Mark-XP

The exexutable's original name was 'yt-dlp_linux' (which incorporates a python-3.10+ runtime) so there's no need to build/brew the python-3.12.5 runtime here (which i did for testing - but only on one development disk). And yes, i'm planing an upgrade to bookworm/trixie/forky later this year (but not very enthusiastic tbh. )

I'm getting too old for much of the new stuff too - a 1GB+ build environment would definitely be unacceptable for me. Btw.: your yt-dlp (version 2026.01.29.071857) runs flawlessly here since end of january (on "Bullseye")

Congratulations @Dietmar! And meanwhile i think that Linux is really much superior to Windows (XP or 7): Lately found a Kingston mSATA ssd in the attic and - without any other application for it - had the idea to use it as transport media from the pc to the notebook utilizing this mSATA-USB3-adapter - mainly to avoid lost file attributes when transferring files to Linux. Had to wait about 2 weeks for the shipping from ShenZhen ($ 2.50 shipping costs? ) so i had enough time to wonder if i could additionally apply some OS's on that stick. After assembly i (MBR-) partitioned the 'disk': A first ntfs system partition for XP, an Extended (ntfs) partition (with Data + Win-7 sub-partitions) and a third ext4 partition for linux. This is what happened: Win XP: xxCloned mbr + bootrecord + the system from a standard ssd (like a dozend times before) in 3 minutes: really fast - but booting from the usb/mSATA resulted in a 07B bsod (0xB84C3524,...). Win 7: it presented all the partitions of the mSATA stick accurately - but stated immediatly that it's uncapable to install on a "scsci" media. Deplorable! Linux: installed flawlessly. Without any single issue. No noticeable performance difference compared to a standard ssd. Fun-fact: GRUB found the first Win-XP partition as bootable and offeres it in it's boot-menu...

Excuse me, is this the crap-message what you get there too: "Turnstile Failed Despite Passing All Checks"? And "All browser compatibility checks passed, but Turnstile couldn't complete successfuly". Tested just out of curiosity, on linux with Basilisk 64 - Luckyly never saw that before on the sites i visit:

I'd like to underline that suggestion to Q4OS Trinity (+ some minimal tweaks). Ok, it remains a Linux, but hey:

Lately i found out, that TinyWall 3.2.5 isn't compatible to .NET 4.5 in Win7 - the admin-panel wasn't accessible, clicking on the admin-item in the menu caused TW stop working. After un-installing the two .NET 4.5 packages (.NET 2.0.x remained) the issue was resolved.

Hello, as regularly i did some backup to an ssd lately. It has the following (MBR) partitioning (The drive letters are as seen in Win-7): 1. (X:) Win-XP: System, Primary Partition 2. Extended Partition: (D:) Data: Logical drive (C:) Win7: Startpartition, Logical drive (F:) Free: Logical drive 3. Active Primary Linux (ext4) Partition The BCD resides in X:\boot\. As usual in my setup, there is no additional 100MB System Partition on the disk. This setup works perfectly on several disks. Now, yesterday after the backup, i booted successfully in all the three OS'es consecutively (for some configuration tasks): into Linux, then in Win-XP and finally into Win 7. Everything fine. Then, by chance, i opened the Visual BCD Editor - but it didn't start: "Run Applicatin with Admin privileges" error message. Just like described here. Though, it is possible to start Visual BCD Editor with "X:\boot\BCD" as argument from the commandline - but it can't Backup/Save it then. There is no Registry Hive HKLM\BCD00000000 in the registry (See here). However, it's possible to mount the BCD (from the Win-XP primary partition X:) with reg load HKLM\BCD00000000 X:\boot\BCD But that doesn't change anything for (Visual) BCDedit: bcdedit /enum BOOTMGR The Memory for the Startconfiguration couldn't be opened The System can't find the specified file bcdedit /enum OSLOADER The Memory for the Startconfiguration couldn't be opened The System can't find the specified file What strange kind of amnesia is this: Win-7 can (dual-) boot perfectly, but it lost the information, where the BCD is? Where does it believe that the BCD is? And, more importantly: how can i tell him, that it's in X:\boot? Edit: Please don't bother! After cloning the first (X:) Win-XP partition, i forgot to remove the existing X:\Boot folder. So Win7, while installing, obviously was aware of another Windows on the System. After deleting the existing X:\Boot folder, a fresh installation went flawlessly...

XP + Win7 + Q4OS in a convenient 3-boot setup here, on pc & notebook (and several backup disks)

Oh NICE, thank you, yes they do work! Tbh, i simply ignored them until now, as the plain yt-dlp executable always did the job. Edit: Now, as i wanted to copy/deploy Python-3.12.5 from my dev-disk to the main System i discovered, that i didn't only ,brew' the executable (of about 36MB) but also an additional Python-3.12 folder in /usr/local/lib - of about 287 MB! So i decide to utilize this (now even more appreciated) Py-integrated version in the nearer future. And i'm wondering a bit, how the new XP-Version can be so small in size!

For the latest version of yt-dlp (2025.10.23) i had to update even my oldstable Debian 11 system (from 3.9 to 3.12.5 - as described nicely here). A little bit complicated and CPU-intense, but it seems to work: yt videos now downloading again without the 403/permission error. Cordial thanks @nicolaasjan and all the other developers!

Well, we've been cautioned betimes : https://msfn.org/board/topic/184368-who-here-has-a-youtube-dl-compile-for-winxp/page/14/#findComment-1277080

i do hook up with Mr. Thornton! Since there was an update to FF 128.10 a week ago (with some "high impact" security fixes) i'd like to ask you @K4sum1 whether / when there will follow updated releas(es) of r3dfox 128. Thank you!

Neither Basilisk/Palemon do, nor Firefox 102. At least Firefox 115 here.

Many thanks again @Saxon, ok, but 1.5 % of 60 W TDP = 0.9 W (in my case!) is far less than the 11W shown in HW-Monitor & Core Temp, the GPU load is 0%, and GDDR5 memory + fan won't suck the remaining 10W... But please don't bather, since this is a 710 thread i will shut up now

Many thanks, @Saxon, i promise to do so! Yes, and that's exactly what what i meant: in my (750 Ti) case i's 1.5 % of 60 W TDP, but it does schow 11-12 W in HW Monitor (and in Core Temp too). Could you please state, what the consumption is in absolute, in your case (GTX 780 - 28nm as well). Thanks!

Really strange - today, the next day, without any change: no problems at all with these two sites... Anyway, i'm a little bit disappointed regarding the power consumption of such a GTX 750 idleing on the desktop: 11 Watt is quite a bit more than i hoped/expected (on Win7 the same) - in comparison to the internal HD530's 0.5 Watt. @Dietmar (or any other user with a 730): How much power does your GT 730 (actively/passivly cooled) consume, while idleing on Windows desktop?

Many thanks again @D.Draker, you're right, i surely must have mixed up something, (and the screenshot of the 344.75 i used due to have better comparison to the functional 344.75 on Win-7). But now the important fact: the 347.26 iCafe version indeed works for the GTX 750 Ti on XP/32 - Many thanks for the suggestion! And on top of that: it still provides the possibility to modify to a custom screen resolution - GREAT, as mandatory for me! Now, after having get rid of the drivers annoying splash animations, with nice grafics on this XP-machine, i started the browser (Spt. 52.9 of late 2024) the first time here, connected to the net, and typed: https://msfn.org (to respond the good news immediately here in the forum): Server not found! All other sites i visit regularly work fine, but msfn.org and https://www.telepolis.de/ do not. Crazy . On Win-7, same machine, same browser, same DNS no problems... Maybe, after having used Linux and Win7 exclusively for web&mail for a while, i suspect to have missed the latest XP certificate news, but that's another story...

Thank you @D.Draker, neither driver worked unfortunately. The only ,advice' regarding that error ["I/O process stopped because of end of a thread" (in german: E/A Vorgang wegen Threadendes abgebrochen)] on the web was: new installation of XP. And that's not practical for me, because of 2 other OS's ,behind' XP, on several backup ssd's. See below the GPU-Z info after the latest unsuccessful driver installation on XP/32 (SP3): There are problems regarding Bus Width, Memory Size an Bandwidh:

Yor're right - but it's the identical subnet mask an the identical BIOS version! (see Win7-GPU-Z below). What do you mean? I did try the (latest XP) 368.81 version with the same result, and - thank you - will try 347.26 later too. But as for me downscaling the resolution is obligatory, and afaik this was omitted after 344.75, so the test will be more academic... 1GB would have been (more) than enough for me, the hunger for VRam is sick imo - but i can only acquire from the (2nd hand) markets...

Well, this is weird: whilst the 344.75 driver could easyly be istalled for an (MSI) GT710 1GB GDDR3 here, it's not possible to install it for the Gigabyte GTX 750 TI 2GB GDDR5 Rev A2 (Low profile). My system: Gigabyte Z170. Did try it first while ,working' on the internal (Intel HD 530) DVI port. And then with the Monitor plugged to the external DVI of the GTX750 too. Tried it by adding it from the device-manager, and with NVidia setup.exe. Anytime the same error: "I/O process stopped because of end of a thread" (in german: E/A Vorgang wegen Threadendes abgebrochen). The Device Id: 10DE 1380 , Subsystem Id: 1458 3667. It's definitely no HW issue: On the same system/disk no problems on Win7 (same driver version 344.75 / 64 bit) and for linux neither! Could be the 2GB Memory of the card be too much for XP/32? Maybe the modern GDDR5 memory? Many thanks in advance for your suggestions!

The ASUS PRIME B860M-A CSM ? Edit: No, probably not, as CSM in this case means "Corporate Stable Model - a commercial program designed to provide stable and reliable motherboards." (see here).

But that little throttle hopefully won't detract too much playing ,Moorhuhnjagd'? No kidding: i wouldn't see any reason to upgrade from Alder to Raptor Lake at all (a fortiori not for an XP user)!

That was to suppose, and you may consider yourself lucky if there's no degradation in that core. Did the board get the latest bios update? i knew this already since over a year, because i read computerbase regularly !

Could you detect whether it was an E- or a P- core which crashed?

What about cloning the 13th gen XP-partition to a separate nvme and trying to boot the 14th gen system with it?