UberGeek451

It's funny you mentioned this . I really wanted an X48 board but they are very hard to find. Instead, I bought the Intel DP45SG Extreme.... both are very similar gaming boards. Although I don't think drivers are the cause of my current installation problems, the USB driver issue could affect many W2K users. After installing W2K 6 weeks ago, USB was the first major problem I experienced. I had no USB keyboard or mouse and since the motherboard had no PS/2 ports, I was stuck at Windows login screen ! It took days to fix this problem but I'm not sure this was an Intel driver issue. The OS was from another ICH10 machine and before the image was copied, the pci bus was uninstalled so it would re-enumerate on the new machine. In this particular situation, the problem was caused by Windows Plug & Play process and its enumeration of USB and \HID\Vid_ device combinations. I was planning to post a seperate topic on msfn to determine the default enumeration behavior of these devices in W2K and XP. For years I have noticed this irratating behavior: a) when you boot a machine with a USB mouse or keyboard plugged into a different USB port they are not automatically recognized. B) In Windows the behavior is similar, you plug these devices into a NEW port, P&P kicks in, and you have to keep clicking yes, yes, yes...before they will work. In otherwords, these inf's are already located in C:\WINNT\inf, and the Mouse and Keyboard should install automatically, but they don't without user input. I assume the same thing would happen before a user login. Q: Do know what causes this particular issue and would your custom usb inf's fix the problem? BTW, I used Chipset inf v9.1.2.1008. I have never experienced any other USB problem with Intel's infs. Usually I run a wildcard search with a PCI\VEN string in my extracted chipset directory and select the latest drivers. As I recall, the last version where every W2K driver for the ICH10 series was found was in v9.1.2.1008. This is also what I'm using on my other ICH10 Desktop, a Q45, HP DC7900. POST ADDENDUM: I just noticed something very interesting. My other Desktop, the HP 7900 (not the problem PC), has a perfectly functioning W2K OS. Even though Windows USB drivers appear to be normal in Device Manager and elsewhere on the system, and has never caused major problems, their flaws may simply be hidden. Running a PCI Bus sniffer reveals a series of empty IO address spaces and unreported details on each of the 6 UHCI Controllers identified in Device Manager as ICH10 Family USB Universal Host Controller (3A64-3A69). Registry identifies oem6.inf (aka ICHAUSB.INF) version 9.0.0.1005 It looks like there's a bug in your modded USB drivers. I downloaded a copy of your Chipset inf from /wlu v9.4.2.1020+v9.4.4.1006+AHCI Bug Fixed version (infinst_autol6.cab) On a fresh W2K installation with Microsoft's native drivers, I ran the automatic installer and immediately lost USB2. Device manager showed 2 yellow bombs on both USB2 controllers, Intel® ICH10 Family USB Enhanced Host Controller – 3A3A, and 3A3C. Rebooting and reinstalling the driver didn't fix the problem. After updating and then selecting a different driver (Windows native Standard Enhanced PCI to USB Host Controller) USB2 was fixed. Please have a look at your ichausb.inf files. In the July 31, 2013 Version 9.1.9 Build 1006, under the section [iNTEL.NT.5.1] the 4 EHCI device drivers are missing a “Services” descriptor which is defined in each of the Intel defaults. The error code in Device manager references Device could not start. I've made some interesting discoveries about why W2K is causing problems on the DP45SG machine which I will be post separately.

Will do. Memory tested ok but this board is VERY picky about DDR3 SDRAMs. Ordered new sticks today by crossing to the exact Elpida chip ID. All hardware is new except for CPU. Motherboard was original Intel in sealed packaging with last revision (probably an Intel factory refurbish) The BSOD's only happened yesterday so I'm not too worried. I just reset the memory timings to default and disabled the sound tray drivers so I don't expect anymore crashes. I'll try running some more tests but this doesn't explain why XP installed on the same drive and has been running perfectly. In the meantime I have been downloading your cab files and will try to test your HFSLIP soon. Thanks

Hi Blackwingcat, I'm happy you responded. Unfortunately, there are no BIOS configurations on the Intel DP45SG motherboard for IRQ interrupts, video shadowing, etc. The only options are to disable onboard PCI devices which I already tried. I really liked the latest v11d update and never had a major problem until now. Yesterday I experienced several minidumps including Page_Fault_In_Nonpaged_Area and PFN_LIST_CORRUPT BSOD's. The RAM tests good and PAE settings don't change anything. This is an Intel Quad Core so I'm going to update the microcode with the KB936357 patch I found on your site. I also noticed during driver loading there was an unusual hang at acpitabl.dat which I have never seen. This looks like an IOCancelIrp error found in Server 2003. Do you know what this is? Last week I was able to force device detection by installing two PCIe video cards and adding the nvidia registry values by hand. I became very frustrated and decided to yank one of the cards in Windows. After loosing video, I rebooted into Windows where Plug & Play kicked in, detected the GT 610, loaded the drivers, enumerated the remaining registry settings, and I had perfect, high resolution video. Of course this didn't last and on the next reboot Windows vga.sys took over again. This looks like a bug in W2K's memory mapping or how it assigns resources to the PCIe bus. I agree about upgrading the kernel and looked at the recent postings on MrMaguire's thread but would prefer creating a new slipstreamed OS. After wasting so much time on this machine I'm not quite ready to jump into another project.... yet. Many Thanks

Hi, I'm hoping someone here can help me figure out this problem. Don't know why this is happening but after installing a fresh copy of W2K, my video card can not be detected on one particular machine. Device manager shows code 12 on the Intel® 4 Series Chipset PCI Express Root Port - 2E21. This port should be shared with the video card but instead it's locked to VgaSave with conflicting memory and I/O addresses. The machine has a P45 ICH10 motherboard with 2x2GB DDR3 RAM and a Core 2 Quad 2.8 GHZ CPU. Video card is a MSI GeForce GT 610 2GB 64-Bit DDR3 PCIe 2.0 x16. Installing an older NVIDIA Quadro NVS 290 did the same thing. Os is W2K Pro Sp4 with UURollup-v11-w20130323 (tomasz86’s HFSLIP program), which was updated to the latest daily, v11-d20141130. Same copy has been used to install W2K on other ICH10 & Core2 Duo machines without a problem. Windows XP Pro SP3 installed on the adjoining partition without a single hitch using BlackWingCat's nVIDIA 310.70 WQHL drivers. Tried everything but W2K has been fighting me for weeks. Appreciate any suggestions.

bluebolt Thanks for bringing this to my attention. I had read the post before and afterwards attempted three repairs based on this info, however on a second reading I saw some comments I may have completely misunderstood. The post began with a graphics card installation which I realized can sometimes cause audio issues. Therefore my first repair was to confirm the onboard video chip had no audio functionality and to make sure, I uninstalled each device and used the preferred reinstall sequence Chipsets - UAA - Audio - and lastly video. The second repair was to re-register the files by running regsvr32 commands on all the exe's and dll's. After reinstalling the sound drivers, the third repair attempt was to simply reinstall the most recent v11 daily (Jan 2014) which I have used and genuinely enjoyed for months. Unfortunately nothing changed. Now I'm confused.... It looks like part 3 should have been to uninstall Windows2000-UURollup-v11-d20140103-x86-ENU, migrate back to the older v11 weekly, reinstall and test for functional sound, then install v11 daily (Jan 2014). BTW, I posted this question as a topic in the main Windows 2000 forum: http://www.msfn.org/board/topic/172216-help-with-onboard-sound-failures-win2k-uur/ and don't want to be guilty of double posting. No answers as of yet but I did post a crude solution today which temporarily fixed the problem on one of the PC's..

I wanted to provide some updates and a temporary solution in case anyone else is experiencing the same problems. ISSUE: Onboard sound will work on XP but not Win2K UUR v.11. installed on the same system. Issue has been duplicated on 2 separate PC's. Every trick in the book wont bring it back to life. SYMPTOMS: SoundMax and Sigmatel drivers written for XP-W2K will install without a problem. - On reboot there is no sound and speaker icon is missing from the system tray. - Install logs and registry entries show driver installations were successful. - Device Manager lies with the statement "This device is working properly" HOWEVER - In the "Sounds and Multimedia" menu in Control Panel the sound volume slider is at 0 and can not be adjusted and on the Audio properties page, all 3 fields are greyed out. - The system file wdmaud.sys and Universal Audio Architecture (UAA) file sysaudio.sys are not running in Task Manager On the desktop machine, I added the /bootlog option to startup and noticed the SoundMax file aeaudio.sys was not loading (Andrea - Audio Noise Filtering Driver). I was finally able to get sound with an embarrassing hack by dropping an earlier version of the file into the drivers folder. Wdmaud.sys and sysaudio.sys are now running and everything is normal except the volume is about half of what it should be. I've given up on the Sigmatel laptop since it showed no driver issues on the bootup log and could be a real PITA. In some small way I hope this helps in finding a solution.

Can anyone provide help on troubleshooting onboard sound issues with UUR v11? I have 3 Windows multiboot systems on my network, a P4 laptop, a Core 2 Duo laptop, and a Core 2 Duo desktop. Each run XP, the latest Win2k UUR v11d, and other OS versions. The older P4 has sound but I can not get sound to work on either of the onboard sound cards in the Core 2 Duo machines, however, sound is functioning on both these machines in XP. The P4 doesn't use a High Definition Audio chipset or UAA and the sound works fine. The Core 2 Duo laptop uses SigmaTel High Definition Audio and was incrementally updated from UUR V10 to V11d. The Core 2 Duo desktop uses SoundMAX High Definition Audio and I recently used Tomasz86's HFSLIP program to install UUR V11 and then update to V11d. I've tried all the usual fixes but nothing seems to work. Both the SigmaTel and SoundMAX drivers support W2k & XP and both install without any noticeable errors but on reboot sound will not function.....no yellow bombs, setupapi.log errors, or other useful clues. The usual speaker icons are missing in both startup trays. After logon a popup window appears with the message "The SoundMAX audio driver did not load. You may need to reinstall SoundMAX". Later the SoundMAX tray icon is visable with a red cross and the message "No SoundMAX audio present". My only guess is that the Microsoft UAA driver or a dll is not compatible but all the version numbers and file sizes seem to match up. Everything else is working beautifully but this is driv'in me nuts. Windows 2000 has always been my favorite OS so any suggestions would be helpful.

Greetings and many thanks for the fine work here. Can anyone provide help on troubleshooting onboard sound issues with UUR v11? I have 3 Windows multiboot systems on my network, a P4 laptop, a Core 2 Duo laptop, and a Core 2 Duo desktop. Each run XP, the latest Win2k UUR v11d, and other OS versions. The older P4 has sound but I can not get sound to work on either of the onboard sound cards in Core 2 Duo machines, however, both work fine in XP. The P4 doesn't use a High Definition Audio chipset or UAA and the sound works fine. The Core 2 Duo laptop uses SigmaTel High Definition Audio and was incrementally updated from UUR V10 to V11d. The Core 2 Duo desktop uses SoundMAX High Definition Audio and I recently used Tomasz86's HFSLIP program to install UUR V11 and then update to V11d. I've tried all the usual fixes but nothing seems to work. Both the SigmaTel and SoundMAX drivers support W2k & XP and both install without any noticable errors but on reboot sound will not function.....no yellow bombs, setupapi.log errors, or other useful clues. The usual speaker icons are missing in both startup trays. After logon a popup window appears with the message "The SoundMAX audio driver did not load. You may need to reinstall SoundMAX". Later the SoundMAX tray icon is visable with a red cross and the message "No SoundMAX audio present". My only guess is that the UAA driver or a dll is not compatible but all the version numbers and file sizes seem to match up. Everything else is working beautifully but this is drivin me nuts. Windows 2000 has always been my favorite OS so any suggestions would be helpful.

Thanks for the reply BlackWingCat. It's not a huge issue but one that's got me really puzzled. I detest Vista and haven't run Linux, XP, or DOS on this laptop (different partitions and bootloaders) for many months, but something has started to call this XP function and I certainly can't find it. The error was believed to be the cause of reboot problems, a possible memory resident issue in VMware by a Ramboot XP forum user. It was subsequently answered by jaclaz (could this be the same jaclaz here at msfn.org?), providing some thoughtful solutions http://reboot.pro/topic/2231-ramboot-xp-reboot-on-acpitabldat/ ,including a link to issues with hard lockups following XP SP2 upgrades (Intel Prescott series, microcode issues, update.sys, etc) http://cquirke.mvps.org/sp2intel.htm Today, I ran a boot log and reviewed the driver loading but there was no record of the two "ghost files". Oddly they appear at the bottom of my screen on every boot into Safe Mode. They are also clearly related to increased boot delays in this specific version of UURollup v11 (other v11 was fine). You're doing some very excellent work and I want to see it continue, and to succeed. I've never had a single BSOD, lockup, or uninstall issue and my only regrets have been a long history of chronic Firefox problems. After some registry modifications I now have W2K running blistering fast which includes exceptional page loading and a big smile that is hard to erase. UberG

Sometime after recently upgrading a Toshiba P4 Laptop from UURollup-v11-w20130323 stable to UURollup-v11-d20130902, I began to notice excessively long bootup times, approximately 3-5 times longer than average. At the time, I was preoccupied with other system configurations and held off on any further troubleshooting. Today, during a visit to safe mode, I watched the drivers load, noticing the system began to hang immediately after agp440.sys, attempting to load two files which looked nothing like drivers.....shown below: C:\WINNT\AppPatch\drvmain.sdb C:\WINNT\System32\acpitabl.dat A quick search confirmed that neither of these two files existed on my system partition. Going a step further, I ran searches for all system files, containing these names, including registry without a single hit. WTF! A broader search in all partitions, on both the Laptop and Desktop, returned hits for only drvmain.sdb, located in XP system partitions under \WINDOWS\AppPatch\, \WINDOWS\system32\dllcache\ and various XP installation sources. Online sources indicated this problem is usually associated with XP or Server 2003 compatibility files and tables used by the OS during driver loading. My identical error is best described in the following Server blog: http://blogs.technet.com/b/askcore/archive/2011/08/19/system-hangs-at-splash-screen-or-boot-freezes-on-acpitabl-dat.aspx Is this a code issue, a file import failure, or just another Microsoft moment? UberG

I'm relieved there's someone else out there with this problem besides me! A lot of people could probably live with this aggravation but I use multiple partitions, hyperlinks, and write most of my documents in HTML so this was nearly driving me mad. BTW, BUG SOLVED!! Just figured this out an hour ago and I'm tickled pink. The cause of the problem was a bug in DDE shell integration and the ddeexec registry keys I mentioned in the previous post. DDE is a miserable attempt to shift control of various File Handlers (.htm, .html, .shtml, .xht, .xhtml) and Protocol Handlers (ftp, http, https) depending on the default browser (IE or Firefox) and various file associations. When Firefox dropped their support of Windows 2000, the software developers had an opportunity to remove some of the dysfunctional DDE dependencies without wrecking the browser for W2K users. UUrollup liberates W2K users to do some very cool stuff, and one of those benefits was upgrading to FF18. For many users, DDE perpetually coexists in a semi-corrupt state without ever being noticed. Unfortunately, after my FF 18 upgrade it became problematic once again. This issue has been a continual headache for Firefox developers since 2004 and after reading the following bug report I found a solution which worked.( https://bugzilla.mozilla.org/show_bug.cgi?id=491947 ) The fix involved using a script to delete a number of FF related registry keys and another to rebuild the configuration.If anyone is interested I'll post the specifics.

BTW, scratch the last post. It's relevant but probably not the source of errors. Here's the latest. I just clicked a shortcut (Firefox icon) to an html document located on another partition. Windows explorer threw the same error popup, "Cannot find the file 'K:\Web_Docs_Business\A&E_Project.html' (or one of its components). Make sure the path and filename are correct and that all required libraries are available" This time however, Internet Explorer which I almost never use, launched a browser in a second window returning a different error, complaining that it couldn't find my document in the Bing search engine. IE had munged the document name into a short file name with a .htm suffix. I took snapshots of both errors and confirmed an hour ago my registry values for Firefox file associations were identical to those in XP. I certainly never had these problems before but still don't have a single suggestion on what to do next. Fortunately it's only a minor headache so I'll let it rest for a while.

Sorry bout that, I sensed it may have hit a raw nerve. In the meantime I've been looking into several Firefox registry values which determine default browser behavior. They may or may not have been skewed which could be the source of my errors. Does anyone know if Firefox is still using the values in bold under either of these keys: HKEY_CLASSES_ROOT\FirefoxHTML\shell\open\command or HKEY_CLASSES_ROOT\FirefoxURL\shell\open\command with either "-osint -url" or "-requestPending -osint -url" inserted after the path Examples: "D:\W2Program Files\Mozilla Firefox\firefox.exe" -osint -url "%1" OR "D:\W2Program Files\Mozilla Firefox\firefox.exe" -requestPending -osint -url "%1" And also, any ddeexec keys or subkeys. Examples: HKEY_CLASSES_ROOT\FirefoxHTML\shell\open\ddeexec HKEY_CLASSES_ROOT\FirefoxHTML\shell\open\ddeexec\Application HKEY_CLASSES_ROOT\FirefoxHTML\shell\open\ddeexec\Topic Many Thanks

Thanks for the response. W2K SP5.0 is just an arbitrary name for my post SP4 version of Windows 2000. After upgrading for years, In May 2010 I decided to do a fresh install using nLite to slipstream my W2K SP3 disk with SP4, MS Rollup v2, recent hotfixes, and a variety of ad ons based on a detailed MSFN posting, hence SP5. Other than a few glitches after the initial installation, this OS has been running rock solid on both my machines. This is an unusual problem that has me really puzzled. Both machines exhibit the same problem so I'm certainly willing to experiment. Not sure if the next move should be another FF install, a downgrade, a different UURollup, play with the Application Compatibility Launcher, etc. Any suggestions would be appreciated.

Greetings to all, Last week I upgraded two machines running W2K SP5.0 with UURollup-v10d. Prior to installing the UURollup, the Image File Execution reg patch was applied and a copy of uxtheme.dll added to the Firefox program directory from BlackWingCat's_KDW-fcwin2k package as per the Firefox 13+ in Windows 2000 instructions posted on Mozillazine (http://forums.mozillazine.org/viewtopic.php?f=7&t=2482475). UURollup installed without a hitch and after a short test drive, Firefox 12 and its crippled Clingon cousins, Java and Adobe flash were uninstalled. After taking a deep breath, I downloaded Firefox 18 which installed perfectly, reserving the final acid test for the Clingon's. I was able to find the pair on BWC's website and downloaded the most current patched versions of Java and Adobe flash. After recently learning these patches were no longer unnecessary, my second machine was updated normally to Firefox 19.x with the most current versions of Flash and Java! These installations went flawlessly, providing a major upgrade that would make Bill Gates blush. I wanted to express my gratitude to the members of MSFN, particularly BlackWingCat, tomasz86, and others who contributed to this remarkable achievement. I use several operating systems but have always loved Windows 2000 the most. The loss of Firefox support and frequent "XP only" executable's have been about as welcome as a series of automatic transmission failures in the Bronx. Thanks for bringing a little sunshine to the winter of 2013. Now to address the current problem: When double clicking on an HTML file, Windows immediately throws an error box in the launching directory "Cannot find the file "X:\yadayada.html" (or one of its components). Make sure the path and filename are correct and that all required libraries are available" This of course is BS since Firefox has already launched the html file in the browser. Unfortunately, each one of these must be cleared by hitting ok, otherwise the blinking warnings accumulate in the taskbar. Oddly, this error can be avoided by the equivalent context menu action, Right-Click-Open-Open with-Firefox. My immediate hunch was to suspect a file association problem which could not be resolved using the conventional bag of tricks. This brought back unpleasant memories of 2004-06, when Firefox experienced a series of similar bugs associated with "who's boss, Firefox or Internet Explorer?" This began with URL hyperlinks, later mutating to include html documents. The culprit of the Firefox bug involved Microsoft's method of handling outside competition, a convoluted process called ddeexec to control file associations and their dependencies. As suspected, clicking a URL hyperlink, reproduced nearly identical errors, except that file http://www.yadayada.com listed instead. I spent the last two nights trying to force the OS to recognize the default browser, deleting Firefox's registry associations and rebuilding the stack several times. Although URL associations have been fixed, I've had no luck with changing html behavior. Is anyone else having problems like this? Did I do something wrong during installation, or is this a dll dependency like Shell32? Chaz

Good software, like good literature, lives on no matter what course the author has taken. As we can all see, nLite has taken on a life of its own. BTW, The driver problems were resolved and I now have a wireless connection on the laptop. The peanut butter trick is what finally fixed my problem. Immediately after posting this veiled threat of annihilation, my Intel wireless msi installers mysteriously began responding after several dozen odd downloads and three days of sheer hell! Gosh.... hope I don't have an ndis keylogger!

Hi jaclaz Thanks for the encouragement. Re: a.) Installation path was WINNT. All others behaved nicely as well. nLite properly redirected my Program Files and Users (Documents & Settings) to alternate partitions and did this very well, encoding the unique prefixes (W2Program, W2Document) so they're kept separate from the others. Best of all, the usual \common files\ path screwups from MS legacy apps like notepad, Outlook, etc. did not break the rules. BTW, I need to confess that I snagged the copy of WINNT.SIF generated by nLite and made a single edit before running my Unattend command. My dllcache is always redirected to another partition (to reduce OS bloat and fragmentation) by adding a variation of the following: [systemFileProtection] SFCDllCacheDir="G:\W2_WFP" SFCShowProgress=1 SFCQuota=FFFFFFFF Could this be the source of my WFP sfc issues? Never had any problems using this option before but also don't know the rules or inner workings of nLite either. Perhaps a trailing backslash or internal expand command??? I hope nuhi will consider this option and add it to his next build. It's very clear that a lot of nLite users enjoy a mean and clean OS. It's also a safe but portable option and there have been plenty of times when the directory was lost after a familiar Windows game, playing hide and go seek with the user during it's traditional logical drive letter reshuffling act. Another copy always appears, automatically re-created by the OS. Assuming the pointers to the new dllcache / i386 are correct, this might have distinct advantages, if I'm correct in assuming a single mouse click could perform a purge/delete/rebuild. Re: b.) Nope, not on an NTFS. Like the MoonGoon, I despise the msi for the sole reason that MS created an entire industry based on unforgivable confusion and complexity. Took me hours to learn the State secret of how to coax this creature out of silence and talk to me. Thanks to a Symantec post, I learned the "CLEVER ones" do permit a verboose mode called "squawkbox". Installing an msi produced some 600kb of "valley speak" with a single line code alluding to some error...so back on the trail to find the Holy Grail! Only information of merit indicated VM's or some cryptic nonsense about 32 bit processors using 64 bit masquerading and setting special reg keys for "registry reflection". This seems to happen continually on INTEL driver downloads from various sites, which either fail to install or self-extract to folders bearing the "E" suffix (ie:IA32E), the 64 bit masquerade. Intel made my processor, they wrote the software and they seem to be one of the few who bundle 32 & 64 bit code into single installers. A search of their site produced zilch but twice this week I've caught their msi's writing incorrect paths in my registry. When I get back to town I'll be on a mission to find my Atheros mini PCI. I wont toss the card but decided to cover it in peanut butter and feed it to the squirrel's instead. This should provide a moment of great pleasure to watch them go nuts instead of me! http://www.msfn.org/board/public/style_emoticons/default/tongue.gif ^^^^^^^^^^^^^^^ I love Win2K far more than the rest but I'm not out of the woods yet. Any tips on how to refresh the CLSID stack? I'm a bit over my head on this one and don't know vbs. Still got some broken COM+ junk (don't we all?)but never heard of any method to rebuild this. Figured a COM re-install would definitely do the trick but don't know if existing objects would be lost in the process. If the Regsrvr UnRegsrvr type commands can do this for dll's and exe's, then certainly there must be a registry listing, ini, or inf which would specifically target these keys using a batch file or script. If so, then file and path listings like those found in the 0.map key like the one listed below would look promising: HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Advanced INF Setup\mshtml.DllReg\RegBackup\0.map Thanks, Uber_G

After spending many hours, I finally managed to get a copy of Windows 2000 to install in unattend, using nLite to create the distribution. nLite would quickly produce working slipstreams but something in the addition would cause invalid data popups. The problem seemed to be resolved once the drivers were eliminated This package was then used to install fresh copies of the OS onto two functioning but separate machines running multiboot setups with multiple partitions. After some minor corrections and preferences, I was delighted by the results and invested a great deal of time perfecting the system. Unfortunately, over the next few days I noticed some troubling device glitches in DCOM/WMI. These were eventually traced to an improper Adobe Flash installation with the name FlashUtil10b.exe, a common addition to many nLite distro's. This was NOT an nLite error, or a virus, but something which demands equally serious evaluation since it will slowly destroy any operating system through a war of attrition. The Flash utility may incorrectly set it's HKEY_CLASSES_ROOT permissions with the "Deny Anonymous Logon" by including the "Deny Everyone", then going for the jugular by failing to limit this behavior to its own CLSID, wrecking a cascade of COM+ values on keys directly below. I had problems with this same issue a while back after loading a different version of this rotten little app in a much different setting. This was installed in "very attended" mode, by me, the administrative user, in an actively working system. Never figured that one out, but this time was different. There were a number of partial installations involved, with many details invisible to registry searches due to these permissions. After spending many hours on the mind numbing task of comparing CLSID's, rebuilding the registry, and re-enumerating various entries, I wouldn't touch this file with a ten foot pole. I can still recreate the issue and will try to post the details later if anyone is interested in a life insurance policy against death by a hundred "silent failures". ^^^^^^^^^^^^^^^^^^^^^^^^^^ Now for the questions: I now have a pair of delightful and very stable operating systems with a couple of lingering problems. a.) After the sfc /scannow command, the WFP appears to play nice but returns a string of event log notices indicating that Windows failed to find C:\WINNT\System32\ BUT NO FILE NAME! b.) Has anyone ever experienced a rare form of msi bungling where some installations either fail or are only extracted as 64 bit files. I've had a helluva a time with Intel Pro/Wireless drivers during the last three days. Hint: My other system (the one with a working internet connection) is an IBM 64 bit P4 although I have never touched a 64 bit application or loaded a single driver. Is the Web Nanny at Intel or Microsoft capable of failing me as well by incorrectly deciding that I need 64 bit drivers or other code. c.) Am I insane for going this far when most normal people would throw in the towel and start all over? During a visit to Microsoft I was convinced that my mind was failing after deciding on a second download, then noticing the following confession: kb/898628/ ) Windows Installer fails silently after you upgrade to Windows Installer 3.1 WTF??? Suggestions anyone?

In reply to Sp0iLedBrAt: Your question was one that puzzled me too. I ended up using my original SP3 CD, slipstreaming a copy of SP4, saving the distribution and moving in sequence to Kurt's SP5.1 Thank God for big fat hard drives!

I finally managed to straighten out this problem by changing to a standard nLite procedure. What a lovely little program! What a joy to see a fresh installation appear on my network. After fighting Windows for a week over a bungled COM+/Update/Security/Installer Failure/Crippled DCOM from a simple update, I was in a very surly mood, fully expecting to have to rebuild Framework 2 BTW, the W2K flat file unattend went perfectly, installing to the same drive and right next door to the partition containing my tattered OS without Windows raising a bloody tantrum!!! A bottle of wine later and the system is running stable, clean, and simple. Ooooh Lucky Day! ^^^^^^^^^^^^^^^^^^^^^^^^^^^^ I'm fairly sure my snag was the result of an errant driver install. What's the definitive word from the pro's on driver packaging? I was always under the impression that the following applied: inf's are a must cat's are a plus dll's are ok sys's are ok exe's are VERBOTTEN! ^^^^^^^^^^^^^^^^^^^^^^ This still doesn't solve the nagging audio, video, and modem drivers which run on executable's. Kurt's bat files got me gears a turnin, shedding some valuable insight into a more complicated multi-batch process I could never manipulate, Dell's Media Center installation on top of Win XP Pro. It may be far fetched but wouldn't it be sweet if links to executable's could be placed on run once registry keys or cmd files pointing to locations outside of the installation package (ie:like program installer's). Don't know about most people but I typically keep copies of the exe's from my favorite app's in specific directories, and we all have program files. Wouldn't this be the ultimate trick for a future nLite release! A link to an executable / copy / paste to file sequence...D:\XProgram Files\Java\jre6\bin\java.exe is just one example.

I thought I'd post here since this seems to be an unusual error that seemed to be resolved a while back. I'm creating an updated Win2k SP5 installation using using Kurt's posting "Up-to-date Windows 2000 install media creation guide". Although I have been using Unattend installs for several years I decided to give nLite a try. After spending three long evenings trying to get this procedure to work, every time I run my commands the setup fails with the error "Setup was unable to build the list of files to be copied. The data is invalid" Google points to a few obscure postings of earlier releases along with suggestions to change the command parameters which I have done with no success. Another tip was to switch to the same OS, but this didn't work either. Before I blow another night on this perhaps some of you might point me in the right direction. Here are the basics: a.) As I regressed back to a simple slipstream, the installation succeeded and began dumping files as expected. b.) No other variations of nLite seen to work, although I'm continuing to test. c.)It is necessary to run the unattend command since I'm running multiboot setups on several machines and MUST point Windows to the selected partition or wage another battle. I also prefer to run this from a HDD flat file, not a CD/DVD install. d.) I've checked the WINNT.SIF for errors and see none EXCEPT FOR the line: [GuiRunOnce] "%SystemRoot%\System32\nLite.cmd" Although nLite.cmd can not be found, deleting the entry in WINNT.SIF makes no difference either. Win2K command sequences are variations of the following: G:\WIN2K_SP4CD\I386\winnt32.exe /unattend:G:\WIN2K_SP4CD\I386\Winnt.sif /syspart:O: /tempdrive:O: I'm pulling my hair out trying to figure why this always fails and questioning my sanity for wasting so much time to coax a simple installation sequence that was much easier the old fashioned way. I must be doing something dreadfully wrong. Is it possible that I've overlooked the obvious, and the only way to get this this to work is from a CD/DVD?