All Activity

Definitely! I can't get that video to play in Chrome v144, Chromium v144, or Edge v149. Finding it very funny that we have found something that works in a Firefox Fork but does not work in Chromium.

Atleast give me the info of the models that you have.

Sorry, but if you publish a driver, e.g. on github, you should first test it on at least several computers. I'm not interested in your Asus x509FA laptop. I would like to use I2C drivers on other computers.

With what hardware? I only have ASUS x509FA. I can't afford another device with I2C.

@reboot12 Enable kernel memory dump from XP's registry (offline). Boot with i2cctrl set from manual to disabled, enable and disable from device manager again after service is enabled through serviwin/etc. Now on BSOD you should have a prompt for MEMORY.DMP.

@UsefulAGKHelper I'm sorry, but I have to wrote your attention: do not quote the entire post, select only important lines using e.g. ... ... do not quote entire logs or other things, e.g. what AI wrote - select important lines from them test your drivers before posting public if you need to post the entire log, use sites designed for this purpose, e.g. pastebin.com

Hello @Reino! Is there already a new address for your great releases? 🤔

This is what Copilot said about your BSOD: That blue screen you’re seeing — PAGE_FAULT_IN_NONPAGED_AREA (STOP 0x00000050) — means a kernel‑mode driver or system component tried to access invalid memory. In your context, since you’re testing your I²C controller driver, this BSOD almost certainly originates from a bad pointer or MMIO access inside your backend (for example, dereferencing devctx->MmioBase before mapping or writing to an invalid offset). Here’s how to narrow it down: 🧩 Likely causes Invalid MMIO base — MmioBase not yet mapped when IntelReadReg or IntelWriteReg runs. Null devctx — a failed allocation or uninitialized FDO context. Incorrect BAR offset — writing to a register that doesn’t exist on the ES controller (e.g., LPSS‑only offsets). Accessing pageable memory at IRQL ≥ DISPATCH_LEVEL — common if ISR/DPC touches nonpaged data incorrectly. Removing the code will get you directly back to error 31. The backend is still hardware specific...

Thanks for your releases and for adding ffplay! Please keep it alive! ffplay is a great media player , and my next release of ytBATCH for Windows XP will use it as I implemented a MediaBrowser in the form of a BAT file. BTW, do you also provide static urls to your XP releases?

@UsefulAGKHelper Your latest i2cctrl driver install OK but after about some seconds BSOD 50 appears:

Up until now, there has been no reason to use quickjs in combination with yt-dlp.exe. I use yt-dlp.exe in many places. Primarily, of course, in ytBATCH for Windows XP, in VLC with Lua scripts, in 3D YouTube Downloader and so on. In what situations do you need quickjs?

@reboot12 https://github.com/UsefulVideos/I2C-Controller-XP/releases/tag/build-20260621-111658 Try this compiled version. The change is only applied in i2cctrl.sys. Remember that the backend is still hardware specific... Will be configured later.

Implementing: /* Intel Serial IO I2C ES (Atom/Cherry Trail/Braswell) */ { L"ACPI\\808622C1", 0x00,0x04,0x08,0x0C, 0,0,0,0, QUIRK_ACPI20, BSOD_NONE }, { L"ACPI\\808622C2", 0x00,0x04,0x08,0x0C, 0,0,0,0, QUIRK_ACPI20, BSOD_NONE }, { L"ACPI\\808622C3", 0x00,0x04,0x08,0x0C, 0,0,0,0, QUIRK_ACPI20, BSOD_NONE }, { L"ACPI\\808622C4", 0x00,0x04,0x08,0x0C, 0,0,0,0, QUIRK_ACPI20, BSOD_NONE }, { L"ACPI\\808622C5", 0x00,0x04,0x08,0x0C, 0,0,0,0, QUIRK_ACPI20, BSOD_NONE }, { L"ACPI\\808622C6", 0x00,0x04,0x08,0x0C, 0,0,0,0, QUIRK_ACPI20, BSOD_NONE }, { L"ACPI\\808622C7", 0x00,0x04,0x08,0x0C, 0,0,0,0, QUIRK_ACPI20, BSOD_NONE }, { L"ACPI\\808622C8", 0x00,0x04,0x08,0x0C, 0,0,0,0, QUIRK_ACPI20, BSOD_NONE },

// // Table of supported controllers (explicit ACPI/PCI devices) // Match against full HWID substrings only // const I2CCTRL_DEVICE_ID g_I2cControllers[] = { /* ACPI-based controllers (NO LPSS) */ { L"ACPI\\INT3446", 0x00,0x04,0x08,0x0C, /* BAR0 DW-I2C */ 0,0,0,0, /* LPSS BAR2 */ QUIRK_ACPI20, BSOD_NONE }, { L"ACPI\\INT3447", 0x00,0x04,0x08,0x0C, 0,0,0,0, QUIRK_ACPI20, BSOD_FORCE_PIO }, { L"ACPI\\AMD0010", 0x00,0x04,0x08,0x0C, 0,0,0,0, QUIRK_ACPI20, BSOD_NONE }, { L"ACPI\\AMDI0010", 0x00,0x04,0x08,0x0C, 0,0,0,0, QUIRK_ACPI20, BSOD_NONE }, /* PCI-based Intel Serial IO controllers WITH LPSS BAR2 */ { L"PCI\\VEN_8086&DEV_9DC5", 0x00,0x04,0x08,0x0C, /* BAR0 DW-I2C */ 0x200,0x204,0x208,0x20C, /* LPSS BAR2 */ QUIRK_NEEDS_RESET_WORKAROUND, BSOD_EXTRA_RESET }, { L"PCI\\VEN_8086&DEV_9DE8", 0x10,0x14,0x18,0x1C, 0x200,0x204,0x208,0x20C, QUIRK_BROKEN_CLOCK_GATE, BSOD_MASK_INTERRUPTS }, { L"PCI\\VEN_8086&DEV_9DE9", 0x20,0x24,0x28,0x2C, 0x200,0x204,0x208,0x20C, QUIRK_NO_DMA_SUPPORT, BSOD_FORCE_PIO }, { L"PCI\\VEN_8086&DEV_9DEA", 0x00,0x04,0x08,0x0C, 0x200,0x204,0x208,0x20C, QUIRK_SLOW_CLOCK, BSOD_NONE }, { L"PCI\\VEN_8086&DEV_9DEB", 0x00,0x04,0x08,0x0C, 0x200,0x204,0x208,0x20C, QUIRK_NO_D1D2, BSOD_NONE }, /* Legacy PCI controllers requiring ACPI 1.0b fallback (NO LPSS) */ { L"PCI\\VEN_8086&DEV_A160", 0x00,0x04,0x08,0x0C, 0,0,0,0, QUIRK_ACPI10, BSOD_DELAY_INIT }, { L"PCI\\VEN_8086&DEV_A161", 0x00,0x04,0x08,0x0C, 0,0,0,0, QUIRK_ACPI10, BSOD_DELAY_INIT }, { L"PCI\\VEN_8086&DEV_A162", 0x00,0x04,0x08,0x0C, 0,0,0,0, QUIRK_ACPI10, BSOD_DELAY_INIT }, { L"PCI\\VEN_8086&DEV_A163", 0x00,0x04,0x08,0x0C, 0,0,0,0, QUIRK_ACPI10, BSOD_DELAY_INIT }, /* Other Intel PCI controllers (NO LPSS unless proven otherwise) */ { L"PCI\\VEN_8086&DEV_A2E0", 0x00,0x04,0x08,0x0C, 0,0,0,0, QUIRK_NONE, BSOD_NONE }, { L"PCI\\VEN_8086&DEV_A2E1", 0x00,0x04,0x08,0x0C, 0,0,0,0, QUIRK_NONE, BSOD_NONE }, { L"PCI\\VEN_8086&DEV_A2E2", 0x00,0x04,0x08,0x0C, 0,0,0,0, QUIRK_NONE, BSOD_NONE }, { L"PCI\\VEN_8086&DEV_A2E3", 0x00,0x04,0x08,0x0C, 0,0,0,0, QUIRK_NONE, BSOD_NONE }, { L"PCI\\VEN_8086&DEV_31AC", 0x00,0x04,0x08,0x0C, 0,0,0,0, QUIRK_NONE, BSOD_NONE }, { L"PCI\\VEN_8086&DEV_31AE", 0x00,0x04,0x08,0x0C, 0,0,0,0, QUIRK_NONE, BSOD_NONE }, { L"PCI\\VEN_8086&DEV_31B0", 0x00,0x04,0x08,0x0C, 0,0,0,0, QUIRK_NONE, BSOD_NONE }, { L"PCI\\VEN_8086&DEV_31B2", 0x00,0x04,0x08,0x0C, 0,0,0,0, QUIRK_NONE, BSOD_NONE }, { L"PCI\\VEN_8086&DEV_31B4", 0x00,0x04,0x08,0x0C, 0,0,0,0, QUIRK_NONE, BSOD_NONE }, { L"PCI\\VEN_8086&DEV_31B6", 0x00,0x04,0x08,0x0C, 0,0,0,0, QUIRK_NONE, BSOD_NONE }, { L"PCI\\VEN_8086&DEV_31B8", 0x00,0x04,0x08,0x0C, 0,0,0,0, QUIRK_NONE, BSOD_NONE }, { L"PCI\\VEN_8086&DEV_31BA", 0x00,0x04,0x08,0x0C, 0,0,0,0, QUIRK_NONE, BSOD_NONE }, { L"PCI\\VEN_8086&DEV_A368", 0x00,0x04,0x08,0x0C, 0,0,0,0, QUIRK_NONE, BSOD_NONE }, { L"PCI\\VEN_8086&DEV_A369", 0x00,0x04,0x08,0x0C, 0,0,0,0, QUIRK_NONE, BSOD_NONE }, { L"PCI\\VEN_8086&DEV_A36A", 0x00,0x04,0x08,0x0C, 0,0,0,0, QUIRK_NONE, BSOD_NONE }, { L"PCI\\VEN_8086&DEV_A36B", 0x00,0x04,0x08,0x0C, 0,0,0,0, QUIRK_NONE, BSOD_NONE }, { L"PCI\\VEN_8086&DEV_02E8", 0x00,0x04,0x08,0x0C, 0,0,0,0, QUIRK_NONE, BSOD_NONE }, { L"PCI\\VEN_8086&DEV_02E9", 0x00,0x04,0x08,0x0C, 0,0,0,0, QUIRK_NONE, BSOD_NONE }, { L"PCI\\VEN_8086&DEV_02EA", 0x00,0x04,0x08,0x0C, 0,0,0,0, QUIRK_NONE, BSOD_NONE }, { L"PCI\\VEN_8086&DEV_02EB", 0x00,0x04,0x08,0x0C, 0,0,0,0, QUIRK_NONE, BSOD_NONE }, { L"PCI\\VEN_8086&DEV_02C5", 0x00,0x04,0x08,0x0C, 0,0,0,0, QUIRK_NONE, BSOD_NONE }, { L"PCI\\VEN_8086&DEV_02C6", 0x00,0x04,0x08,0x0C, 0,0,0,0, QUIRK_NONE, BSOD_NONE }, { L"PCI\\VEN_8086&DEV_06E8", 0x00,0x04,0x08,0x0C, 0,0,0,0, QUIRK_NONE, BSOD_NONE }, { L"PCI\\VEN_8086&DEV_06E9", 0x00,0x04,0x08,0x0C, 0,0,0,0, QUIRK_NONE, BSOD_NONE }, { L"PCI\\VEN_8086&DEV_06EA", 0x00,0x04,0x08,0x0C, 0,0,0,0, QUIRK_NONE, BSOD_NONE }, { L"PCI\\VEN_8086&DEV_06EB", 0x00,0x04,0x08,0x0C, 0,0,0,0, QUIRK_NONE, BSOD_NONE } }; const ULONG g_I2cControllersCount = sizeof(g_I2cControllers) / sizeof(g_I2cControllers[0]); This is the list of supported controllers in i2cctrl.c.

uartctrl.log [21/06/2026, 09:13 AM] AddDevice: PDO=FFFFFADFD0B5E060 [21/06/2026, 09:13 AM] AddDevice: checking HWID 'ACPI\8086228A' [21/06/2026, 09:13 AM] AddDevice: checking HWID '*8086228A' [21/06/2026, 09:13 AM] AddDevice: no supported HWID matched i2cctrl.log [21/06/2026, 09:11 AM] DriverEntry begin [21/06/2026, 09:11 AM] DriverEntry complete (mask=0x00000009) [21/06/2026, 09:11 AM] AddDevice: begin [21/06/2026, 09:11 AM] AddDevice: scanning HWIDs for controller match [21/06/2026, 09:11 AM] AddDevice: HWID candidate: ACPI\808622C1 [21/06/2026, 09:11 AM] AddDevice: HWID candidate: *808622C1 [21/06/2026, 09:11 AM] AddDevice: unsupported controller, skipping [21/06/2026, 09:21 AM] DriverEntry begin [21/06/2026, 09:21 AM] DriverEntry complete (mask=0x00000009) [21/06/2026, 09:21 AM] AddDevice: begin [21/06/2026, 09:21 AM] AddDevice: scanning HWIDs for controller match [21/06/2026, 09:21 AM] AddDevice: HWID candidate: ACPI\808622C1 [21/06/2026, 09:21 AM] AddDevice: HWID candidate: *808622C1 [21/06/2026, 09:21 AM] AddDevice: unsupported controller, skipping

did you check the i2cctrl.log file too from windows\system32? It has more details

@UsefulAGKHelper I tried your drivers - all using force because no compatible ID's and only GPIO install OK. I2C or UART - errors Code 31 (cannot load drivers...) setupapi.log #I163 Device not started: Device has problem: 0x1f: CM_PROB_FAILED_ADD.

https://github.com/UsefulVideos/I2C-Controller-XP Compile with WinDDK 7.1.0. 7600.16385.1 from I2C-Controller-XP (using cd to I2C-Controller-XP in WinDDK XP x64 build env) directory. Remember that since I don't use WinDBG (single hardware), the log is shown at \Windows\System32\i2cctrl.log My driver tries to do the job (it assumes that the controller is powered up, out of D3 cold mode) but can't because it's a firmware thing that disables certain ACPI tables based on OS version (OSI). It's not even close to being "finished", and although it has universal options, it was more done for my hardware. Things like error code 10 will never happen because it always returns status success (doesn't try to force to enable something that isn't powered on until it's powered on by ACPI tables first). Can you figure out how to power the controller out of the D3 cold state using Clover first? If isn't out of this powered off state, I can't forward to ACPI child enumeration of ACPI/PNP0C50 (that's a very important step to reach a I2C device like the touchpad). Remember that it was mostly intented to try to use I2C touchpads on XP than anything (it does have features of a I2C controller though), and that the mechanism is done from WDM instead of KMDF. FYI, this is how the logging works on the driver: /* ----------------------------------------------------------------------- * kernel logger with printf-style formatting + timestamp prefix * ----------------------------------------------------------------------- */ VOID I2cCtrl_Log( PCSTR Format, ... ) { CHAR buffer[512]; CHAR final[600]; va_list args; NTSTATUS status; UNICODE_STRING path; OBJECT_ATTRIBUTES oa; IO_STATUS_BLOCK iosb; HANDLE hFile; LARGE_INTEGER sysTime, localTime; TIME_FIELDS tf; PAGED_CODE(); // // Hard safety guards: prevent use-after-free crashes // if (KeGetCurrentIrql() != PASSIVE_LEVEL) { return; } if (Format == NULL) { return; } // // SAFE pointer formatting: // Convert all %p to 0x%I64X BEFORE calling VPrintf. // This prevents the CRT from dereferencing freed pointers. // { CHAR safeFmt[256]; SIZE_T i = 0, j = 0; while (Format[i] != '\0' && j < sizeof(safeFmt) - 1) { if (Format[i] == '%' && Format[i+1] == 'p') { safeFmt[j++] = '0'; safeFmt[j++] = 'x'; safeFmt[j++] = '%'; safeFmt[j++] = 'I'; safeFmt[j++] = '6'; safeFmt[j++] = '4'; safeFmt[j++] = 'X'; i += 2; continue; } safeFmt[j++] = Format[i++]; } safeFmt[j] = '\0'; va_start(args, Format); status = RtlStringCbVPrintfA(buffer, sizeof(buffer), safeFmt, args); va_end(args); if (!NT_SUCCESS(status)) { return; } } /* Get local time */ KeQuerySystemTime(&sysTime); ExSystemTimeToLocalTime(&sysTime, &localTime); RtlTimeToTimeFields(&localTime, &tf); /* Format timestamp prefix: [DD/MM/YYYY, HH:MM AM/PM] */ { CHAR ts[64]; ULONG hour = tf.Hour; BOOLEAN pm = FALSE; if (hour == 0) { hour = 12; } else if (hour == 12) { pm = TRUE; } else if (hour > 12) { hour -= 12; pm = TRUE; } RtlStringCbPrintfA( ts, sizeof(ts), "[%02u/%02u/%04u, %02u:%02u %s] ", tf.Day, tf.Month, tf.Year, hour, tf.Minute, pm ? "PM" : "AM" ); RtlStringCbPrintfA( final, sizeof(final), "%s%s", ts, buffer ); } /* Open log file */ RtlInitUnicodeString(&path, L"\\SystemRoot\\System32\\i2cctrl.log"); InitializeObjectAttributes( &oa, &path, OBJ_CASE_INSENSITIVE | OBJ_KERNEL_HANDLE, NULL, NULL ); status = ZwCreateFile( &hFile, FILE_APPEND_DATA | SYNCHRONIZE, &oa, &iosb, NULL, FILE_ATTRIBUTE_NORMAL, 0, FILE_OPEN_IF, FILE_SYNCHRONOUS_IO_NONALERT, NULL, 0 ); if (!NT_SUCCESS(status)) { return; } /* Write timestamped line */ ZwWriteFile( hFile, NULL, NULL, NULL, &iosb, final, (ULONG)strlen(final), NULL, NULL ); ZwClose(hFile); // // Mirror to ETW/WPP without the timestamp prefix. // TraceEvents( TRACE_LEVEL_INFORMATION, TRACE_FLAG_BUS, "%s", buffer ); }

It was very difficult for me to abandon XP 'cause I had customized it so much -another reason why I'm so happy most of the software that worked on XP is now working on Windows 11- and I'd been using it since it was released. It was a tough decision. I also admit that during these past couple of months I struggled quite a bit to adapt to the new OS. Now, seeing all the benefits I've gained, I'm satisfied. A conversation with a friend was decisive. He uses Windows at work but Linux in his personal life. He thought I should try Windows 10 or 11 'cause learning Linux at this point would be very tedious for me. In that same conversation, he told me that Windows 11 is Windows 10 in disguise and I verified this myself. I thought Windows 10 was lighter but comparing these X-Lite versions which have no extra cheese, Windows 10 and Windows 11 consume exactly the same resources. Only differences are cosmetic improvements.

Many thanks for your tests and confirmation ; with some more free time on Saturday, I decided to explore how ALL stable VLC versions after 3.0.8 behave when fed the above audio HLS live stream; one can find older stable VLC releases in their "download" server: https://download.videolan.org/pub/videolan/vlc/ FTR, v3.0.7.1 (immediately previous to 3.0.8) behaves almost identically to 3.0.8; HLS streams (at least the demuxing part) are handled by the libadaptive_plugin.dll module; here's the CPU usage history of my Core2 Duo with ONLY VLC v3.0.8 streaming BBC R2: libadaptive_plugin.dll versions 3.0.9.2 and 3.0.10 are actually BROKEN when it comes to handling the BBC R2 HLS stream; I can hear the audio for ca. 30-40s, but then the stream turns muted (continues to play otherwise) ... From v3.0.11 onwards, the BBC R2 HLS audio stream does play consistently, but at elevated CPU consumption ; with each newer VLC release, the file size of the above DLL increases, but the CPU it uses remains constantly high; below is the CPU usage history of my Core2 Duo when ONLY VLC v3.0.23 is running, streaming BBC R2: I'm sure you can see the CPU spikes there ... If you're using yt-dlp.exe without a JS runtime (quickjs is the only option under XP SP3 x86), then the default player client there is android_vr (aka ANDR-V) and the formats it returns are ALL "https", not "m3u8" (PROTO column): yt-dlp -F "ffcitRgiNDs" [youtube] Extracting URL: ffcitRgiNDs [youtube] ffcitRgiNDs: Downloading webpage WARNING: [youtube] No supported JavaScript runtime could be found. Only deno is enabled by default; to use another runtime add --js-runtimes RUNTIME[:PATH] to your command/config. YouTube extraction without a JS runtime has been deprecated, and some formats may be missing. See https://github.com/yt-dlp/yt-dlp/wiki/EJS for details on installing one [youtube] ffcitRgiNDs: Downloading android vr player API JSON [info] Available formats for ffcitRgiNDs: ID EXT RESOLUTION FPS CH | FILESIZE TBR PROTO | VCODEC VBR ACODEC ABR ASR MORE INFO --------------------------------------------------------------------------------------------------------------- sb3 mhtml 48x27 0 | mhtml | images storyboard sb2 mhtml 80x45 1 | mhtml | images storyboard sb1 mhtml 160x90 1 | mhtml | images storyboard sb0 mhtml 320x180 1 | mhtml | images storyboard 139 m4a audio only 2 | 1.33MiB 49k https | audio only mp4a.40.5 49k 22k low, m4a_dash 249 webm audio only 2 | 1.46MiB 54k https | audio only opus 54k 48k low, webm_dash 140 m4a audio only 2 | 3.51MiB 129k https | audio only mp4a.40.2 129k 44k medium, m4a_dash 251 webm audio only 2 | 3.77MiB 139k https | audio only opus 139k 48k medium, webm_dash 160 mp4 256x144 25 | 1.46MiB 54k https | avc1.4d400c 54k video only 144p, mp4_dash 278 webm 256x144 25 | 1.79MiB 66k https | vp9 66k video only 144p, webm_dash 394 mp4 256x144 25 | 1.41MiB 52k https | av01.0.00M.08 52k video only 144p, mp4_dash 133 mp4 426x240 25 | 3.08MiB 113k https | avc1.4d4015 113k video only 240p, mp4_dash 242 webm 426x240 25 | 2.68MiB 99k https | vp9 99k video only 240p, webm_dash 395 mp4 426x240 25 | 2.53MiB 93k https | av01.0.00M.08 93k video only 240p, mp4_dash 134 mp4 640x360 25 | 5.77MiB 213k https | avc1.4d401e 213k video only 360p, mp4_dash 18 mp4 640x360 25 2 | ≈ 9.28MiB 342k https | avc1.42001E mp4a.40.2 44k 360p 243 webm 640x360 25 | 6.80MiB 251k https | vp9 251k video only 360p, webm_dash 396 mp4 640x360 25 | 4.58MiB 169k https | av01.0.01M.08 169k video only 360p, mp4_dash 135 mp4 854x480 25 | 9.80MiB 361k https | avc1.4d401e 361k video only 480p, mp4_dash 244 webm 854x480 25 | 9.37MiB 345k https | vp9 345k video only 480p, webm_dash 397 mp4 854x480 25 | 7.91MiB 292k https | av01.0.04M.08 292k video only 480p, mp4_dash 136 mp4 1280x720 25 | 18.56MiB 684k https | avc1.4d401f 684k video only 720p, mp4_dash 247 webm 1280x720 25 | 17.27MiB 636k https | vp9 636k video only 720p, webm_dash 398 mp4 1280x720 25 | 13.88MiB 511k https | av01.0.05M.08 511k video only 720p, mp4_dash 137 mp4 1920x1080 25 | 49.02MiB 1807k https | avc1.640028 1807k video only 1080p, mp4_dash 248 webm 1920x1080 25 | 30.88MiB 1138k https | vp9 1138k video only 1080p, webm_dash 399 mp4 1920x1080 25 | 24.70MiB 910k https | av01.0.08M.08 910k video only 1080p, mp4_dash I have found the elevated CPU usage (in v3.0.11+) to be specific to the HLS streaming protocol ... Wise choice; and you should probably stay there, too... Words of wisdom ! Kind regards (and protect yourself from the heatwave ... )

For my "old" Chrome and Chromium v144, the video only plays if MUTED. Any volume increment above MUTE causes the video to stop playing. I did not try in Edge (the only browser that I keep on the most-recent version).

So, despite that the video itself ISN'T DRM'ed (e.g., when you watch it in FxESR-115.37.0 under Win7SP1, the DRM icon doesn't show up in the URLbar), for the embedded player to load it needs the browser to come with Widevine support, in case another video on CTVNews.ca uses DRM? This is so badly conceived ...

older builds shows these instead: navigator.requestMediaKeySystemAccess promise rejected 0x80530009 'Key system is unsupported' MediaKeySystemAccess::GetKeySystemStatus(com.widevine.alpha) result=available msg='' and it still doesn't play here.

Thanks for this detailed info ; in case anyone wants to salvage/archive those builds: vlc-3.0.23-win32-a5b72d67 https://artifacts.videolan.org/vlc-3.0/nightly-win32/20260512-0219/ vlc-3.0.24-beta1-win32-4c608da1 https://artifacts.videolan.org/vlc-3.0/nightly-win32/20260529-0214/ It's unclear why they broke XP-compatibility in 3.0.24b1 after May 29th ; it might've been an inadvertent breakage, like they did with VLC-3.0.21, or as part of a conscious decision to drop NT 5.x support (recent 3.0.24b1 builds continue to work under Vista SP2) ; I guess very few people (especially on "legacy" WinOSes) monitor the "nightly" artifacts, so the XP-breakage, if not remedied in the meantime, will only hit the "masses" when 3.0.24 sees a final/stable release (and there's probably still time for that ) ...

The embedded video player doesn't even load in latest NM28 when one accesses: https://www.ctvnews.ca/toronto/video/2026/06/20/germany-fans-take-over-streets-of-toronto-ahead-of-world-cup-game/ TypeError: navigator.requestMediaKeySystemAccess is not a function in newrelic.js:6:8007 I had to use r3dfox (v140.0.4) to view the clip; BTW, the clip is neither DRM'ed nor geoblocked; NM28 should've worked!