user57

hmm did you try the code i posted in page 1 ? for HKEY_LOCAL_MACHINE/System/CurrentControlSet/Services/FltMgr, make sure that it has "Tag" [DWORD] value of "1" (0x00000001). = "HKLM\0, System\\CurrentControlSet\\Services\\FltMgr\0, Tag\0, %REG_DWORD%\0,01,00,00,00", others : "HKLM\0, System\\CurrentControlSet\\Control\\Filesystem\0, NtfsDisable8dot3NameCreation\0,%REG_DWORD%\0,00,00,00,00", "HKLM\0, System\\CurrentControlSet\\Control\\Filesystem\0, Win31FileSystem\0,%REG_DWORD%\0,00,00,00,00", "HKLM\0, System\\CurrentControlSet\\Control\\Filesystem\0, Win95TruncatedExtensions\0,%REG_DWORD%\0,01,00,00,00", it would allow you to make a fix directly

// h265 --enable-libvpx --enable-libx265 --enable-libaom // new image formats --enable-libwebp --enable-libjxl do ffmpeg handle these well in xp ? what with the LAV filters - guys, we would need people what make a list what works with this one BUT lav is a classical "programming engine" windows 10/ect. users are using in their programs to play h265 but its a engine written for windows10/ect. later on it get to DX11/CUDA ect. - where cuda at best has a few support but to really handle this controlment what comes afterwards we would need the handling for the cuda in ring0 it wont work with a limited function set that is provived in XP so either libde265 or maybe a reverted x265 would make the right decoding progress, cpu power wise you dont need that hardware acceleration (grafic card wise) for a decoder - a decoder useally can be just considered a player also h265 useally has parts written in MMX/SSE/AVX - those are a lot faster - it would be enough of speed if these things still need a discussion, people should come out with a discussion

nice one, but you said the exact oposite according to this they do not need a real name, real email, no smartphone number to gather this information -> witch means these services dont need to ask us these things to do so then dont ask us for our real name, our mail address or phone number -> just putting the logic right and a company should not do a such thing at all

well what you wrote now just goes to what i said in the past i dont join other projects like win2k or vista , nor i make a version war but it keeps ringing these questions - you followed that and you drained out i told sam and others a similiar thing - but yet he wants to go for win2k the todo list is far to big to jump around - in the end nothing is done while dibya for example started a vista version - what made me like angry because i know it leads to something like that

well technical speaking actually it would be possible to make a check what technology in the XMM registers are possible there was recently a problem like that with a microsoft file (in the supermium posts) where actually SSE4.1 instructions are used more professional programms (and useally you would expected something like that from a very big money concern like microsoft) check what instructions are available (the x265 encoder/decoder does so for example) after that then the right routine is called up often it is a compiler setting problem one method to avoid this one would be to create an .obj file (what contains the functions without SSE/MMX ect.) then at first you make the check if SSE can be done if not you call up that .obj file´s function if you have SSE you call up your normal function thus you can avoid creating multiple complies like having a SSE version and a non SSE version

you building it into the explorer ?

is the missposition of the fonts still persistend ? in the past there was a smaller discussion, i reverted that GDI problem once and saw that they changed the distance handler from float to int i didnt have the problem of missposition, however in that gdi-chromium i think the int variant was choosen but if you dont revert it to float then probaly some fonts are missplaced or even not visable who dont know what a font is : a font is like every sign ABC123@|^ ect. - it is key for a webbrowser

USHORT is a short syntax for unsigned short signed short (or just short) goes from −32767 to +32767 unsigned short goes from 0-65535 https://en.wikipedia.org/wiki/C_data_types in newer compilers i saw that is sometimes a problem if you put a unsigned char vs a BYTE - it starts to complain and fail in VC6 it dont do that - because BYTE is a macro name for unsigned char you have to understand it that way: 16 bits (that is what short represent) (Length) 16 bits (MaximumLength) buffer (32 bits/4 bytes) (PWSTR Buffer) then here useally is that string

i find interesting when i talked to sam and dibya and i came up with the idea ffplay would be the option (also on page 13 here), you just came up with that one you more or less can forget about LAV - LAV is a software engine that use up like dx11, cuda and other things the thing what is handling that internal is something they call CUDA (in kernel mode - not the usermode functions) (look what he writes on his github: NEW: D3D11 support for HEVC 4:2:2 and 4:4:4 hardware decoding v0.80.0) LAV is not something that provides such en or decoding for h265 and such - its a engine what use a next engine (like dx11, cuda ect.) (its like a script or prescript - what meet its demise later on) support matrix (NVENC, NVDEC) https://developer.nvidia.com/video-encode-decode-support-matrix you actually dont have h266 there - and thats forever (all cards up to 21.12.2025) for that grafic cards because its a print <-- what you could do for us is making a last like nvidia did - what LAV and what modes can be done with it, like the one from nvidia grafic card "encoding" wise there is a big problem , hardware units doing so are a print - so there are many modes grafic from grafic card that is not supported and printed is printed - while software can be upgraded the other option like having a programmable core is close to the same thing as if having just 1 cpu core more with software instructions cpu decoding dont need that much power even with normal instructions (the XMM registers are a lot faster) you dont need a grafic card to encode a video or image here i wrote exactly this (no video card, no weird dll´s, no weird os functions, no directx, no cuda, no lav): https://msfn.org/board/topic/185879-winxp-hevcheifheic-image-encoderdecoder/ it use normal instructions if (hardware acceleration is turned off) - if on it checks what XMM registers are available (these are like 2-150 times faster) also from consideration would be the vLC player - it suppose to work with video´s from H265 - how many modes are not known to me to CUDA and XP to say is that the CUDA versions that are available on XP are like old, the functions are probaly very limited the only real option would be to either have the source code of the nvidia driver (how it controls its cuda) or how to give that cuda kernel code its controlment (like control/input/output) so after all of that - yep you right ffplay was my idea too but ffmpeg in the past was very self-contained - today they already use also engines, dx11 and such things - so it might also have its limitations if you really want a h265 support on mpc-hc you have to write a connection from mpc-hc to libde265 (that one can decode h265) https://github.com/strukturag/libde265

programming-wise normal strings get translated to the unicode string (windows do so internal with its deeper function) https://learn.microsoft.com/en-us/windows/win32/api/ntdef/ns-ntdef-_unicode_string theoretical these can have a bigger value of 260 - max_path has not 256 for a reason of terminating 0´s an example would be CreateFileA/W this one gets translated to ZwCreateFile(NtCreateFile) to create this unicode string there is : RtlInitUnicodeString( &fileNameUnicodeString, L"\\Device\\Harddisk0"); and then: InitializeObjectAttributes( &objectAttributes, &fileNameUnicodeString, OBJ_CASE_INSENSITIVE, NULL, NULL ); then it looks something like: ZwCreateFile( &hFileHandle, SYNCHRONIZE|FILE_ANY_ACCESS, &objectAttributes, &IoStatus, NULL, 0, FILE_SHARE_READ|FILE_SHARE_WRITE, FILE_OPEN, FILE_NON_DIRECTORY_FILE, NULL, 0); explaining that unicode-string: typedef struct _UNICODE_STRING { USHORT Length; USHORT MaximumLength; PWSTR Buffer; } UNICODE_STRING, *PUNICODE_STRING; this one got a buffer and a Length and MaximumLength - ushort is 65535 in size but that are theorecial numbers it raise questions to make so many folders so they extend the length of 256 signs

if you want to give a installer to us you can make it by using this project file you know about the kb-upgrades ? they are compressed like a zip file (Intrapackage Delta compression or IPD) - the most common unzip files like winzip, winrar, 7z can unpack this format inside is useally a .inf file and that .inf file contains both the registry and the file entrys typically they are easy to find because they useally come over this 2 entrys: CopyFiles = filesnumber1 AddReg = regnumber1 then somewhere is a string called filesnumber1 and there are the entrys the same goes for regnumber1 so the common idea is to read out that information (what is stored in that .inf file) and combine more kb-upgrades in 1 upgrade - a problem with that is that this .inf engine is like a script its not very fast - so its not very good for many entry´s so i came up with a own installer that can do the same thing (but many times faster, having full control of the code, dont need other things to be installed first) so if you want to make us a file that install that fltmgr.sys you can make it with this project file - might be something for you installer.zip

hmm trying to get a view do you use a kernel extender ? some are partial functional and use the same folder what software you are using ? if you turn off many of them and limit the software, would be something the OS in this case xp has background activity for the system - but first would be the software to be turned off you know how to handle a file logger ? maybe microsoft´s filemon a lot of these files are not cookies - your search is for changes - activity also caused changes like stats, or that deskop.ini is not a cookie file https://www.techbloat.com/what-is-desktop-ini-on-windows.html

the same discussions where around in the past from what the xbox solution was just to increase the sector size from 512 to 4096 - also there is some information about that this has problems either a problem with system based buffers or with user buffers also the one core api is saying to have that problem solved having current informations would be fine

well then just move enough files on that harddrive until it extends the 2 TB limit, better would be to use the entire disc if the files are then corrupted or the checksum test has different value´s it dont work

hmm i do not know the details but these are 2 different error messages: "api-ms-win-crt-convert-l1-1-0.dll was not found" -> this means that dll/file was not found "ucrtbase.mbstowcs could not be found in the dynamic library api-ms-win-crt-convert-l1-1-0.dll" -> this means that the dll/file was found but not the function "ucrtbase.mbstowcs" in that dll ucrtbase refer to the c-runtime (ucrt - Universal C Runtime Library) so you might can give it a try by installing the vc-runtime https://learn.microsoft.com/en-us/cpp/windows/latest-supported-vc-redist?view=msvc-170 but i do not know if that one needs sp3 - its worth a try but there might also be a different version or modded version of that api-ms-win-crt-convert-l1-1-0.dll - what can give you function maybe if you have one (these files had different compiles for different os´s of windows) (what could explain why that error message says it found the file but not its function) if the c-runtime works - a good method to avoid this problem is to ship the vc-runtime with the executable (in this case ytdl)