Leaderboard

both are possible, but I think they didn't test that PR letting it to run for a longer time and browsing more sites and pages. (since I don't have a way to 100% reproducing the crash in JitCodeTables destructor je_free() here)

the landing of #2083 in upstream causes some crashes to me, I'm trying to partly reverting it to see if crash will go away.

I have them. See below and bookmark this post, so that you can always see them when logged in. [Edit] The link to yt-dlp doesn't change any more. It is hosted on GitHub for quite some time now.

... Well, those builds use a CPython 3.7.1 version "hacked"/patched to work specifically on WinXP, ONLY (whereas official CPython 3.7.x, by the PSF, requires WinVista as minimum [client] WinOS) ... The PSF EoS'ed WinXP SP3 with the CPython 3.4.x branch; this patched py3.7 compilation borrows modified "system-like" files from either Wine/ReactOS/OneCore API (not sure exactly which ) projects (files bcrypt.dll, kernelXP.dll, ntext.dll, psapi.dll, ws2_xx.dll), which specifically target NT 5.1 and aim to backport missing functions from NT 6.0 (where py3.7 runs natively ) . It's the same (type of) CPython 3.7.x "hack" that is incorporated into the latest ProxHTTPSProxy versions (maintained/released by MSFN member AstroSkipper); this nice project runs successfully ONLY under Windows XP; for Vista [and higher, though Win7 updated to EoS shouldn't need ProxHTTPSProxy ], the embedded Python has to, somehow, be switched to one of the official py3.7 releases...

@un user Here is the XP SP3 acpi.sys version with OS fakes from 21 December. This is the one that Ramsey uses with its *.pdb. Without OS fakes I can do later, I need to save the files in whole folder Acpi with all new entries before I can do this Dietmar https://ufile.io/hfxc70iv

WinNTSetup 5.3 Beta 7 - workaround windows 7 Dism /add-driver bug 30 (AMD, NVIDIA drivers with LZSS compression) - workaround EFI NTFS driver bug - fixed capture problem with OneDrive On-Demand files - added combobox script selection to diskpart window - added -diskpart command line switch to bring up diskpart window - added WIM_MSG_ERROR and WIMLIB_PROGRESS_MSG_HANDLE_ERROR Messagebox choice

I have changed the link in my signature to the Python 3.8 version now.

The linked version on dropboxusercontent works for me and is fast again. I have YouTube-DL integrated into New Moon, which I use from both WinXP and Win2008, so I like that it works in both.

For your Windows NT6 you could also use one of the official builds listed there. ("yt-dlp_x86_Windows-XP.zip" is the only one manually uploaded after GitHub's build process)

No idea. I'm not a programmer. But there is good news. The last couple of weeks I'm testing another build of yt-dlp, compiled with a modified Python 3.8. Just tested on Windows 7 and no error: [debug] Command-line config: ['-vU'] [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 2023.02.01 [8b008d625] (win_x86_exe) [debug] Python 3.8.13+ (CPython AMD64 32bit) - Windows-7-6.1.7601-SP1 (OpenSSL 3.1.0-dev ) [debug] exe versions: ffmpeg N-109674-gc0bc804e5-2023-01-26-nonfree (fdk,setts), ffprobe N-109674-gc0bc804e5-2023-01-26-nonfree [debug] Optional libraries: Cryptodome-3.16.0, brotli-1.0.9, certifi-2022.12.07, mutagen-1.46.0, sqlite3-2.6.0 [debug] Proxy map: {} [debug] Loaded 1761 extractors [debug] Fetching release info: https://api.github.com/repos/yt-dlp/yt-dlp/releases/latest Latest version: 2023.01.06, Current version: 2023.02.01 yt-dlp is up to date (2023.02.01) Worked on Vista and Windows 10 as well. Could you please test? https://dl.dropboxusercontent.com/s/0b0r4vbdu70njyz/yt-dlp.7z. It's also listed on my yt-dlp GitHub fork page under Releases. (click on "Show all 19 assets") When using the New Moon browser, you need the Palefill extension to navigate GitHub.

Fixed in [jsinterp] Support if statement thanks to @VistaLover's observation. Link in my sig already has the fix (for yt-dlp).

Thank you. Looks like they need to update youtube-DL because it works slowly again. "Unable to decode n-parameter"

Probably most Chromium browsers are linked against static Visual C++ runtime, so it doesn't matter whether Visual C++ Redist is installed or not, the runtime code is in the compiled binaries. No .NET components either. Don't quote me on this either, but I think you can only emulate such hardware. "Using it for real" would mean host OS would have to give it up. Possible AFAIK, but then VirtualBox would work differently and would be overall different experience. But most of these are available in up to 1080p resolution, you should get the same results in all of them at one common resolution (select it explicitly!), so 720p or lower.

As far as I can see, the number of dropped frames only depends on the recording quality of the video... I have found some videos with (almost!) no dropped frames, starting with 0 dropped frames from 2000/5000... see for yourself these 3 videos:

Requires Win7 x64. Sends data to x.ss2.us and packages.axiacoin.com. "Axia Member Center" seems a bit untrustworthy, though I have no real "proof". Not a fan of built-in VPN's (too many folks "abuse" them!) Based on Chromium v107.

There is no mountain, OP is in a climbing gym. As long as there is awareness.of the futility, it's alright. Still, floppies are not a good substitute to recreate the ZIP disk experience, and it is not like your friend played with floppy disk images in a VM. A small gift for your friend, zippier: https://www.flickr.com/photos/dvsjr/albums/1489687 jaclaz

New build of Serpent/UXP for XP! Test binary: Win32 https://o.rthost.win/basilisk/basilisk52-g4.8.win32-git-20230128-3219d2d-uxp-fd45d6324-xpmod.7z Win64 https://o.rthost.win/basilisk/basilisk52-g4.8.win64-git-20230128-3219d2d-uxp-fd45d6324-xpmod.7z source code that is comparable to my current working tree is available here: https://github.com/roytam1/UXP/commits/custom IA32 Win32 https://o.rthost.win/basilisk/basilisk52-g4.8.win32-git-20230128-3219d2d-uxp-fd45d6324-xpmod-ia32.7z source code that is comparable to my current working tree is available here: https://github.com/roytam1/UXP/commits/ia32 NM28XP build: Win32 https://o.rthost.win/palemoon/palemoon-28.10.6a1.win32-git-20230128-d849524bd-uxp-fd45d6324-xpmod.7z Win32 IA32 https://o.rthost.win/palemoon/palemoon-28.10.6a1.win32-git-20230128-d849524bd-uxp-fd45d6324-xpmod-ia32.7z Win32 SSE https://o.rthost.win/palemoon/palemoon-28.10.6a1.win32-git-20230128-d849524bd-uxp-fd45d6324-xpmod-sse.7z Win64 https://o.rthost.win/palemoon/palemoon-28.10.6a1.win64-git-20230128-d849524bd-uxp-fd45d6324-xpmod.7z Official UXP changes picked since my last build: - No issue - implement js::NativeDefineDataProperty helper (2db0386e6) - No issue - reformat GlobalObject::skipDeselectedConstructor (7d2b83faf) - No issue - Remove "code" from jsprototype.h macros (41c2b3454) - No issue - Cleanup unused function exports to self-hosted global (123080858) - No issue - add API to tell Profile Timeline Recording state to JS engine (531906eb8) - No issue - Throw error when resolving or rejecting promise returned by async function with testing function (021610819) - Issue #2089 - Use JS engine stack if necessary when reporting errors (6b50dd5d0) - Issue #2089 - In Promises, use a C++ version of SpeciesConstructor when calling from C++ (8322304fb) - Issue #2089 - Avoid copying/recreating iterator result, AsyncGeneratorRequest and GeneratorObject expression stacks (ece049698) - Issue #2091 - Parse \p{Extended_Pictographic} too. (849523139) - Issue #2093 - Trace HeapPtr<RegExpShared::groupsTemplate_> for GC move/compact events, regression from #2072 (70279e1eb) No official Pale-Moon changes picked since my last build. No official Basilisk changes picked since my last build. Update Notice: - You may delete file named icudt58l.dat inside program folder when updating from old releases. * Notice: From now on, UXP rev will point to `custom` branch of my UXP repo instead of MCP UXP repo, while "official UXP changes" shows only `tracking` branch changes.

The other way around actually, they work but won't install as the installer became XP incompatible. 8u152 is the last with a working installer as you say, but versions up to 8u241 do work on XP if you do a rather tedious manual install process. 8u251 contains one XP incompatible file (sunmscapi.dll) which has to be replaced with the version from 8u241, but otherwise works. After 8u251 the file incompatibilities got so much worse it wasn't worth the trouble of trying to get it to work anymore!

Why ? Have something to hide ? Yes , support forum and I provided with fully working instructions, yet you started to imply the're somewhat bad . Why ? "...roots updater goes up to late 2016: Let's Encrypt sites might be broken still (i.e. about half the web) due to the DST Root X3 expiring in September 2021" That's simply not true.

I'm terribly sorry if asking a person to read again means hostility for you. What "drama" are you talking about ? Yep , we both updated the certs that way , but no drama had happened during the update. xD . Who are you , lol ?

Duh !? That's why I wrote he needs to insert the new certs he downloads directly from MS. Did you read what I wrote ? Well , try again (reading out loud with a thick British accent might help). This method proved to be fully working by many MSFN members.

Debian still builds Firefox-ESR without the SSE2 requirement for i686, and I can confirm it works just fine on my Athlon XP 3000+. As for chromium, I have no idea.

I disagree with this one. I got athlon64 that lacks sse3 but does very decently as guest pc with windows 7. Modern sites produce no issues with it (expect new youtube layout). Some peoples even use win95 today and I see no problem on that.

95% of the websites will still work with another browser, or the older version of Chrome. Of course the big websites are the incompatible ones. But anyways, they've grown too much to be useable on such old hardware. I'm always surprised how well this bulky Pentium 3 of 2001 performs online, although the single-core is bad at doing things simultaneosly. The hardware without SSE3 deserves respect! The Pale Moon forks might jump in the place of the best option for older processors. Just as they are now for Windows XP and NoSSE-Processors, like a Celeron from 1998. Pale Moon also has it's own browser engine.

It's sad to see legacy support dropped yet again for legacy hardware. Although I agree PC's older than 2005 are pretty slow, cutting them off will make it hard for those who can't afford new PCs...

No, those builds are only needed by computers that don’t even support SSE2, which is only a fraction of the remaining Windows XP diehards. I think browsers that support SSE2 will be around for a long time - they just won’t be based on the latest Chromium.

A counterargument to that is there are many poorer countries like China and India where the people simply can't afford newer boxes. They're going to use what they got regardless and so it does everyone a great service to at least provide them with a working browser with some security protections. Now I get that browser developers can't do this forever but Goggle has the money, although they do what they feel like because they have the power. This is where Roytam and others deserve a good deal of credit for keeping the older versions of Firefox patched and up to date as best they can since for these users it's better than nothing. And the side benefit is that the rest of us get to keep using them as well.

The state of play, I believe, is: We have ProxHTTPSProxyMII 1.5 working with Cryptography 2.4.2, which runs even under Python 3.4. For now, this works well. But TLS continues to evolve, so to avoid eventual obsolescence we'd like to get ProxHTTPSProxyMII working with newer Cryptography versions. This required a newer version of Python, so we've got Python 3.7.1 working on XP; but more is required as well (at least a user-written version of bcrypt.dll to substitute for the Win 7 module, but apparently even more than that). The problem is, once you move beyond Python 3.4, developers assume you're running on a newer Windows version, so all packages developed for newer Python versions may also use newer Windows functions. It may not be practical to go much further.