe file (basically as above, with a trip via Safe Mode). It deleted a bunch of registry entries. When I re-installed the 4.00 drivers, there was a 'click' sound from the speakers. I'm pretty sure that was the first signs of life from them. Sound still didn't work, but after checking the MB manual, I realized there were two jumpers missing from a pin header labelled "F_AUDIO_I". That was the final piece of the puzzle and sound now works! Joe.

Another little candidate for the SP : REG.EXE : Extract with 7-Zip from 'sp4supporttools.exe' from : http://www.microsoft.com/download/en/confirmation.aspx?id=18614 Copy to %windir% I needed REG.EXE to install SAP GUI for Java. Although this example needs KernelEx as well, there may be other Java-based packages that don't need KernelEx but do need REG.EXE to do their registry stuff. Joe.

I know this is an old post but ... I have the same (?) situation with a GA-8IEX Rev 2 motherboard that I'm trying to get going. I initially tried the Realtek drivers from Gigabyte, which correspond to version 4.06, then found this post and tracked down a copy of the 4.00 drivers. So basically, I uninstalled the 4.06 drivers via Add/Remove Programs in Control Panel, then installed the 4.00 drivers. Now, the drivers are installed and report as OK, but there is no sound. So, did you need to do anything special to retrograde these drivers (or do I have faulty hardware)? Joe.

Confirmed! According to Wikipedia, the current stable version is 2.2.6000. I run that version and have had no problems. What version has problems with ActivateActCtx? Hi jumper, Mine says it's 2.2.6000 (MD5 of EXE = b6051fdab7dc811a2d6be64a1579c735, MD5 of DLL = e72eff1b793fe064f068e715efb1b5cd). Joe.

Another problem child from the update @ post #10 : ;ActivateActCtx=f2e // buggy! This one prevented Dependency Walker from running (unless I disabled KernelEx for it). Joe.

OK, I've found the problem with update @ post #10, it's : FindActCtxSectionStringW Now my 'stubs.ini' looks like : [#prefs#] StepLoad=0 ;32/64-bit return value codes: ; t:-1, f:0, z:0, o:1, p:parameter1 ;Stack pop count: ; 0:'C' function; 0..31:# of WINAPI parameters ;SetLastError value codes (only when needed): ; e0:0/ERROR_SUCCESS; e,e1:1/ERROR_INVALID_FUNCTION ;MSDN notes for return values: ; STATUS_SUCCESS: 0 ; S_OK: 0 ; INVALID_HANDLE_VALUE: -1 [Advapi32.dll] AddAccessAllowedAceEx=z5e ChangeServiceConfig2A=z3e ChangeServiceConfig2W=z3e GetTraceLoggerHandle=t1e QueryUsersOnEncryptedFile=o2 [Avicap32.dll] capCreateCaptureWindowW=f8 capGetDriverDescriptionW=f5 [Cygwin1.dll] setgroups _ctype_ [Kernel32.dll] ActivateActCtx=f2e CreateActCtxA=t1e CreateActCtxW=t1e DeactivateActCtx=f2e FindActCtxSectionStringA=f5e ;FindActCtxSectionStringW=f5e // buggy! GetGeoInfoA=z5e GetGeoInfoW=z5e GetUserGeoID=t1 GetVolumeNameForVolumeMountPointA=z3e GetVolumeNameForVolumeMountPointW=z3e HeapQueryInformation=f5e InitializeSListHead=f1 InterlockedFlushSList=f1 InterlockedPopEntrySList=f1 InterlockedPushEntrySList=f2 QueryDepthSList=f1 ReleaseActCtx=f1 [Msvcr9.dll] ??0bad_cast@@QAE@ABV0@@Z= ??0exception@@QAE@ABQBD@Z= ??0exception@@QAE@ABV0@@Z= ??0exception@@QAE@XZ= ??1bad_cast@@UAE@XZ= ??1exception@@UAE@XZ= ?what@exception@@UBEPBDXZ= mktime= [Msvcrt.dll] ___lc_codepage_func= ___lc_handle_func= ___mb_cur_max_func= __crtGetStringTypeW= __crtLCMapStringW= __iob_func= __pctype_func= __uncaught_exception= [Netapi32.dll] NetApiBufferFree=o1 NetFileGetInfo=o4 NetGetAnyDCName=o3 NetGetDCName=o3 NetGroupEnum=o7 NetGroupGetInfo=o4 NetGroupGetUsers=o8 NetLocalGroupAddMembers=o5 NetMessageBufferSend=o5 NetMessageNameAdd=o2 NetMessageNameDel=o2 NetMessageNameEnum=o7 NetServerEnum=o9 NetShareEnum=o7 NetShareGetInfo=o4 NetUseAdd=o4 NetUseEnum=o7 NetUseGetInfo=o4 NetUserDel=o2 NetUserEnum=o8 NetUserGetGroups=o7 NetUserGetInfo=o4 NetUserGetLocalGroups=o8 NetWkstaGetInfo=o3 [NTDLL.DLL] NtAllocateVirtualMemory=t6 NtCreateProfile NtOpenProcessToken NtQueryInformationProcess NtShutdownSystem RtlInitAnsiString RtlInitializeCriticalSection RtlInitUnicodeString RtlIpv6AddressToStringA RtlLeaveCriticalSection _stricmp sscanf=z0 strlen=z0 wcslen=z0 wcsncat=p0 [Ole32.dll] CoAllowSetForegroundWindow=t2 [Secur32.dll] GetUserNameExA=z3e GetUserNameExW=z3e [SHELL32.DLL] SHPathPrepareForWriteW=t4 StrChrA=f2 StrChrW=f2 [Shlwapi.dll] SHRegisterValidateTemplate=t2? Joe.

Thanks for the tips, jumper. I'll do some experimenting as soon as I can and report back. Joe.