File: ruby2.png (201 KB, 600x600)

201 KB PNG

>>78710095
I just came here to tell you you're doing a great job.
Don't give up and ignore the trolls.
Special thanks to Russian friends.

https://youtu.be/-DAojx2Hgec?t=569

This guy talks about the ability to run OS/2 and POSIX 1 applications in Windows NT in subsystems

>>78710528
Weren't those compatibility layers removed in Windows 2000?

>>78710528
>ability to run OS/2 and POSIX 1 applications
I think that's not really useful. There's cygwin, mysys2, mingw and all that crap also not many programs are POSIX-compliant.

shell/themes/packthem can be used to unpack existing *.msstyles:
packthem -u <file> -a

Fair bit of documentation in shell/themes/themedir


base/ntsetup/textmode/kernel/spconfig.c SpSaveSKUStuff controls the CPU socket limitations per-SKU. This can actually be increased post-install by setting the correct bits on HKLM\System\Setup\SystemPrefix


base/ntos/mm/mminit.c MiMemoryLicense sets the maximum memory per-SKU.

>>78710535
subsystem personalities of NT

jw, has anyone here made a gitignore that can help git work with this? eg to ignore the obj folders & other build detritus?

>>78711734
probably won't happen until a full razzle audit is done.

>>78711734
Seems WRKRazzle has a gitignore that skips obj dirs & build.log etc: https://github.com/stephanosio/WRKRazzle/blob/master/.gitignore
Likely needs a lot more added to it for the full srv03rtm tree though, maybe I'll look into this later if nothing else is posted in the meantime.

>>78710528
os/2 source code when ;_;

>>78710095
>Glowie torrent

repack anon spends months collecting source code and then releases it to the community. textbook example of a glowie.

meanwhile... NOTREPACKED:

>making false accusations
>deliberately spreading misinformation
>basically doing nothing but re-uploading a file that was already uploaded.

yeah, that's our savior, saving us from repack anon. clearly the logical part of the brain is switched off for many who visit these threads.

they've bought into the misinformation campaign hook, line and sinker.

>>78712619
Please don't start the discussion again, it only derails the thread.
If you really want to contribute, then work on some improvements or provide a diff/comparison of the two torrents.

>>78712712
>please don't state facts. we're quite happy living in a lie. stating facts derails the thread.

>>78712712
it's useless, he started spamming his cringe posts even more now that we started including his torrent in the OP too

better to remove the glowie one altogether and leave it with the original, after all that's the only src this thread is focused on anyway, nobody here is working on xbox or dos 6, just going to confuse people even more

>>78712712
if you don't know how to do a recursive MD5 hash comparison of the extracted directories, then maybe you don't have the necessary skills to participate in this project?

multiple anons have already confirmed the extracted files are identical, there's a script for that purpose in the repack torrent which is very easy to use. but keep acting as if the issue has not been settled.

>>78712855
It also contains the 2000 source, which is incredibly useful. The winlogon replacement is based on the Windows 2000 winlogon.

>>78712855
source code for NT 3.5, NT 4, 2000, WRK and OpenNT.

yeah, totally not relevant for this project.

>>78712885
so does the first result for "windows 2000 source code" on google: https://github.com/pustladi/Windows-2000, you don't need to grab 40GB of crap for it

>>78712735
Read again my post:
>If you really want to contribute, then work on some improvements or provide a diff/comparison of the two torrents.
Add your facts and evidence to the FAQ section in OP. Currently your post contains only unverifiable opinions and is not convincing.

>>78712923
nigger you can confirm this yourself instead of spreading false rumors. use the script.

>>78712880
>if you don't know how to do a recursive MD5 hash comparison of the extracted directories
I know what's MD5, but I thought the glowie torrent contains some extra files.

Anyway, I think discussing this is counterproductive and that anon should stop.

>>78713013
the repack* torrent does contain extra stuff, but not in the nt5src.7z archive. the extra stuff is in the other archives, additional source code for other microsoft software, including multiple versions of NT.

>>78713013
not to make the argument go longer but wit all of this file do MD5 avoid any collision??? Like we have probably around 1M file inside the unpacked dir. Like probably there more hashing method for this big stuff

>>78713073
>not to make the argument go longer but wit all of this file do MD5 avoid any collision???
Getting a collision there would be like winning a lottery, I wouldn't worry much about that.
There are also sha256 sums to make sure.

>>78713073
MD5 collisions are so astronomically unlikely that they're not even worth considering.

if you're really that concerned, you can modify the script to use some other algorithm that you trust more. you'd have to change one or two lines of the code at the most.

>>78713142
>sha512
That's even better.

Any exinit/systime news?

>>78713229
Gonna bisect the exinit crashes now, starting from original openNT exinit.c then slowly add my updated functions on top, hopefully can point to whatever function is responsible, just in middle of a clean build atm.

I'll look into systime after that, got the .obj to be ~95% matching too but haven't tested yet.
I'm guessing even though it's matching well it'll probably have its own issues like exinit did though, so I'll try that out separately once exinit stuff is solved.

File: file.png (70 KB, 632x827)

70 KB PNG

>>78713288
Also looking into gitignore too, seems there's a lot of build crap left all over the tree, won't be easy to write rules for, probably will need to exclude a lot of these one by one.

File: file.png (46 KB, 373x758)

46 KB PNG

Someone brought up tinykrnl last thread, pretty interesting, aionescu from reactos did RE on SP1, before WRK was available it seems.
Is likely closer to actual 2003 SP1 than WRK is in some parts.

Was hoping to find that he'd reversed exinit/systime from SP1, but the only link I can find doesn't seem to have it sadly: http://web.archive.org/web/20070214135342/http://svn.reactos.ru/tinykrnl/tinynightly.7z

Still quite a bit there though, including some HAL code which iirc WRK doesn't have at all, pretty neat.

File: file.png (184 KB, 2058x952)

184 KB PNG

I managed to add the _except_handler4_common function to MSVCRT.DLL using code from https://github.com/Speedi13/ManualMapped_SEH_32bit

However ... testing in a Windows XP VM and a Windows 2003 VM, they both hard shutdown after one second before the login screen.

Any idea why? Is Windows blocking this? Is there a log file I can checkout?

Making changes to KERNEL32.DLL did not cause this startup interruption.
Event Viewer does not show an error message
Memory dumpfile is not written
There is no BSOD
The VirtualBox log file is in the screenshot as well as the evidence it has been added

>>78713925
weird, maybe try attaching windbg to it, can be done fine with vmware so I'd guess virtualbox can do it too, if it's a crash it should let you know where it's happening

Postbuild makes a lot of cab files, wonder if newer cab tools could speed it up at all

>>78713925
also yesterday by EH i meant SEH of course :) Well you should attach the debugger and see what is going on. I would build a checked build too.

>>78713957
don't forget to point windbg to the symbols.pri folder too - I think it should work fine without needing to move any symbols around, but I might be wrong about that

>>78714084
what I would like to know is how the fuck we modify the code to turn on debugging automatically when we install windows. Because every time we have to modify the boot.ini file to enable debugging.

File: file.png (72 KB, 1334x945)

72 KB PNG

>>78713957
Yes, Windbg was not with working for me (with the settings in one of the previous threads) it cannot connect/attach, Windbg & Virtualbox are both on the same machine. went to F8 > Debugging Mode and it is waiting
Am I missing something? The error and settings are in the screenshot

>>78714115
yeah that'd be nice, it's something in the base\boot\ folder that handles setting debug flag IIRC.
although if debug is always enabled it'll slow down boot since it has to check for KD which takes a few seconds, so boot.ini option might be better, maybe it'd be nice if we can edit the default boot.ini it creates so that it always has a debug option there

>>78713925

does the ordinal number matters? i see that except_handler4_common does not follow the sequence

>>78714153
as said in prev blog you have to add too boot.ini /debug /debugport=COM1 /baudrate=115200
(check online it's sometime on that line now I can't rember)

>>78714153
Also for pipe debugging you have to run windbg as admin. What we really should do is adding ethernet debugging is faster and don't require admin priviliges

>>78714208
hmm, I never had to run windbg as admin for pipe debugging with vmware though

>>78714219
>>78714208
but yes any other method would still be better than serial/pipes, I did see one place mention using 1394 with virtualbox but I haven't looked into it much

>>78714219
but probably that pipe is arleady use. try something good like ms_fuck ahah.

>>78714259
maybe, it could depend on the VM host app that opened the pipe too
eg. if VMware was running as admin, pipe would need admin access, but I'm running VMware as same user as windbg, so admin wasn't needed
maybe virtualbox was admin for that other anon or something

btw does anyone know which version of vmware tools works on our server 2003 RTM build? latest one it tries installing needs 2003 SP1 ;_;

>>78714379
I was able to use hyper-v wth no problem. but with that you need windbg as admin (at least that was for me)

>>78714379
VMware-Tools-10.0.12-legacy-4448496.zip seems like it might be it

File: file.png (10 KB, 407x193)

10 KB PNG

>>78714415
well those installed fine, but then I got SFC popup...
gonna need to look at disabling this

Any one is working on missing amd64? I'm doing a postbuild right now but boy we miss millions of stuff ):

File: file.png (138 KB, 1880x847)

138 KB PNG

>>78714182
Yes, that did the trick using these settings now. Thx!

>>78714574
You are welcome :) It is annoying that we do not know how to get this working by default in checked builds. when there is no trolls this chat is also good, I'm glad of it

File: problematic.png (87 KB, 503x1126)

87 KB PNG

Wow, the issue with new exinit.c seems to be a single line
OpenNT exinit:
> for (i=0; i < 8; i++) {
New exinit (and also the original exinit.obj):
> for (i=0; i < 23; i++) {

With the 23 changed to 8, kernel boots fine, no BSOD or anything.
No idea how this would cause a stack error, or why using the same value as exinit.obj would break things. Maybe something to do with the constants above it.

File: file.png (25 KB, 362x476)

25 KB PNG

>>78714886
lel yep, one of the constants is wrong, can /g/ spot it?
good one is on left

>>78714937
Make sense maybe we were stack overflowing and try to execute the code of that Master table

>>78714886
Also weird: if I replace ntoskrnl/ntkrnlmp with changed one, it also doesn't boot, but adding /kernel=ntosnew.exe to boot.ini and booting with that works fine.
maybe more SFC shit to blame? another anon did mention having trouble replacing kernel too. guess I need to go thru postbuild and create a new ISO to be sure...

anyone here understand nmake syntax?
need help with a technical q

>>78715071
IS DIFFICULT TO SEARCH NMAKE MSDN AND READ DOCUMENTATION?

>>78715071
just ask, hopefully someone here might know the answer

>>78714937
I CAN'T DEED

File: file.png (14 KB, 540x201)

14 KB PNG

>>78714886
unfortunately systime.c currently makes it bring up this msg after a minute or two, so guess I'll have to work on that, at least it doesn't BSOD though :)
atm my NtLockProductActivationKeys func is only 0.97 similarity, so most likely something wrong there causes this message

>>78715243
hit it with antiwpa in safe mode

>>78715174
okay say we have a line
X : Y
cmd

is X the thing being generated, or is it the thing being processed? what is Y?
i'm asking this because i ran modified scorch on the repo and it deleted services.tab, and the pattern on safedelpattern implies that .tab files are generated, and i'm trying to figure out if that's true or not for services.tab

>>78715318
his is a little condensed, but it's mostly what I already have there, not sure why it wouldn't work properly
it could be an issue with constants again, I'm trusting the anon who worked on systime.c last added them properly, but I should probably check them to be sure...

>>78715328
afaik X would be a target, Y is things target depends on
then make would go through all the Y's, and likely in another makefile like /tools/makefile.inc there's targets that match those and show how they get built

this isn't really nmake specific, normal makefile stuff, it can get a little confusing sometimes though

File: file.png (66 KB, 889x774)

66 KB PNG

>>78715372
ah crap, might have found the systime.c problem actually...

>>78715243
Or you can try to implement a fake one, as done here
https://reactos.org/archives/public/ros-diffs/2014-September/053826.html

>>78715390
Anyone have a am non SP1 ISO? Or MS only ship amd64 starting from sp1?

>>78715411
yep amd64 only started at SP1 sadly
there's beta amd64 builds from like 2 months after RTM that might be closer to RTM than SP1 though, dunno if anyone looked into them much

>>78715428
Have a link?

>>78715428
So basically it is impossible to have a missing_am64... because the work would be to copy stuff that had not changed and patch the stuff that change to be in RTM state. It would be a pain.

>>78715481
Would this one help?
>Windows XP x64 2003 Beta Build 3790.1069

https://archive.org/details/5.2.3790.1069.srv03spbeta.0309051850amd64freclientprofessionalretailenusab1pxfreen

>>78715481
other anon here. maybe these are useful. a lot of betas here
https://archive.org/details/BetaPackWindows2020

>>78715640
It seem 2004, the other anon say two month after shipping rtm. So I guess there is prev version?

>>78715722
https://betawiki.net/wiki/Windows_XP_Professional_x64_Edition_build_3790.1069 says compiled 2003-09-05

>>78715722
compiled 2003-09-05

>>78715748
it's pro, we no server builds? (I know xp is based on server though just be pretty sure but I will give a try)

>>78715870
looks like there's a server one of the same build too: https://betawiki.net/wiki/Windows_Server_2003_build_3790.1069

Fuck I need an AMD processor, intel CPU BSOD with this, fuck

Bet theres some fags over at BA keeping different XP code to themselves.. I mean look at that Longhorn prick

>>78716211
He is just a liar clown. He has nothing. No one has any code for newer windowses... Outside the GSP packs made by MS themselves. Forget it.

>>78716008
try qemu/tcg.

>>78716008
Only QEMU seem to do the trick and I was able to install....

>>78716250
I don't want 'newer' OS code lol

>>78716295
"Newer" includes everything from Win2000 and onwards, except the GSP packs. No beta sources, no winxp/2003 sp source not longhorn, no vista, no 7 sp sources, no win8 and no win10 source

Since we have a (somewhat) working winlogon replacement, what else is missing?

>>78716342
exinit/systime, pretty much all the stuff in missing.7z, probably other precompiled crap in the src too

Seem to have exinit working now, hope someone here can test it out to make sure though, been having strange results but it might be because I've been testing systime too.

https://pastebin.com/8c0R86T5

systime still has issues it seems, probably related to that LockActivationKeys func, I'll look into it some more tomorrow.

>>78716670
Oh whoops, accidentally left in some #ifdefs there

>exinit.c 0.81
https://pastebin.com/XHiAPUQp

btw with this everything should be mostly 1:1 with the original exinit.obj, except for ExpWatchProductTypeInitialization function which is missing a tiny bit of license related crap, should be fine without it, but I'll hopefully get that added later

>>78716705
good than the next job for us is to do the same for amd64 and see what changes in that build! Because we need code that can be compiled for both arcs (three in reality bu ia64 is dead)

>>78716838
also what the hell is hydra I see defines such that all over the codebase! What type of machine is that?

>>78716838
There's nothing really arch-specific in there, but I haven't checked if all the pointer shit is setup to work with both 32/64 yet... can think of 1 thing that might be broken there actually.

Since we do have an ia64 exinit.obj maybe I can BinDiff against that to see if anything breaks when building for 64-bit, dunno if we can even build ia64 yet though, but hopefully should at least be able to build the small part of ntos that this is in.

>>78716865
I think it's a codename like daytona, at least I think I've seen hydra & daytona folders next to each other before, daytona was NT3.1/NT4 codename, maybe hydra is something between those

>>78716949
>>78716865
hydra is codename for termsrv apparently
>Terminal Server adds "multiheading" support to Windows (the ability to run multiple instances of the graphics subsystem), and the hydra is a mythological monster with multiple heads.

daytona was NT 3.5

>>78716949
but yet IA64 is a quite different from amd64, I don't think if it can help.

>>78716865
Hydra is Terminal Services, tightly integrated by default since WinXP to give it Multiuser capabilities. It was mostly stolen Citrix code before it got rewriten in longhorn/vista transition.

>>78716968
well there is hydra stuff even in ntgdi/ntuser stuff

>>78717025
Both use 64-bit pointers, if anything pointer related is broken that should show us, that's really the only thing I can think of that moving it to 64-bit might break

There's a chance that amd64 might have extra code added though I guess, but I didn't see any indication of that since x86 didn't need any special code for it, but if it comes to it I guess we can look into amd64 ntoskrnl.exe

>>78717078
yeah maybye that asp1 beta version of 2003 released two month after rtm. (there is no amd64 release of rtm) that we were speaking before on the thread.

>>78717069
Terminal Services was a huge overhaul of NT - more so than it was really reported in the media at the time.

>>78717109
yeah I can see the changes happened it's spread all over the codebase.

>>78717069
Ofc. Citrix used to offer all these components in their Winframe product, the only official known Windows distro besides MS product itself. When MS decided that Citrix had became too dangerous for them, they went apple way, making impossible to get any new contract for NT4 distros. Then Citrix had no other way, but deal with MS, giving any Kernel/Win32k code they had, in exchage for a permission to sell "addons" and "plugin" products for MS "Terminal Services". They called such addons "MetaFrame". They had to do than in order to get new private libs and headers for NT4+ win32k and kernel, so they could compile their modules.

>>78717094
If there's a chk build of that sp1 beta it'd make things a lot easier if we need to look at it for anything, seems betawiki doesn't say if a build is chk/fre for some reason though

>>78717142
I had toi install it on qemu (if runned on intel CPU it crash at setup), I will check tomorrow if it is checked or not

File: file.png (29 KB, 807x162)

29 KB PNG

>>78717142
Ah yep, it's a chk build, sweet, still stupid that betawiki doesn't mention that tho...
Lets hope there's symbol pack released for this build too, they usually release that with chk builds so hopefully

>>78717166
Have you a link for those?

>>78717207
Those are on BA so nope sadly, seems the enterprise server one is on archive.org though
https://archive.org/download/5.2.3790.1069.srv-03-spbeta.-030905-1850-amd-64chk-server-enterpriseserver-retai

that has chk/fre for server, and fre for professional, not sure if chk professional is uploaded anywhere else

>>78717207
Found nevermind :)

>>78717237
chk pro is at https://archive.org/download/xpx64betas

no symbols ISO on either of these though, but BA has it apparently

>>78717248
what is BA?

>>78717261
https://www.betaarchive.com
lots of builds and stuff on their FTP, but you need to upload something to get access

>>78717261
Yeah having symbols make all easier.

>>78717274
>https://www.betaarchive.com
No one here have access????

>>78717274
Someone should dump all the interesting old builds and other goodies that aren't on the web

>>78717295
Pretty sure I've seen people post in these threads about BA access before

If anyone does have BA, https://www.betaarchive.com/database/view_release.php?uuid=b7d0ec55-5adb-465a-9ea8-6e6b270ef2ba is the symbol pack we're looking for...

Maybe someone can find it somewhere else, couldn't find anything on google about it though

>>78717311
People usually mirror BA's stuff to other places like winworld if it's a new build or something that was hoarded like old longhorn builds (there were some new ones released this year, 16+ years after they were built...)

But seems people don't care much for symbol packs, or even source code.. BA had an awesome source code collection once but I haven't seen it mirrored anywhere.

>>78717354
well for devs symbols are the juice of life ahah

>>78717324
If you will do this, just don't abuse it. Last time some faggot abused leaking their stuff, everyone else got hard to get access to their servers. Abusing was the reason why the relaxed 10 posts requirement to get an FTP account there disappeared.

>>78717384
In fact in this can we just wanna one file, that iso with symbols. Nothing more. So leak he file and nothing more :)

File: WindowsServer2003-5.2.379(...).png (31 KB, 419x347)

31 KB PNG

Ha, nice build

>>78717439
.NET server is the build supposedly able to be build with popfilter in tools folder along with XPSP1 files and some unlockxpfiles scripts in the tools folder

Has anyone heard from OpenNT guy recently?

>>78710095
Winlogon anon and >>78716705 need to put their code on the hidden git. That way, instead of searching through potentially ancient threads just to find one thing, people can just go to the git, making it easier to contribute/test.

File: err.png (46 KB, 979x512)

46 KB PNG

Who knows where i can find ichaud.sys for AMD64?

>>78717883
They already said and have shown no motivation to do that. Someone else will have to do it.

>>78717883
This git not even working (at least for me)

>>78716211
i'd bet there are multiple gatekeepers reading the thread right now, who have access to source code that could really help this project. but they won't release it because their whole identity hinges on the fact that they have access to things that others don't.

>>78718455
im a dev and wanted to clone the repo and start working there but i don't want to be exposed, maybe we can setup some auth shit with encryption because at first glance even the onion site looks like a honeypot

File: Ed8Q2N4UYAExOE1.png (718 KB, 720x1200)

718 KB PNG

Is this the Clover OS?

>>78718602
Packed/Not-Repacked Clover OS?

>>78718455
>i'd bet there are multiple gatekeepers reading the thread right now, who have access to source code that could really help
no doubt, hell I'd wager the only reason XP/2003 leaked is because something better got out, don't forget that the nintendo leaker who was known to share things privately with his friends also hacked MS a year before

File: file.png (26 KB, 640x480)

26 KB PNG

Hmm strange, exinit/systime worked fine when using /kernel=ntosnew.exe, but now creating ISO with the updated kernel gives this BSOD when going to gui setup...
Guess I'll have to enable prerelease flag and debug it.. tomorrow maybe

Looks like different BSOD to the one the other anon had, 0x7E = SYSTEM_THREAD_EXCEPTION_NOT_HANDLED apparently, hm

>>78718697
well, hopefully once NOTREPACKED OS gets released it will be packaged in a CAB file instead of an ISO, for authenticity.

>>78718889
Oh, found one small issue where ExpSystemIsInCmosMode was getting set to TRUE instead of FALSE, maybe that broke setup, not sure

File: sea of green.png (85 KB, 724x906)

85 KB PNG

>>78719073
also after this, systime.c is now matching systime.obj 100% :)
(or maybe 99.9999%, because some stack variables are moved around and I can't fix them atm... shouldn't matter tho)
just hope there's nothing non-code-related breaking shit now

>>78719098
Nice work, bro!

File: IE11 - Win7-2020-11-16-09(...).png (203 KB, 1718x928)

203 KB PNG

>>78708238318
It looks like you can set the desired parameters in your file in developer/%USERNAME%/setrazzle.cmd

>>78719098
Seems systime.c works fine too, at least setup didn't crash with it
Adding exinit.c brings back the 0x7E BSOD though, damn... I'll look into it more tomorrow
If anyone wants to take a look, >>78716705 is still pretty much the latest one, should work fine if you load it via boot.ini /kernel=bleh.exe param

If anyone here has PRERELEASE flag already setup & windbg etc, it'd be nice if you could get a trace of the setup crash... I haven't set up that flag yet, and afaik I need to rebuild a lot for it, so yeah, tomorrow for me :)

>>78719365
You can build WinPE it's has debug that can be activated by F8 without PRERELEASE flag

Clueless AMD fag here, trying to build my first iso. Then I will probably try getting a WinPE iso from what others anon have said.

>>78719648
Oh yeah good idea, I'll give that a try soon
I think the crash could be caused by setup itself though, exinit does a bunch of checks to see if it's in setup mode etc, but maybe WinPE also passes those checks

>>78718741
yeah. i think it's very selfish and shortsighted to gatekeep. what if a hdd crashes and everything is lost? at least if you distribute the files to others long-term preservation is guaranteed. especially if it's done over bittorrent.

>>78719726
I don't get it either, but I doubt there're just on one hdd, much one RAID
>>78719746
What about a semiliterate user willing to test shit for you? I been working on just getting it compiled atm, just made my first iso and about to make a vm to test then they on older hardware.

>>78719707
Instructions for the WinPE build here >>78687696 I would not have been able to figure that out.
But I feel like a retard, where does ISO spit out? I'm working off my actual D: drive.

>>78720107
oscdimg makes iso in %_NTDRIVE% (D:\3790.1.x86fre.winmain.201116-0958_winpe.iso)

>>78719719
>There are 2 floppy images that boot
Do you have any link for those? I can only find that incomplete tinynightly.7z code

>>78720131
Ok, so it's not. That where I figured it'd be, but when I did as stated in >>78687696 it didn't throw any errors, said it completed. If it did fail does it make a log?

We should probably put together a list of 'essential' mods, stuff like
>disable SFC
>always enable KD over com1
>PRERELEASE flag
>disable shutdown-reasons, before build

Can you think of any others? I can look into the 'how' part of them, just need ideas of what people might want to change

>>78720182
You did postbuild before makewinpeimg? (also makewinpeimg has no amd64 support only ia64 and i386)

File: file.png (26 KB, 580x449)

26 KB PNG

>>78720193
Huh, there's actually an ifdef for disabling SFC, TURN_OFF_WFP
I'm not sure how we pass defines to builds without editing the source file tho, can we just use build -cZP -DTURN_OFF_WFP maybe?

>>78720221
So yeah, I'm amd64, so can't make it on this machine then.

>>78720262
i mean windows that you build should be i386 or ia64

>>78720262
you can build on amd64 fine, as long as you used prepatched ZIP and _x64 folder

File: XenixLarge.png (71 KB, 200x200)

71 KB PNG

Any winlogon news?

>>78720484
It's boots to desktop but i think logout still not fixed

Having really weird postbuild error, anyone ever had these before?

(filechk.pl) ERROR: dtc missing dc240usd.dll
(filechk.pl) ERROR: dtc missing dc260usd.dll
(filechk.pl) ERROR: dtc missing kodak_dc.icm
(filechk.pl) ERROR: bla missing cyyport.sys
(filechk.pl) ERROR: bla missing cyyports.dll
(filechk.pl) ERROR: sbs missing camdro21.sys
(filechk.pl) ERROR: sbs missing camexo20.ax
(filechk.pl) ERROR: sbs missing camexo20.dll
(filechk.pl) ERROR: sbs missing ovca.sys
(filechk.pl) ERROR: sbs missing ovcam2.sys
(filechk.pl) ERROR: sbs missing ovcd.sys
(filechk.pl) ERROR: sbs missing ovce.sys

and so on, what's weird is those files are in my binaries.x86fre fine, and I even ran postbuild a few times before this without problems, just randomly started getting these...

Trying to boot whatever it generated just gives "usbochi.sys is corrupted" error, that file isn't even mentioned in postbuild error log ;_;

Dunno why this shit started happening, gonna have to nuke my tree and start over again I guess...

>>78719365
>PRERELEASE
what is PRERELEASE that ?

>>78721122
Inside base\prerelease.inc file is a flag you can change
>PRERELEASE=0
setting it to 1 will make setup try connecting to KD on com1 automatic before setup starts
it changes a few other things too but don't have a list atm
pretty sure you need to rebuild pretty much all of base\ folder to use it though

>>78718540
It's obvious that MS don't give a fuck on this leak, so don't worry to be exposed.

>>78721212
see you tomorrow anon

File: 1583309908214.jpg (185 KB, 1080x1349)

185 KB JPG

If you want to more with the code you need to host it somewhere with a versioning system, a distributed versioning system.

https://github.com/mfyuce/GitTorrent

Easily hosted, distributed, censorship resistant.

>>78721354
maybe we don't use git... ms had to modify in order to have the windows source code posted there.

>>78721508
stil

File: error.png (48 KB, 677x579)

48 KB PNG

did we know how to fix this one?

>>78721354
>>78721508
maybe something like gogs?
>https://gogs.io/

>>78721555
If you can search for different versions without explicitly knowing an address is not censorship resistant

The best hosting is no hosting

>>78721722
*can't

https://radicle.xyz
Is also worth seeing for hosting our project source code using p2p technology.

If one of the coders here decide for a decebtralized hosting i might bother looking

Until then this is quite messy and OP doesnt help by talking about a glowie link

>>78721722
>The best hosting is no hosting
I agree, why bait yourself up with syncing with a hosted repo when you can just do it all offline, manually patch files that need it and profit in the knowledge you aren't being duped by a honey pot vibe repo

>>78723040
>forgetting about the improvements other people can do
nice try fbi

>>78723143
Surely the FBI would want you to use an online repo?

>>78723241
>denying the inprovements people can do with a network of collaboration
>implying even people offline wont benefit
not today cia

Bump for NTDEVs source code

File: 1605338110828.jpg (12 KB, 274x182)

12 KB JPG

personally I wonder how long the code would stay up when someone uploads it on microsoft owned github

(basically making microsoft share its own code with everyone openly)

>>78723784
the irony

>>78723784
Last time someone tried it, it didn't even lasted 30mins.

File: file.png (16 KB, 704x344)

16 KB PNG

>>78723784
http://web.archive.org/web/*/https://github.com/shaswata56/Windows-Server-2003
was up for a few weeks

>>78723784
https://github.com/PubDom/Windows-Server-2003

>>78724964
Lol

This whol day I have been searching for something to allow us communicate chat, and share files over P2P without a server. Here is a perfect implementation if somebody is interested for it. Tor access is also possible.
https://retroshare.cc/

bump

bump

just use the git

Oh the retarted people are back in this, fuck! Yesterdady night was soo good without thid dump/ntpack/norepack idiots. Fuck you idiots

>>78710095
How are you, my (not only) Russian friends?
Happy hacking.

>>78726441
I purged the thread by r*porting them all.
Consider this to be my contribution to this project.

>>78726516
Thanks :)

A general question about the NT build system.. what's the general causes for NMAKE 'don't know how to make' errors?

File: lol.png (104 KB, 677x613)

104 KB PNG

Future is here!

>>78726875
that seems to happen when a targets dependency failed to build

eg. ex.lib depends on exinit.c, if exinit.c fails to build, you get a exinit.c build error followed by "don't know how to make ex.lib"

Do you guys create custom libraries?

I'm starting to work on an article for a near-future PhD application and decided to create one to contain folders for my PDF and e-book files, which got me wondering what else could be done to optimize the Windows-User experience?

>>78726955
>156%
lol

>>78727022
and what the fuck you post in this blog? We are doing other things and we don't have time to waste on stuff like that, sorry

>>78726958
Okay thanks

>>78723784
attach it to their dmca repo like some fag did for youtube-dl

What does 'rtl' do inside ntos?
Is rtl something specific?

>>78728249
Kinda needed then lol

>>78728155
yeah kinda the most important think the common logic shared across the kernel/hal/ntddl/ntuser/ntgdi ecc it's I guess one of the most important think. Don't mess arounf with that

>>78728328
Right okay, not gunna go into detail because some here will rave on about offtopic lol but I have got stuck with the last 2 errors with xbox source and was curious

>>78726516
i hope you at least reported the NOTREPACKED trolls first.

Also can eventually have a missing.zip also for checked builds? I know the fre work also but having also for check would be greater.

>>78728764
Seconding this too

File: file.png (85 KB, 714x906)

85 KB PNG

>>78718889
Aha, finally got exinit fixed under setup too, no more BSODs at all now :)

Was just a wrong variable being used, two had similar names but one was NULL while other wasn't, of course code was trying to use the NULL one...
Sadly I couldn't even get KD connected to debug this, probably would have helped a lot, but I guess this crash maybe happened before KD was inited? not sure.

Anyway exinit is like 95% complete now, only function different is ExpWatchProductTypeInitialization which has some licensing crap missing, doesn't seem that it's needed for system to run, but I'll try looking into making it a 100% match like systime.c later.

I think someone who belongs to closed community made windows_xp_source.rar and wrote about "base64 password" and that archive has been in chinese imageboards

>>78729642
I will soon, need to make sure they'll both work together once ExpWatchProductTypeInitialization is finished though.

The exinit.c from >>78716705 is mostly complete, except line 1255 needs ExpProductSuiteValueInfo changed to ExpProductTypeValueInfo to get rid of setup crashes.
Really hope someone here will give that exinit a try...

>>78729683
It's a useless rename of NT 3.5 where ntos folder is named xpos

>>78729683
the correct password is 'internaldev'. content is fake.

>>78720484
>>78622467
Hey there Mr. Winlogon Anon - Here is a bug report for you when running our winlogon under winpe sp1 x86 / MiniNT w/ WRK x86 kernel - hopefully you can figure this out since it seems important for WinPE builds???

Looks like it is coming from "SetUserProcessData" in either secutil.c secutil.h or winlogon.h

https://0bin.net/paste/pLbivAnd#n91p5I-dILsWANZWUKWETCRSWtkGDCU9HgJUSwiRWcL

>>78728799
I delete binaries fold to strart frsh now I have this missing files (after importing missing.zip and run postbuild -full)


(adminpak.cmd) ERROR: D:\srv.srvrtm_03.x86chk\cladmwiz.dll file is missing
(adminpak.cmd) ERROR: D:\srv.srvrtm_03.x86chk\clnetrex.dll file is missing
(adminpak.cmd) ERROR: D:\srv.srvrtm_03.x86chk\cluadmex.dll file is missing
(adminpak.cmd) ERROR: D:\srv.srvrtm_03.x86chk\cluadmin.exe file is missing
(adminpak.cmd) ERROR: D:\srv.srvrtm_03.x86chk\cluster.exe file is missing
(adminpak.cmd) ERROR: D:\srv.srvrtm_03.x86chk\msclus.dll file is missing
(adminpak.cmd) ERROR: D:\srv.srvrtm_03.x86chk\cluadmmc.dll file is missing
(adminpak.cmd) ERROR: D:\srv.srvrtm_03.x86chk\clcfgsrv.dll file is missing
(adminpak.cmd) ERROR: D:\srv.srvrtm_03.x86chk\clcfgsrv.inf file is missing
(adminpak.cmd) ERROR: D:\srv.srvrtm_03.x86chk\debugex.dll file is missing
(adminpak.cmd) ERROR: D:\srv.srvrtm_03.x86chk\vsstskex.dll file is missing
(adminpak.cmd) ERROR: total 11 files are missing to build the adminpak.cab file

Never had issue such that, I miss admin stuff?

>>78729846
Hmm, could you upload your binaries.x86fre\winlogon.exe file somewhere?
symbols from binaries\symbols.pri\retail\exe\winlogon.pdb would be useful too

>>78729908
Looks like most of those are built from src (seems cladmwiz/clnetrex are at least), maybe in chk they don't get built though, or maybe there was error during build that skipped them

>>78729908
The only sutff different is that I net PRELEASE to 1 in the base folder makefile. But I dunno that change what happen in the admin folder...

>>78729955
Now but before I was running and compile checked build with no problems. Now I have recopy back the admin folder. But it is very strange.

Where is the file that filecheck use as reference?

>>78729982
I had a similar issue earlier where some exe/dlls that I built & postbuilt fine with suddenly just got deleted, had to rebuild them again for postbuild to succeed
No idea why that happened though, maybe postbuild isn't really meant to work more than 1 time per clean build or something

It gets the filenames from binaries\layout.inf afaik, which is generated from mergedcomponents\setupinfs\layout.inx

>>78729955
Or maybe old missing contains those file and than they got delete by mistake?

>>78729846
> under winpe sp1 x86 / MiniNT w/ WRK x86 kernel
If you're building it in the srv03rtm tree and then copying into WRK/SP1 it probably won't work well, since it'd be using RTM structs and ordinals etc which probably don't match SP1

We had to do some hacks to get it working on SP1 amd64, those are only setup for amd64 atm though

Maybe SP1 x86 can be made to work the same way, maybe just replacing the msgina.def with msgina_sp1.def & userenv.def with userenv_sp1.def will fix it (the _sp1 files should be included in v3b zip), you'll need to cleanbuild the whole security\gina\ folder after changing them though

>>78730123
In fact the idea of mixing this stupid is absolutely fucking dump, I'm saying from weeks. SPI1 was the major update in term of kernel export functions. If we continue this trend we result in this project going to shit. like longhorn did for ms.

>>78730123
>>78729846
If it still don't work after that, maybe installing SP1 DDK and copying security\gina\ into there & building with DDK tools would help, that would make sure it's using SP1 structs etc.
We didn't need that for amd64 SP1 support though so hopefully x86 wouldn't need it neither

>>78730123
>>78729928
>>78729846

Also of note to the winlogon maintainer,
here is an x86 postbuild error log:

(delayload.cmd) ERROR: DLCHECK : error DL000000: d:\binaries.x86fre\prerebase\winlogon.exe imports WINSTA_WinStationNotifyLogon which is not handled.
(delayload.cmd) ERROR: DLCHECK : error DL000000: d:\binaries.x86fre\prerebase\winlogon.exe imports WINSTA_WinStationNotifyNewSession which is not handled.
(delayload.cmd) ERROR: DLCHECK : error DL000000: d:\binaries.x86fre\prerebase\winlogon.exe imports WINSTAWinStationDisconnect which is not handled.
(delayload.cmd) ERROR: DLCHECK : error DL000000: d:\binaries.x86fre\prerebase\winlogon.exe imports WINSTA_WinStationCallback which is not handled.
(delayload.cmd) ERROR: DLCHECK : error DL000000: d:\binaries.x86fre\prerebase\winlogon.exe imports WINSTAWinStationNameFromLogonIdW which is not handled.
(delayload.cmd) ERROR: DLCHECK : error DL000000: d:\binaries.x86fre\prerebase\winlogon.exe imports WINSTAWinStationQueryLogonCredentialsW which is not handled.
(delayload.cmd) ERROR: DLCHECK : error DL000000: d:\binaries.x86fre\prerebase\winlogon.exe imports WINSTAWinStationShutdownSystem which is not handled.
(delayload.cmd) ERROR: DLCHECK : error DL000000: d:\binaries.x86fre\prerebase\winlogon.exe imports WINSTA_WinStationWaitForConnect which is not handled.
(delayload.cmd) ERROR: DLCHECK : failed on d:\binaries.x86fre\prerebase\winlogon.exe

>>78730147
Eh, well I wouldn't advise it neither, that's also why I left original exports as msgina.def/userenv.def and only made it use SP1 stuff for amd64, better to keep things as they are already if possible, merging SP1 stuff should only be last resort (like what amd64 needed)

But if that anon really wants to try it on SP1 x86 I don't mind telling him what I know about it, it'll probably take even more work than I mentioned though

>>78730237
Yeah that's known, updating sources.inc should fix it: >>78622467
Change DELAYLOAD line to match the second one in that post

File: ntdev-leaked.png (121 KB, 1500x700)

121 KB PNG

Friendly reminder of who originally leaked the code.

ironically, since we have source code for nt 3.5, including bits for Object File System, the precursor to winFS, I wonder if /g/ might be able to implement OFS better than the abortion that was WinFS (to actually improve upon NTFS, unlike what WinFS became and eventually fizzled out of)

>>78730600
nice photoshop. i was in that thread and it didn't say NTDEV. why are you so insistent on framing him?

>>78730600
archived()moe/g/thread/77874231/#q77874482

just stop it...

File: file.png (21 KB, 912x393)

21 KB PNG

>>78713925
After trying with Windbg. it looks like a STOP c000021a because I replaced MSVCRT.DLL
like encountered here https://blog.enrii.com/2008/07/06/troubleshooting-window-xp-c000021a-error-blue-screen/
My WinDbg kept on crashing/disconnecting every time.. so I could only capture this once
if only I could find the location where it checks it to make it ignore this "problem"...

>>78730908
Could be SFC maybe, try replacing msvcrt.dll inside binaries.x86fre and run postbuild -full, should create proper sfc files for you

>>78722027
>glowie

see >>78730600

anyone who's gullible enough to believe NTDEV is the leaker, is also probably gullible enough to fall for the NOTREPACKED meme.

this board is frequented by lots of psychopath trolls who are mentally ill, and who play sick little mind games with people to entertain themselves.

don't rely on second-hand information from trolls on this board. verify things for yourself.

File: file.png (9 KB, 645x153)

9 KB PNG

>>78730969
Nope.
After replacing the file in binaries.x86fre, running postbuild, and installing again from the new ISO, the GUI setup does not come up - same error, but this time it is a BSOD, with STOP: c000021a
It does complain about the Windows Logon Process. Not sure why ...

File: file.png (7 KB, 649x77)

7 KB PNG

>>78731325
Yep already changed those, but ty anyway.

Currently sitting at 0.94 similarity with ExpWatchProductTypeInitialization now, function is pretty much all implemented, just need to bash it into shape so it matches the .obj.
Late here now though so I'll look into it tomorrow.

>>78731428
Write me at lojiwa6866@aalyaa.com, I have something for you....

I need rtm of this dll that make postbuild fail clusocm.dll
and my rtm install did not have this file

>>78732135
Isnt that to do with Cluster components? What 'rtm' ISO are you using?

>>78732279
I don't know like enterprise what is strange is that I had never had errors before just now after I delete the bin folder. I guess missing v8 don't have the file but an older version do.

Because where that stuff is in the source? I could have delete the folder by accident maybe?

>>78732321
Just try with a normal build SKU.. or find an ISO with the file you need..

>>78732335
Best bet with finding where it is, is to do file and/or grep searches

>>78732439
can you link one?

>>78732465
Archive.org will most likely have a ton

>>78732135
clusocm.dll is built in base\cluster\mgmt\clusocm

>>78732539
Yep I deleted the that folder by error, that is the reason. Thanks :)

File: file.png (62 KB, 955x769)

62 KB PNG

>>78732567
np, really it'd be nice if we had a list of all the DLL/EXEs that we can build, along with the path to the src for them

atm the only other way to find source dir is grepping inside sources* files (i used astrogrep here, agent ransack seems nice too)

>>78730600
>>77874482
Yeahhh, about that.

>>78732645
Well as I said for some reason the folder got renamed to t so that was why I missed those file. Interestly this cluster thing is in base but no one else used it as I was able to build all the thing with no error.

>>78730718
>actually going out of your way to find the original thread because you got triggered by a meme
found NTDEV

>>78733002
>thinking someone's triggered when we're all actively digging though all these archives.

>>78733002
there are so many morons who visit these threads that it's hard to tell who really believes the stupid shit, and who doesn't.

not NTDEV, but... judging from his Twitter you've caused NTDEV unnecessary stress with this stupid meme, because some people believe it. it's a pretty serious crime to be accused of no matter what we think of the ethics of leaking XP.

>>78731428
>>78729175
>>78719098
>> 78717069

Contac us at bibleman@atheist.com we have somthing for you related to amd64. You can imagin what if you rember last nigh chat. Have a good night :)

bibleman@atheist.com

>>78714937
Dear anons, if you diff values like this a lot, try merging the two sides into one by crossing your eyes. The differences between the two should shimmer letting you know where the issue is.

>>78733487
Classic trick. Just don't strain too hard and break if you get a headache.

>>78710535
OS/2 was; the POSIX one was replaced by Interix (which was developed by another company before it got bought by Microsoft), and was later removed on 8.1

>>78733887
pretty sure a lot of the tools got shoveled into RSAT for 8.1+

File: exinit.png (66 KB, 668x756)

66 KB PNG

>>78719098
Another sea of green :)

>>78734704
Also checked manually against MP exinit.obj/systime.obj, every opcode matches up fine :)
Only differences are some larger funcs have different stack variable offsets, but that don't really matter. (If I could fix it I would, but nothing I tried seems to let me change them...)

Here's exinit.c, really hope someone here can try this out for me, so I know it can work on other setups too...
https://pastebin.com/shRezZNA

I need to cleanup systime.c, once I know exinit works hopefully it'll be ready.

Finally got XP Pro installed on a VM. So with just the basic build stuff from rentry winlogin doesn't work right?
I always used to just reset the comp when asked to make a user, and figured I'd do it here. Is it abnormal that I got to the desktop with no issue?

>>78734914
Tiny update for exinit.c, tidied a few things and used enums instead of int in a few places (eg. 3 -> REG_BINARY)
Nothing should be different functionally, just improved the code a tiny bit:

>exinit.c 5.2.3790.1
https://pastebin.com/H9UJ8Xvy

>>78735111
If you used the missing.7z then it came with a winlogon.exe from retail ISO, with WPA etc inside it

>>78735150
Ah, thx, good to know.

bump

>>78735459
bumpy mcbumpface

Betaarchive went demoted the two people that worked to get us the symbol pack. We got it though. If we need any BA stuff from now on, please contact someone in private about getting it, and don't talk about where you got it from.

>>78735518
You're right... Now I think, was really stupid to talk about it here. Many bA sissies lurk here, and one of them got tattle tale with her bA mums.

File: nice.png (113 KB, 802x674)

113 KB PNG

>>78735150
It's works!

File: Снимок.png (452 KB, 1026x842)

452 KB PNG

>>78735150
Testing single core kernel works fine for 2 hours

File: Снимок.png (323 KB, 1026x842)

323 KB PNG

>>78735150
Multi Core kernel works fine too!

>>78737104
It crashes beyond 32 cores

Would it be possible to remove some more niche drivers that aren't really important to reduce the dependence on precompiled software?

>>78737389
has that been tested on datacenter sku versus the pro sku?

The reason I ask is that the licensing shit sets the limitations normally. There may also be a bit of code necessary to get more than 32 cores functional on all sku's.

>>78735150
So if I want to get away from the precompiled winlogin I'd need to do a dirty build with winlogon200X_v3b.zip (rentry has that as the current version)?

>>78737389
Windows Server 2003 supports more than 32 cores?

It's seems i386 datacenter supports 32 cpus max not sure if it crashes

>>78735518
>>78735785
It's fucking pathetic to ban two fucking people for a fucking file that a part from the exinit guy no one else in the world care about!!!
They are stupid and If they play like this I really don't wanna inside that community lol. It's like the file in question is a OS that everyone gonna use.

And secondly is not that that dvd was unique anyone in 2003 with a internet connection would have be able to download the symbols from the MS site as symbols got stored in MS server for a long time. So it's not unique anyway!

bump

>>78737389

File: cpus.png (88 KB, 634x564)

88 KB PNG

>>78738733

>>78710095
bump

>>78730255
>>78730160
>>78730123
>>78729928
>>78729846

Here is the winpe build and the symbols - you will need to have windbg running in kernel debug mode on your host and repeatedly spam f8 before / while the starting setup text and progressbar at the bottom loads to get it to connect to windbg.

This has the x86chk build of winlogon and smss in it based on winlogon200X_v3b.

https://gofile.io/d/gAFTfK

>>78735518
didnt the-eye start to archive their stuff but stop for some reason? http://the-eye.eu/BetaArchive/

>>78735150
btw embedded 2009 minlogon has no wpa components

>>78738750
Thanks!

bump for winlogon anon - other anon posted the symbols winlogon anon requested

bomp

File: file.png (88 KB, 1458x843)

88 KB PNG

>>78731289
Got rid of the BSOD trying to get Windows 10 software to run in Windows XP/2003, it was a MSVCRT.DLL version check that is now bypassed. Needed to include a change in MSVCIRT.DLL too.
Now write.exe from Windows 10 runs, now unchanged in my Windows 2003 Server machine. This is now allowed

>>78744039
gofile changed files

Found this doc in the Xbox source, it's a quite comprehensive list of info on the NT Build.exe environments flags/tweaks/what files it uses to build and tips on making said files etc..

Hope it's help to people

https://gofile.io/d/XQY0Jt

File: build.hlp.png (73 KB, 664x692)

73 KB PNG

>>78746238
Hm interesting, that looks similar to the build.hlp we have in this leak (tools\amd64\build.hlp), maybe this .rtf has been updated though, since the .hlp we have seemed to be from NT4 era and was missing a lot of things, ty for finding it.

>>78746297
That's cool, Just stumbled across it while trying to get a successful build of the xbx src (1 error left lol)

>>78735962
>>78736623
>>78737104
Aha nice, glad it works fine, thanks for trying it.

systime.c should be ready soon, there's a few ugly bits I'm hoping to fix up before release though (it's pretty hard to make ugly parts look nicer when the result has to match up though...)
I got a few ideas how I might get that fixed without breaking it though, hopefully should have it ready in a couple hours.

>>78746368
Did xbox tree need many changes to work? I remember BA had a build drone VM that seemed to build fine, dunno if the creator had to change anything for that tho, also think that was the stripped-down leak that only had kernel parts too.

>>78746297
>>78746238
The source for build.exe is in the NT 5.2 leak as well, under sdktools iirc. Build.hlp is the same between nt4 through at least 5.2, so it will be the same file in the Xbox leak as well as the 5.1/5.2 leak.
They never bothered to change much with Build, except for the version that comes in the DDK - it's got a lot stripped out of it.

Where I can download the xbox source tree?

>>78746534
I released a working version of the barnabas leak a long time ago, on BA. They've since removed all source code on their FTP due to threats from Microsoft.

I'm also the asshole responsible for analyzing and trying to make the latest leak work, with little knowledge of programming (I can try to fix missing dependencies, but beyond that, I'm a lost cause).

>>78746534
Not really, I applied the Complex source update, followed up to a point where I was in razzle, then I added a few switches to the suggested build switches

Then I noticed razzle paths needed public, idw and mstools in the xbox root folder so I took them from inside private


I am aware of the build drone, I can't find it anywhere sadly

>>78746692
Is it the same version that's in the torrent?

I found a version on AssembleGames that was bundled with 2k3 ddk tools but I haven't tested that one yet

>>78747041
>I am aware of the build drone, I can't find it anywhere sadly
Like another anon mentioned it used to be on BA for a while, pretty sure they had both barnabus & the build drone itself on the FTP there, too bad they purged all the sources, really wish they'd let somewhere mirror them before they did that.

>>78746629
Torrent in OP has a copy of it, the original leak was on /vp/ called Microsoft.OriginalXbox.zip or something like that

Wasn't it rumoured that there are other leaks (like the larger may leak) floating around between people, would be interesting if any of them have any variation to the Win2000 Minkernel it uses

>>78747094
>the build drone itself on the FTP there

Does anyone know what the original name of the archive/vhd was called?

how does it build registry hive files?

>>78747041
use the version of razzle in the source code, under public\tools
You probably don't want to try to adapt the version of Razzle from nt 5.1/5.2 as there are customized postbuild scripts included with the Xbox source.

I had a build drone at one point that was intended for building Codename Fusion (The WRK + Xbox loader project), but cannot seem to find the files for it in my backups.

>>78747062
That was the version that I was giving out to folks. It missed all the signing keys that are kinda necessary if you want to run any of the compiled bits on a real xbox. The MCPX won't boot unsigned/badly signed code without some softmod hacks. The COMPLEX mods in the code seem to negate the need for the keys, but they appear to be included in the private\cert files.

The version in the torrent is much more complete. The original barnabas leak was missing a _LOT_ of the contents that were necessary to build a fully functioning Xbox OS dump.

>>78747231
"Xbox Build Drone.7z", with a "Xbox Build Drone.vmdk" inside

>>78747315
>use the version of razzle in the source code, under public\tools

That's what I did, I may have worked badly, I added switches to build not razzle (I think I used build -bcDeFZP in the xbox/private dir)

I still need to figure how to use cdimage tool.

Codename Fusion.. I will see if I can find
anything related to that at least, thanks for the info!

>That was the version that I was giving out to folks

Small world lol, handy to know about running any of the code, I may try out with xqemu for now until I set up my old box..

>>78747287
2k3 uses the hive files (filetype .hiv) but I can't remember where in the src they are at the moment, away from that pc

>>78747492
Thanks!

>>78747497
The latest XQemu and Cxbx-reloaded will run binaries compiled with both versions of the source code. That I am certain of. I've tested the banabas version's ability to build a recovery cd, which also appears to function properly on real xbox hardware.

The newer leak has all the bells and whistles, though, including the source code for Xbox Live, which is how someone is working on a replacement service for the Xbox in 2020+

>>78747287
who? opk pe tool? with bldhives.exe and the well known hive INFs dude.

>>78747569
I mean leaked code. How does the build system build the hives there?

>>78747527
I know its hive file, but don't know how does it produce them?

>>78747619
Same.

>>78747551
Interesting about the recovery cd, I may have to explore as I have a few real boxes about

New thread: >>78748378