All Activity

Me personally, I would use the Windows 10 USB/HID driver on the Windows 11 installation. No guarantee, but my hunch is that solves it. At bare minimum, can you at least tell us the driver DETAILS that are being used in each installation? This ( https://www.majorgeeks.com/files/details/drivergrabber.html ) should be able to "copy" the Win10 driver so that you can install it on Win11. You will (from memory, so I'm depending on you to do some of the legwork) have to edit an .ini or setting somewhere to make sure it copies "Microsoft" drivers.

I've asked the dev of Block Start Menu and Hosts Windhawk mod to include both CrossDeviceResume.exe and CrossDeviceService.exe in the mod.

Hi everyone, I’m hoping someone can help me diagnose a very strange Windows 11 HID/input issue. Hardware Motherboard: ASRock Z690 PG Riptide Logitech G502 HERO mouse Logitech keyboard Logitech webcam Operating Systems Tested Windows 11 (full installation): Problem occurs Windows 11 WinPE: No problem Windows 10: No problem I have also performed a completely clean installation of Windows 11, including installing it onto another drive, and the problem still occurs. Symptoms Windows boots normally. After Windows has been running for a while (not immediately after boot), the system suddenly behaves as though the Ctrl key is being held down. A few seconds later it transitions into repeatedly pressing the C key, and from that point it continuously types “C” until I intervene. Important observations This is not a physically stuck keyboard key. I have tested multiple keyboards and the exact same behavior occurs. Once the phantom C key begins, unplugging the physical keyboard does not stop it. Windows continues receiving phantom keyboard input. The only thing that immediately stops the behavior is uninstalling the HID Keyboard Device in Device Manager. However, after approximately 15–30 minutes, the problem eventually returns. HID devices Windows reports three HID Keyboard Devices. I checked the Hardware IDs for all three. All three report VID_046D (Logitech). The Hardware IDs are: Device 1 HID\VID_046D&PID_C336 Device 2 HID\VID_046D&PID_C336 Device 3 HID\VID_046D&PID_C08B So every HID keyboard-class device currently detected appears to belong to Logitech hardware. Additional information I do not have Logitech G HUB installed. The issue occurs using Microsoft’s default drivers. Windows 11 WinPE does not exhibit this behavior. Windows 10 on the same hardware did not exhibit this behavior. The issue survives a completely fresh Windows 11 installation. Current thoughts At this point I no longer suspect the physical keyboard itself because: I’ve tested multiple keyboards. Unplugging the keyboard does not stop the phantom input. Instead I’m wondering if: one of the Logitech devices is exposing a faulty HID keyboard interface, Windows 11 has a compatibility issue with one of the Logitech HID interfaces, or there is some USB/HID driver interaction specific to Windows 11. Has anyone seen a Logitech device generate phantom keyboard input like this under Windows 11? Is there a recommended way to determine exactly which HID interface is producing the Ctrl/C key events? Any suggestions would be greatly appreciated.

... OK, I found some related info over here ; but then, why would MS claim for "EventSetInfomation" to be a Win8+ function?

@panda_zzz Here is a screenshot including the log before freezing:

Because you shouldn't hit ENTER right after typing "qjs"; to get the help details for the app, type in the Windows Cmd Prompt "qjs -h" and then hit ENTER: QuickJS version 2026-06-04 usage: qjs [options] [file [args]] -h --help list options -e --eval EXPR evaluate EXPR -i --interactive go to interactive mode -m --module load as ES6 module (default=autodetect) --script load as ES6 script (default=autodetect) --strict force strict mode -I --include file include an additional file --std make 'std' and 'os' available to the loaded script -T --trace trace memory allocation -d --dump dump the memory usage stats --memory-limit n limit the memory usage to 'n' bytes (SI suffixes allowed) --stack-size n limit the stack size to 'n' bytes (SI suffixes allowed) --no-unhandled-rejection ignore unhandled promise rejections -s strip all the debug info --strip-source strip the source code -q --quit just instantiate the interpreter and quit (apologies for the accidental submission of an incomplete post)

According to Microsoft, this is a Win8+ specific function: https://learn.microsoft.com/en-us/windows/win32/api/evntprov/nf-evntprov-eventsetinformation#requirements Are you sure about your assessment?

I understand that perfectly. But you’ve already been given the code for my youtube.lua script. The ytsearch.lua script is still under development (the title of my post was, after all, "Development progress insights"), and I generally don’t make source code available at this stage. And it’s not yet clear whether I’ll be publishing any code at all at the moment. But that’s a different matter and has nothing to do with you. I will comment on this at a later time. Besides, the code wouldn’t be of any use to you anyway, as our two scripts are completely different. You’ve almost managed to populate the link list (unfortunately, the script still freezes when doing so), and passing the link to VLC fails then completely. You want to control a lot of things in the script that I don’t consider sensible and which I see as potentially triggering JS challenges, and the VLC settings – for subtitles, for example – are completely ignored in the process. You need to decide what your script is actually supposed to do and work through it bit by bit. Whenever one function is working reliably, implement the next one. Under no circumstances should you try to do everything at once. When I’m back at my computer, I’ll try to generate the log you’ve requested.

Still Mapped to 0xF0020000, but read is no longer 0xBDF7 but seems to be normal (to me) after forced install of ACPI and Q233017. It's all in the latest log's.

It might be that KB3080149 is needed to get EventSetInformation. But since this package adds "Diagnostics and Telemetry tracking service," I'm not going to install it. Yes it seems that QJS outputs colored text to ConEmu. Echoed input is green, and an error message is red. I don't know anything about this... some programs can do colored console text straight into Windows console. I like ConEmu because it is relatively small and runs on my computer. But I don't know a practical use for it yet.

very true that nothing lastst forever

It would be helpful to get a detailed log of exactly what's happening in your player. “C:\Program Files\VideoLAN\VLC\vlc.exe” --verbose=3 --file-logging --logfile=“C:\path\to\your\vlc-log.txt” The video is certainly interesting, but the source code is even more interesting for direct comparison and testing.

This looks like raw ANSI escape sequences that the terminal didn't process but instead displayed as plain text. It seems the problem isn't with parsing the input, but rather that the new qjs has started outputting VT/ANSI codes for line editing and highlighting, and the console doesn't understand them. You can check this by running it through an ANSI-compatible terminal or shell: the old ConEmu, ANSICON, mintty, Cygwin, or MSYS.

One of the sites with the list of updates is: https://www.hpcfactor.com/support/cesd/200240/windows_nt_40_workstation_patches_updates_guide/ But the one that would help which I also do not have would be Q326643 (Ntfs.sys hotfix) that was listed on Erpman's website back when it still existed... I don't believe NT4.0 gains NTFS v3.1 access by any of updates, it just might not mess it up with that one. You could try disabling NTFS service (Start=4 in registry) and other services that are responsible for it but I don't know.

The cause of this 0xBDF7 error seems to be that Windows is giving a completely incorrect memory mapping for the HD Audio registers. When things work it is mapping the controller registers at 0xFBEF8000 physical, after disable and re-enable the memory is mapped to 0xD0020000 physical and the specification version number, format capabilities and STATESTS all read back as 0xBDF7. Maybe Windows 98 First Edition doesn't properly release the device memory resources when unloading WDM drivers. I do unmap the PCI BAR space in the AdapterCommon destructor by calling MmUnmapIoSpace(m_pHDARegisters, memLength); but maybe there is something else that needs to be done to unload things properly? In the next version, the driver should bail out earlier if the HDA Specification version number reads back as anything other than 1.0 which would indicate that the mapped memory is not the correct registers. I haven't been able to test extensively in ACPI Mode as Win98 QuickInstall doesn't enable ACPI and 98's ACPI 1.0 implementation doesn't seem to work at all on real hardware out of its normal compatibility range. Especially laptops. Regarding VDMSound, the original Alpha-3 version posted on VOGONS works as well with my audio driver as VDMSound 2.0.4 ever did. I was able to compile the EmuAdlib.dll and EmuSoundBlaster.dll from VDMSound 2.10 and those still work with the old kernel driver. I had to add one line to the config file that was listed in the debug logs to make Adlib use OPL3. Now Doom doesn't crash at least but it also doesn't play sound effects. Still lots to do on that front because I'm not sure how to package things with the original installer and the VDMSound Core part doesn't build yet (the WinHelp compiler crashes and then the whole build is stuck). I'll put a package up here when I manage to get it working better. https://github.com/andrew-hoffman/vdmsound/ EmuAdLib.dll EmuSBCompat.dll

Are you sure you have version from vladimir-andreevich ? https://github.com/vladimir-andreevich/node.js-windows-7/tree/main/v20 Take the zip file for your architecture and extract it to e.g. C:\Bin\Node and then put that folder in your PATH.

Maybe I need some update for running Node 20. The procedure entry point EventSetInformation could not be located in the dynamic link library ADVAPI32.dll.

I had a long conversation with Google AI on this, could possibly be noise in combination with timing/ chipset/ WIN98FE issues. The probabilities in the statistics of the LLM predicted things will become better if I used ACPI on WIN98FE. After a longer process this was really the case for Wave. Only MIDI is not working anymore after Disabling/ Enabling hda.sys. Interesting are the Read-outs of WPCREDIT ('Idea' from Google AI). Some comments are above my level, maybe you can use it. I also tried 'Low' DMA (set to max 16MB), but no difference. More clear for me now. This would be VERY nice, in the past I tried VDMSound with Watler's HDA2.DLL, but Prince of Persia 1 never played me Midi. Not from your driver, or do have you plans to rename to hda2.dll? This was an inactive registry entry mentioned in DxDiag report: 20 43 61 72 64 20 6E 61 6D 65 3A 20 B0 03 0D 0A 20 20 20 20 44 72 69 76 65 72 3A 20 48 44 41 32 2E 64 6C 6C 0D 0A 0D 0A Card name: ° Driver: HDA2.dll I removed the inactive entry before current tests. WIN98FE_ACPI_DX7.0_DUT_WDMHDA_021.zip Google AI on STATESTS and ACPI with WDMHDA.021 on WIN98FE.ZIP BTW conversation is mainly in Dutch, I asked my AI-Agent to translate to American English. Dutch readers can (try to) read both versions. If someone cannot read md-files, just use Total Commander (No-nag-shareware) with the md-Lister Plugin. Note on 'Google AI': still now and then hallucinating, but if asked if 'something' is 'true' or so, the LLM is correcting earlier statements. Seems things are becoming a little bit better.

Agreed, reminds me of updating the Windows XP and 7 update lists back in the day for both this forum and JCarle's WUD website as he shared a exe file to enable me to upload the update lists to his website. Sad that he's no longer active on this forum and that his website is no longer available, but nothing lasts forever unfortunately.

No idea. When I run this version on Windows 7 (with all Server 2008 R2 ESU updates), I get: qjs -h QuickJS version 2026-06-04 usage: qjs [options] [file [args]] -h --help list options -e --eval EXPR evaluate EXPR -i --interactive go to interactive mode -m --module load as ES6 module (default=autodetect) --script load as ES6 script (default=autodetect) --strict force strict mode -I --include file include an additional file --std make 'std' and 'os' available to the loaded script -T --trace trace memory allocation -d --dump dump the memory usage stats --memory-limit n limit the memory usage to 'n' bytes (SI suffixes allowed) --stack-size n limit the stack size to 'n' bytes (SI suffixes allowed) --no-unhandled-rejection ignore unhandled promise rejections -s strip all the debug info --strip-source strip the source code -q --quit just instantiate the interpreter and quit On Windows 7, I use Node 20.19.2 from here as js interpreter. Much faster. Node 20 is not supported by yt-dlp any more since this commit, but in my build I reverted that change. [debug] Command-line config: ['-v', '--js-runtimes', 'node'] [debug] Encodings: locale cp1252, fs utf-8, pref cp1252, out cp1252 (No VT), error cp1252 (No VT), screen cp1252 (No VT) [debug] yt-dlp version nicolaasjan/yt-dlp@2026.06.22.040728 (win7_exe*) [debug] Python 3.14.6 (CPython AMD64 64bit) - Windows-7-6.1.7601-SP1 (OpenSSL 3.6.3 9 Jun 2026) [debug] exe versions: ffmpeg 8.1.2-WIN7 (fdk,setts), ffprobe 8.1.2-WIN7, phantomjs 2.5.0 [debug] Optional libraries: Cryptodome-3.23.0, brotli-1.2.0, certifi-2026.06.17, curl_cffi-0.15.0, mutagen-1.47.0, requests-2.34.2, sqlite3-3.50.4, urllib3-2.7.0, websockets-16.0, yt_dlp_ejs-0.8.0 [debug] JS runtimes: node-20.19.2 [debug] Proxy map: {} [debug] Request Handlers: urllib, requests, websockets, curl_cffi [debug] Plugin directories: none [debug] Loaded 1739 extractors

Nice to see this program still has users it brings back memories

The message to log in to prove I'm not a bot went away by itself later. QuickJS seems to take 10 seconds to do its thing on my computer. When I run it in command prompt on Server 2008 R2, why do I see garbage characters after typing something, like "\h"? t:\working\qjs>qjs QuickJS - Type "\h" for help qjs > ←[32;1m\←[0m←[J←[D←[32;1m\←[0m←[32;1mh←[0m←[J

Development progress insights: The ytsearch.lua script by AstroSkipper One week ago, I have created completely from scratch my own YouTube Search script I named ytsearch.lua. I followed exactly what I stated here: I did what I said. Only searching for YouTube videos and when clicking on it, the rest is up to my already existing youtube.lua script. It runs fine here under Windows XP. @panda_zzz My ytsearch.lua script is very slim, it has only one-sixth the file size of your youtube_search.lua script. And no temporary batch files, and no JS challenges here, of course. Here's a short clip: You just need to focus on the essentials, and then things will work out much better. Brevity is the soul of wit, especially when it comes to coding.

https://github.com/LordOfMice/Tools/blob/master/gpttsd.zip As GPTTSD is mostly generic TSD with 64-bit LBA support: User-defined partitions mount added by request from RayeR.

@panda_zzz Whenever I provide something I’ve programmed myself, I’ve tested it thoroughly. So I don’t understand what you’re doing here. Apart from that, your script – which was already complex – has grown by another 50% and become even more complex. My general advice: less is often more. If you’re wondering what’s so bad about complexity, the answer is quite simple: it makes the code prone to errors. And that, in turn, makes it much harder to fix things quickly and easily. You’re using far too many variables, and the code is no longer easy to follow. That aside, your script ignores the subtitle settings in VLC. The trick is to communicate with VLC and query specific internal VLC variables. Another important thing. You should always test your own toggle buttons to ensure they work properly. And if you still think that a temporary batch file is the only solution, then I’m afraid I’ll have to disappoint you. You can definitely manage without one. Furthermore, your script, if it ever does work, will most likely run into YouTube’s JS wall and have to solve JS challenges. However, I don’t see any integration of an external JavaScript module such as QuickJS inside your script. Personally, I avoid any conflicts with YouTube’s JS wall, as it significantly delays the start of the stream. Anyway! That’s just my assessment, of course.