>>78588821
shut the fuck up. nobody cares about 20 year old proprietary garbage

>>78588834
Clearly people do. the last /wxp/ was busy let them live you low life

is there already some script that gives semi-automated handholding for building?

Whats current state of winlogon and amd64?

>>78589147
The chat has its own script, plus here some anon threw a script on the github.
In fact, the script is not really needed. Unset read-only attributes, drop the files from the patch into a folder, all that remains is to run razzle by

"tools\razzle.cmd free offline"
then run build
"bcz"
then do postbuild
"tools\postbuld.cmd"
then get iso
"tools\oscimage.cmd pro"
Its all.
If I have not forgotten any moment, then everything is simple

outsider here, sometimes I visit these threads to see what's up. What's your endgame/goal anons?

>>78589576
fix

"tools\postbuild.cmd"
"tools\oscdimg.cmd pro"

>>78589605
The goal is to fully reconstruct the source code, then improve it

>>78589605
I'm doing this for fun. I'm interested in exploring the insides of Windows. Each has its own goal - each fixes a specific problem. Recently, anons have learned how build an x64 edition, which is extremely cool.

>>78589605
To have fun

>>78589710
>>78589576
You forgot to prepatch.cmd.. and if the host is x64 you need the x64 patches to build on an x64 machine

Will there be pre-built ISOs?

File: hqdefault.jpg (13 KB, 480x360)

13 KB JPG

>>78589743
>>78589787
>>78589789

fun is the most valid reason ofc. godspeed.

Does this mean you aim to fully reproduce the original OS so we can actually install and use it?

>>78589903
YES!

>>78588855
>took three days to get to autosage
>"busy"

>>78589907
>Does this mean you aim to fully reproduce
I will never reproduce.

>>78589994
that says more about /g/ than about that thread

what are the most interesting finds since the leak?

>>78590027
No, the fact that over 50% of posts in these threads are "bump" says what we need to know about /g/.

>>78590043
bump :^)

>>78590043
bump

Bump

>>78590037
Internal tools, some test certificates, source of programs like calculator :)

bump!

If anyone would like to test the source patcher I have started to add to Easy-Build script that I would appreciate..
Currently the missing.7z files are not included in the patcher but that's next. For now it has the last v9c patch to download, but I am going to update that to the prepatch anons v10 if the patcher works.

It can also try to set read-write to the source root if the script thinks it is unmodified source and proceed to patch it.

If anyone does test this with unmatched source let me know if it works as expected

https://gofile.io/d/J7rkdu

>>78590271
*unmodified

>>78589378
current state of amd64 winlogon: winNOTlogon

>>78590301
Punny

>>78590311
bunny

>>78590390
Bump the bunny

hop hop bunny

>>78589743
>improve
Has any of that happened yet?

>>78591209
It took them weeks to figure out how to compile it.
It took them months to cobble together a barely-working Winlogon.
TL;DR: No, and never will.

>>78591594

It hasn't even been 2 months since the leak yet dude, bloody hell give them some time they'll figure it out.

>>78591594
Based

>>78591594
Idiot we know how to compile from day 0. Go and fuck yourself in other threads little idiot! You just say shit to people and do absolutely nothing. I hate people like you

>>78591594
We had a working build less than a week after it leaked lol

File: 1602960691586.png (342 KB, 675x462)

342 KB PNG

>>78588821
I see we have a new thread.
I archived the thing on an M-Disc in tar.lz and tar.xz formats.
Will also archive your work once it is placed together and produces a working system.

>>78593498
We already can produce a working system

openxp sucks
* noone talking at element
* almost no commits to git
* empty kanban board
* empty flyspray

OMFG xD
NOT GONNA WORK

>>78593619
Did I miss anything since yesterday?
Well then, is everything documented in one place?
Last time I checked the hidden git is filled with not just branches, but repositories of different users.
I don't feel comfortable with guessing what's valuable and what's not.

>>78588834
Calm down your pussy.

>>78593629
Thanks for your input, NTDEV.

>>78593714
I personally avoid the git, it's better to get the original leaked archive and apply patches that way, the guide in the OP will produce a bootable ISO if you follow it, we still need some prebuild RTM files because the leak was partial..

Just look at the thread info pages in the OP

So what's the status on making XP fully compatible with modern programs? Can you run Steam again?

>>78594030
Far far far away.... We in fact cannot compile XP SP1 still. What we can do is compile 2003 and make an XP build from it. And that is something that program understand because it is an hybrid.. XP fail at postbuild due to miss servicepack stuff

>>78594063
Btw, has anyone tried installing sp1, sp2, sp3 on a compiled version of openxp?

>>78594200
I never connect a vm to the internet... But I guess the same problem would happen as the installer cannot understand which os is that

>>78594063
Who would want to wast time to compile NT5.1 ancient codebase? Is an evolutionary dead end.

>>78594200
this would probably work, but you'd also just be replacing all of our code with with microsoft's.

>>78594200
Doing that makes having the source useless. That's why, even getting working builds by replacing winlogon and cia with retail versions of the files, skilled anons are trying to reverse the source of these modules.

>>78594400
yeah i guess the thing is where the line is between OS and microsoft software that runs on OS, and it's kinda blurry for shit like WinFX but for obvious stuff like service packs we should be doing that.

>>78594752
That's one thing I would love to figure out is how to build a service pack from this source

>>78594327
it works better than wine

>>78595056
U sure?

>>78594833
+1

>MS stil hasnt shut this down
Let me guess, they plan on releasing the win10 source code in some years anyway

>>78594833
>>78595097
Looking at the XPSP1 source in tools/sp they are pretty much the same as postbuildscripts

https://github.com/Empyreal96/build-env-info/tree/master/XPSP1%2Ftools%2Fsp

bump bump bump

bumping for amd64 sp0 winlogon goodness

>>78595056
Yes, I never said otherwise... But still not worth to make a full build of it, unless someone gets curious about how to make with the tree an SP exe package, as stated by some anon posts before.

>>78595118
>Let me guess, they plan on releasing the win10 source code in some years anyway
I think they might be dropping NT in favor of the Linux kernel.

>>78597664
Fuck you. the article was fake and you are retarded if you think they will do that. ALERT: Stupid retarted linux people here tonight!

>>78586178
YES - our winlogon DOES work on a 1/4 installed amd64 retail sp1 version of xp professional x64

It boots and starts GUI mode setup!

Windows Server 2003 Kernel Version 3790 UP Free x64
Built by: 3790.srv03_sp1_rtm.050324-1447
Machine Name:
Kernel base = 0xfffff800`01000000 PsLoadedModuleList = 0xfffff800`01197ac0
System Uptime: not available
WINLOGON INIT!
204.208> Winlogon-Trace: ProfileUserMapping Refs = 0
204.208> Winlogon-Trace: Actually opening user mapping. User is not logged on
204.208> Winlogon-Trace: ProfileUserMapping Refs = 1
204.208> Winlogon-Trace: Actually closing user mapping
204.208> Winlogon-Trace: ProfileUserMapping Refs = 0
204.208> Winlogon-Trace-Init: Boot Password Check
204.208> Winlogon-Trace-Init: Execute system processes:
204.208> Winlogon-Trace-Init: Done with system processes:
204.208> Winlogon-Trace-Init: Sync with thread started by SbBootPrompt
204.208> Winlogon-Trace: Switching desktop from Winlogon to Application

>>78597738
The amd64CHK smss.exe DOES NOT work though.

>>78590390
cunny?

>>78597691
You never know. They also were not supposed to drop IE/Edge and they just picked the cheaper option - stealing a ready browser from Google.
Now if they're already moving everything into the "cloud", what stops them from changing the kernel to the cheaper alternative?

>>78588821
so what do you guys do on your XP setups anyway? are you strictly 32-bit or do you mess with the 64-bit XP and use larger amounts of ram

Bump

>>78588821
>chadmed
is this a reference to ahmed?

>>78594063
Ok I really need to ask this, I have not been following too closely the specifics of this leak, but why do people have any interest in building xp when they have win2k3?

maybe should add this to the guide.

Postbuild phase.
before running postbuild
"set PB_COM=FALSE" - we remove the compression, the image becomes about 150MB larger, but the post-build stage is noticeably faster
"set HORSE_POWER=4" Not perceptibly affected, since most scripts still go to 1 thread

Bump for Microsoft Office source code

>>78599029
Gunna test this with a slower device (~8hrs build time) to see if there is any improvements in that respect

>>78599029
Seems like it should be PB_COMP, not PB_COM, btw

File: file.png (26 KB, 735x242)

26 KB PNG

>>78600678
forgot pic

bump

bump for halo CE source code

>>78600678
Yeap, late night makes me make mistakes
https://empyreal96.github.io/build-env-info/WIN2K3/tools/postbuild.txt

>>78599029
fix
"set PB_COMP=FALSE"

>>78589997
not with that attitude

>>78602517
now there's something i'd like to see

bump

"fixing" an absolute shit tier old laptop with a HDD and 4gb of ram that has to use Win 10, it would work okay-ish if the DISK WASNT CONSTANTLY 100% USAGE WHAT THE FUCK, shit is fucking killing it. How the fuck do you fix this

>Also, the 2004 leaked Windows source code[37] was not seen as legal risk for ReactOS, as the trade secret was considered indefensible in court due to broad spread.
Interesting, maybe that anon that said he talked to a lawyer about this (>>78544322) was onto something.
Then again this is from wikipedia, sourced from a reactos blog post, so no idea how valid it actually is.
Don't forget to let all your friends know about the leaks though I guess.

Winlogon anon needs to post on git. Could help development a bit and people won't have to search old threads for downloads.

Anyone heard from OpenNT guy recently?

>>78603592
Replace HDD with SSD. Or replace Windows 10 with 7/XP or some old Linux distro.
If none of that is an option then you can try to turn off a lot of crap in Winodws 10.
1. SuperFetch (SysMain) - nobody needs that shit.
2. Windows Defender scans entire drive 150 times per fucking day. Get rid of that or make it scan only once in a week or so.
3. Some kind of peer-to-peer shit that helps distribute Windows updates. Turn it off.
...
The list goes on and on. And just in general open "Control Panel\Programs and Features" and remove every item that you can't explain purpose of. It won't harm you PC at all. The only exception might be some kind of driver management software, gay codecs etc.

>>78604141
>Replace HDD with SSD
haha, there are so many I/O operations that it will kill your ssd in a couple of compilations, asshole

>>78604274
Windows killing SSDs is a meme from like 2019

>>78604283
I'm not talking about Windows, im talking about compiling Windows on SSD

>>78597664
>what level of brain damaged parroting of retarded shit for literal decades are you on
>hold my beer

File: 1603624202914.png (177 KB, 441x377)

177 KB PNG

>>78588821
I haven't watched this since the beginning. What's the status on it? Can you compile a fully working version? Did people already start putting work into making updates?

>>78604274
>there are so many I/O operations that it will kill your ssd in a couple of compilations, asshole
Not that anon but being build-win2k3 editor I'm probably in the top 5 of people that built this the most times (probably clean-built at least 50+ times by now, maybe closer to 100), and my SSD seems fine, no reallocated sectors + samsung tool still says health is good.

Probably still wouldn't risk setting up some kind of automated build on it, but for most anons that want to just dick around and try building XP from source a few times I think it should be fine.

>>78604274
Oh I thought you tried to just use that laptop with windows 10. Anyway using SSD for compiling should be fine. You don't need super expensive part for that. Just get the cheapest SSD you can get and go for it. And probably you will reuse that SSD afterwards. It shouldn't die that fast.

And by the way, there is some way to setup some sort of one-time FS in RAM. Sounds like a meme, I know. But just maybe it is possible to use this kind of setup to compile NT5 even faster. If you have enough RAM and use it in the most efficient way, without keeping anything that is rarely needed during compilation and is read only.
Sounds interesting at least.

>>78604274
I feel like SSDs are safe being the fact I have done several full builds and postbuild on EMMC/SD Card storage in a Windows tablet

Bump for tainting open source devs

bump for The Witcher source code

>>78606391
Bump for Ciri's source code

>>78603969
That's some retarted shit the anon is referring to.
This only applies to trade secrets, but not to the copyright law.
There are two ways of protecting intellectual "property" in industry - patents and trade secrets.
Patents give "the inventor" (the company) 20 years of monopoly but it expires and everyone can use it afterwards.
Trade secrets are just secrets - if leaked they're not protected, excluding copyright.
Copyright always applies if something is copyrightable.
Surely some technical solutions from the leaks can't be patented anymore, if they become common knowledge, but you can't just decopyright 13GB of source code.

bump for winlogon

>>78608932
Winlogon was fixed IIRC.

currently in the process of rebuilding the OS after running scorch.pl w/ only the SafeDelPat activated. it's not going great. pretty much nothing important links.

>>78609162
what scorch do?

>>78609461
in my modified form, deletes all temporary files left in the repository that are considered "safe to delete"

>>78609506
I take it that you adapted scorch.pl from XPSP1 codebase?

>>78609534
nah, just patched the one in the 2k3 repo.

>>78609571
Shit yeah I didn't see it at first in my 2k3 folder! Lol

>>78609506
does razzle use scorch.pl? I never bothered to check

>>78610078
I think timebuild.pl uses scorch

File: linus_windows_xp.png (891 KB, 1239x636)

891 KB PNG

>>78605352
When Linus didn't have a beard, he also didn't mention that you could use a RAM-disk to build an OS:)
https://youtu.be/bYbCYgYZVT8?t=328

debloated unbotneted release when?

i have the source code for mspaint but how the literal fuck do you go about to compile it?

>>78604283
Windows XP doesn't even have SSD/TRIM support. It would certainly fuck an SSD in a few compilations.

File: mspaint.png (206 KB, 1836x935)

206 KB PNG

>>78610862
1. Make sure that you are in the razzle command line, as explained in the https://rentry.co/build-win2k3 guide
2. run command: cd (your build folder and) \shell\osshell\accesory\mspaint
3. run command: bcz
4. see image, with an example where one thing is changed using Visual Studio 2008 SP1. The output of 1.-3. is in this as well.

with scorched repo:

65568 files compiled - 1564 Warnings - 18959 Errors
3880 libraries built - 245 Errors
112 executables built - 60 Warnings - 662 Errors
12 browse databases built
13543 files binplaced - 30 Errors

>>78612222
holy fuck i got quads
it's not as big as it seems because half of the errors are just This Library Didn't Build Because One Obj Wasn't There And Now It Won't Build Anything Depending On It

>>78612241
Cool fact:
Quad happens every 1111 posts.

>>78612222
Lol that fucked up then xD

>>78605167
we got a working x86 build, work is being done on getting amd64 working.

>>78612506
Do we have to build anything in particular before building the winlogon source?

My build succeeded, but post build is throwing errors like this:
>"(delayload.cmd) ERROR: DLCHECK: error DL000000: binaries.x86fre\prerebase\winlogon.exe imports WINSTA_WinStationNotifyLogon which is not handles"

>>78612506
>>78612762
I didn't mean to tag your post sorry Anon lol

>>78612241
Checked.
>>78611347
Any big use to compiling MSpaint? Just wondering wile giving a bump.

>>78612368
9999 + 1111 = 11110

>>78613198
That's interesting, you proved me wrong.

>>78613673
>>78613198
Fun fact: Quads sometimes happen every 1111 posts!

>>78613698
>>78613673
>>78613198
Soooooo quads happen 9 times every 1111 posts and one time 1112?

Probably wont help us, but I found a neat set of Windows Longhorn SDKs

https://archive.org/details/microsoftwindowslonghornsdkbuild4074en

>>78614216
these have actually helped us, someone ported the longhorn universal AHCI driver from the samples back to server 2003

>>78614289
Oh cool! There are also WinHEC Longhorn isos on there too with checked builds of 4074 longhorn

>>78612762
I tried clean building ds/security/gina but I am still getting this issue (x86 fre build, winlogon v3b)

>>78594030
Steam still works if you use the old builds...and updated root certs...only problem people have is only steam accounts with older games work...seems new games invoke some newer steam 'features' that break old steam client

File: Billgates.jpg (30 KB, 770x476)

30 KB JPG

the irony that this thread is absolutely dead and has been for weeks, only getting bihourly bumps from losers who really think they're gonna get an xp that runs win10 drivers and programs, is utterly hilarious

Bump

bump

>>78618934
I don't think anyone in this thread thinks that we're going to magically get windows 10 drivers running on NT 5. There's a lot of intricate bits of the OS that were rewritten between NT 5.2 and NT 6.4+ (10+++).

What is most interesting is the ability to build AMD64 versions of SP0 NT 5.2 -- that was something that only Microsoft had internal builds of, prior to this (and a handful of leaked AMD64 builds from the SP1 era).

In addition, this entire leak is an educational way to explore what made NT 5.x last so long.

This is also the closest to the initial LongHorn as well as the basis that Omega-13 was built on (SP1 version of NT 5.2).

We probably want to get to SP1-compatibility before going much further -- I believe most of the SP1 enhancements were backports from the ill-fated Longhorn project.

>>78619778
the problem is that kernel SP11 have like a thon of changes and new exported function. It is like biggest change MS have done with service pack releases. And yeah we can't get any win10 driver we miss all of the new driver interface that was introduced in win vista. (So you can have driver running or in user or in kernel mode)

>>78588834
faggot

File: Screenshot_20201110_122321.png (324 KB, 1920x1080)

324 KB PNG

>>78588821
>NOTREPACKED

File: lick.gif (1.38 MB, 498x280)

1.38 MB GIF

>>78597664
>I think they might be dropping NT in favor of the Linux kernel.
I've read enough netorare doujins to know where this is going. Based Microsoft, cucking all freetards.

>>78619778
> thinks that we're going to magically get windows 10 drivers running on NT 5
IIRC ReactOS are doing some shit using the KMDF/WDF portions that MS published on github to try bringing support for newer drivers over to it, since ROS is pretty much a gimped NT 5.2 I wonder if those changes could also be brought over to this some day.

>>78619852
>kernel SP11 have like a thon of changes and new exported function
A lot of those changes/functions were only added in SP1 due to SP1 being the first amd64 release, since this leak has code for a lot of amd64 stuff we might have some of those changes in here already ((albeit in a pre-release state with probably many differences)

>>78620748
kmdf/wdf was ported to XP as a patch for 'portable mobile devices' towards the end of SP2/SP3's lifetime

It was released as a hotfix package, installable by OEM's. IIRC it was pretty unstable, but officially supported and officially backported from Vista/Win7 for XP's use.

>>78612762
>>78614901
Hm, maybe fixable by editing winlogon sources.inc and removing winsta.dll from the DELAYLOAD= line.

>>78620814
Okay thanks I will try

What is the end goal of this project? Is it to get it working on modern systems (e.g. Ryzen, 10th gen Intel)?

>>78621316
not initially.
the point right now is to get it as close as possible to the released binaries -- winlogon and all the missing bits and pieces need to be fully re-implemented via reverse-engineering the original and other binaries from the era that we actually have debug symbols for.

After that, sure. Maybe even some UEFI support -- ReactOS's FreeLDR might be a big help there. Some experiments have been done to get XP running via UEFI bootloaders, and that was without the source code, so it's definitely doable.

>>78621664
Makes sense, thank you anon

>>78612762
I found that you can ignore those errors and still build a working iso.

>>78602517
with all the leaks that's happened this year I hope this gets out soon
or at least a proto of h1/h2

>>78620748
reactos is a russian project backed by traitors

>>78622313
so? plz don't be like repackfag, posting angry comments whenever something you don't like is mentioned, we should be able to talk about related things here without getting shit replies in return

>>78620814
Yeah the delayload line is wrong atm, oops
Seems like RTM-chk uses
>DELAYLOAD=winmm.dll;netapi32.dll;ole32.dll;msgina.dll;authz.dll;rasapi32.dll;mpr.dll

but since our winlogon currently doesn't import from authz/rasapi32/mpr it'd need to be
>DELAYLOAD=winmm.dll;netapi32.dll;ole32.dll;msgina.dll
otherwise link will give a LNK4199 warning about trying to delay-load a module we don't actually import anything from, and /WX will treat that warning as an error.

>tfw built openxp near my disk size limit and accidentally got 29% fragmentation on my 5400rpm 7mm 16mb cache drive

>>78622313
idiot, I really don't get why people is so mad at reactos, it's like they are jelous because they fail and at least reactos have reach something.

>>78620779
yeah but I guess the kernel have changed quite a little bit from where we are to SP1. Someone should see the bin diff. Because now we can compile amd64

>>78620748
yeah the way is to get WDM into the system because is the base of all modern drivers.

>>78622862
Probably because ReactOS is yet another project that only exists to consume donations and get funded by Russian military only to waste that money. You know, like they always do in Russia.

>>78622923
and than? At least is better than have nothing. Have you look at other hobby OS all are terrible. The problem is that to made an OS you need a lot of money.

>>78588834
Not very friendly of you

>>78623128
reactos is pretty terrible

at least other hobby OS boot

>>78623251
>>78623128
i'm way more concerned with reactos's very poor code quality than like, anything else.

>>78622862
>idiot, I really don't get why people is so mad at reactos
Because they were kicked by ROS devs or mods for trash talking in their forums about mascots, wallpapers, themes or spamming "suggestions" they don't even have idea on. Just like it was here when leak hype was fresh and high.

File: Neptune-tan.jpg (42 KB, 189x156)

42 KB JPG

Any winlogon/systime.c news?

>>78622923
>and get funded by Russian military only to waste that money.
Lolz... Someone came from an alternate timeline... In this timeline that never ever happened. In fact there aren't russians in the Core Team anymore. All them are French, Dutch, German, or people from baltic countries.

>>78623636
Oh. Maybe things will get better then.

>>78623565
nah we're just haiku shills

>>78623251
live cd boot fine for me.

>>78623565
no because they are idiot that think that linux can solve world problems because it is free.

>>78622467
>Yeah the delayload line is wrong atm, oops
BTW, delayload seems to depend on kernel32 containing error-handlers for each function that gets exported.

So our winlogon was trying to get functions that aren't in kernel32 (error in >>78612762 says "imports WINSTA_WinStationNotifyLogon which is not handles").

Kernel32 is also the DLL that amd64 seemed to have trouble with, right?
In >>78583611 it says
>Unable to open section for \KnownDlls\kernel32.dll, error c000003a
Wonder if the updated delayload line might help out with it at all.

>>78625786
Also if amd64-anon is still around, it might be worth trying to mess with KdPrint filter masks to see if we can get debug output from the PE loader, whatever is preventing winlogon.exe from being loaded should hopefully be printed there

https://docs.microsoft.com/en-us/windows-hardware/drivers/debugger/reading-and-filtering-debugging-messages#setting-the-component-filter-mask
Even though that says
>create a value with the name of the desired component
you can just set the (default) entry to 0xF to enable all messages afaik.

I'm not sure how you can edit the registry hives of our build yet though, are they in mergedcomponents\setupinfs?

File: file.png (95 KB, 1546x566)

95 KB PNG

>>78626189
Actually you should create a DWORD entry inside
>HKLM\SYSTEM\CurrentControlSet\Control\Session Manager\Debug Print Filter (create key if doesn't exist)
called LDR (uppercase), and set that to 0xF

Even though the MS doc page says 2003 doesn't use this filtering stuff and it's vista only, Ldr code is definitely using the KdPrintEx function for filtering output, and
>base\published\obj\amd64\dpfiltercm.c
also has registry keys for each component, pic related.

>>78626405
Hm think I know how to add it

Edit mergedcomponents\setupinfs\hivesys.inx
and add
>HKLM,"SYSTEM\CurrentControlSet\Control\Session Manager\Debug Print Filter","LDR",0x00010001,0xF
above the "HKLM,"SYSTEM\CurrentControlSet\Control\Setup",,0x00000012" line (at line ~1651)

then make sure to bcz inside setupinfs folder.

Of course this is assuming you're booting into amd64-chk setup, can't remember if you were or not.

File: file.png (82 KB, 641x480)

82 KB PNG

I'm trying to install XP Pro from the Win2k3 source on a new system, I get to the uisetup and get an error that says Untranslatable message, Win32LastError is 14001. It says that one of the components that Windows needs to continue setup could not be installed; then the installation is cancelled.
Postbuild was succesful.
Has anyone seen this before? What causes this? How to fix this?

Build on Windows XP 64-bit, trying to install on a VirtualBox configured for Windows XP 32-bit - I have not applied the logon stuff afaik

>>78627597
Are you used the prepatched pack? without that there's cert issues and other BS you'll need to fix, prepatched will sort all that for you though

>>78627636
>Are you used
fml, before I get the faggot ESL reply I was distracted writing that btw, meant "did you use" ofc

>>78627597
14001 is problem with asms01.cab, we had that in the early days: >>77961579 / >>77961646
Either use prepatched pack which should generate a proper asms01.cab, or replace the asms01.cab with one from RTM iso.

>>78627636
Yes, I applied both win2003_prepatched_v10.zip and 2k3-missing-x86fre-v8.7z

I will try and build from scratch, after trying >>78627755 to replace asms01.cab

Context: I did try to build for AMD64 as well on this Windows XP x64, maybe that messed some things up

Thx!

>>78627597
Why you did no follow the guide before you post here? Just apply the prepatch stuff and the missing archive and everything is fine

>>78622313
>>78622923
Look, yet another urban legends.
ReactOS didn't receive any money from the russian government.

>>78623251
>reactos is pretty terrible
>at least other hobby OS boot
Not really. ReactOS boots on my stinkpad, but Haiku doesn't.

>>78627990
>I did try to build for AMD64 as well on this Windows XP x64, maybe that messed some things up
Yeah it's likely, seems switching build arch breaks a lot of things, even if using bcz to clean the src
Sadly I'm not sure how to fix it besides starting from scratch again.

If anyone here is bored and wants a task to do, it'd be nice if we could replace an MS root cert with our own one (or add our own instead of replacing)
ATM we have our root cert added as a test cert, but that puts it into test-mode with the "testing purposes only" text on the desktop, it's easy to remove that text of course, but the main problem of it being a test cert instead of proper root is still there.

A script that can generate new cert & private key, and then add that cert as a root in the source code, would be sweet.
tools\openssl.txt in the prepatched ZIP explains how to generate certs/keys, if you look at the source files the prepatched pack changes you can see how test root was changed too, adding as root should be very similar afaik, likely only needs changing the same files.

>>78628209
i'm planning on writing an applet to allow for easy generation of new certs. (currently waiting on that defrag i posted about last night)
we also need this in order to distribute official builds with driver signing.

>>78612222
A lot of those errors are probably things that need dependencies built first, normally we don't need to build deps because we have pre-built libs inside public\internal\*\lib folder, but clearing those out means we'll need to build the deps ourselves first.

I'm pretty sure we have code for like 90% of that dir, if you search sources files for all the lib names you should be able to find them.

Anon, is there any collection of variables and changes in the files that configure the resulting build? I looked at various build scripts, but they are obviously not complete, and I didn't understand much

>>78628056
can reactos run Visual Studio yet?
even older versions like 6 or 2003?

>>78628880
https://empyreal96.github.io/build-env-info/ maybe look here

>>78628087
>asms01.cab
replacing this in all occurrences in binaries.x86fre from a Windows 2003 Evaluation Edition ISO worked! and I learning something: none of the Windows XP/2003 keys were recognized as a valid key - only the one from the Evaluation Edition was; "Real product key was found in installation cd unattend.txt is located on d:\i386 and look product key" as stated in one of the comments on archive.org
Generating my ISO with the asms01.cab from Evaluation Edition also did place this file in my ISO... not sure why it did that but def. interesting
https://archive.org/details/Microsoft_Windows_Server_2003_Enterprise_Edition_Evaluation_Edition_Microsoft_20

File: file.png (67 KB, 769x892)

67 KB PNG

>>78629021
picture with unattend.txt and location of key

>>78629021
If you edit timebomb.cmd and change DAYS to 0 it should build normal edition instead of eval
(afaik eval keys are forced because timebomb.cmd swaps out the setupp.ini file)

There's a modded pidgen src I made a while back that allows setup to accept any key, except because we used RTM winlogon WPA would see key was invalid and prevent logging in
Now we have our own winlogon maybe it'd work fine though, still hoping someone can try it: >>78242359

>>78628975
Yes, I've seen it and its auto-build script, but I haven't figured out its code yet, so I think can find a few useful points there

>>78588821
what are some based windows programs that are must have?
I can think of:
7zip
cmder
sharex
sumatrapdf

what else?

File: 1115251308942.png (15 KB, 300x400)

15 KB PNG

>>78597664
I've been waiting for them to go full Apple for twenty years now. OS development is just a money sink.
If they were guaranteed an Office sale with each Linux install they'd pull the plug on Windows in five minutes.

File: Everything.Search.Window.png (82 KB, 660x564)

82 KB PNG

>>78629166
voidtools everything

>>78629166
I think you're in the wrong thread, bro

Try picotorrent, thunderbird, process explorer, keepassxc, everything

File: file.png (112 KB, 1264x1008)

112 KB PNG

>>78629166
astrogrep is pretty based

>>78629166
Why the fuck are you posting here idiot? Go to another forum and don't brake our balls with this fucking idiotic question!

>>78623595
Nope, winlogon-anon died, systime.c-anon died, Sizzle dev got black-bagged.
Sad state of affairs.

>>78629810
fuck you idiot. Go to another forum, we are doing hard work and we don't need people to waste our time with idiot like you! Night .)

>>78629166
010 editor & cff explorer are good RE tools.
CFF hasn't been updated in years though since the dev switched to making a paid PE explorer app instead, I keep wishing something new could replace it but still haven't found anything yet.

>>78629166
Clipx

bump

>>78588834
You're gonna need to go be bottom bothered somewhere else.

>>78628771
all of the code is definitely there, there's just tooling that needs to run.

>>78614289
link? source code?

Tiny source fix for
>mergedcomponents\wlnotify\sources

On line 66, change
> {$(O)\wlnotify.h=$(DS_INC_PATH)\wlnotify.h} \
to
> {wlnotify.h=$(DS_INC_PATH)\wlnotify.h} \

Allows public\internal\ds\inc\wlnotify.h to get updated with mergedcomponents\wlnotify\wlnotify.h automatically, and gets rid of a p1001 error in the build log (which didn't get logged in build.err)

Might be more publish errors like this that nobody noticed since they don't show in build.err, I'll check once this build is done.

>>78629226
>I've been waiting for them to go full Apple for twenty years now.
Fuck, anon, you're old. Good I'm not that old yet.
>If they were guaranteed an Office sale with each Linux install they'd pull the plug on Windows in five minutes.
Office is the last thing they will port. I think Wine having 99% of usability will make Microsoft switch.
Then they could use their own Wine-like proprietary Windows compatibility layer on top of the Linux kernel.
This isn't as unlikely as it might seem - Valve is pumping money into Steam development, more and more games are usable on GNU/Linux.
Why compete when they can EEE? Linux development is far cheaper than NT development plus there are more people knowing Linux on the market.
For me they could just go the way of Android - proprietary garbage on top of the kernel.

bump for windows neptune build 5000

bump for amd64 anon trying out >>78626788

>>78633177
er i mean that's what I'm hoping for, not that I'm trying it

Did anyone ever manage to boot a chk build at all?

>>78633270
yes

>>78633484
any proof? I've only seen people booting partial ones here, full chk builds bsod on setup afaik

>>78588834
>>78633580
STOP BUMPING THIS BULLSHIT YOU MOTHER FUCKERS

has anyone found anything spooky or comfy yet? is it possible now to build your own customized nt win os? fix exploits and bugs? remove junk????

>>78633614
go back to whatever shitty consumer thread you came from

>>78633646
FUCK YOU

FUCK WINDOWS XP

FUCK CLIPPY

FUCK MICROSOFT

FUCK PROPRIETARY SOFTWARE

FUCK YOU

>>78633580
I'm doing that for x86 since weeks. It only need a debugger but after login because an assert is firing but you can ignore it and everything work fine

>>78633656
epic post bro

>>78633663
>It only need a debugger but after login because an assert is firing
ahh that would probably explain it yeah, thanks for letting me know, probably wouldn't have thought of that

>>78633663
did you use the missing.7z for the missing files or something else?

>>78633747
yes, last version as the last version of the prepatch stuff found in the guide page

Bump for Windows NT for PPC source

Bump for bump

windows is gay lol

>>78634520
Doesn't Bristol NT4 leak includes PPC sources already, along with some ARC firmware examples?

>>78634789
iirc, the compiler is missing more so than the source code itself.

File: изображение.png (101 KB, 1027x403)

101 KB PNG

Hi, anons. What im doing wrong?
Last patches - 10v

>>78635242
You forgot prebuild, you'll need to start over with fresh code tho

>>78635286
Oh wait you're using easy build, hm, not sure then

>>78635296
Thank you, i think you're right,
I didn't run the item with prebuild, just now I saw it, it seemed to me that he should have done it himself, lol.

empyreal96, if you're here, perhaps it should be launched at the first launch?

>>78626405
open up regedit and mount the hive under the file menu. (tester-anon)

>>78628087
I just open up the amd64 razzle64 window after the x86 build is complete - this gets both versions built.

File: изображение.png (52 KB, 1086x689)

52 KB PNG

>>78628087
Im not sure, but do you try this script?
tools\postbuildscripts\deletebuild.cmd

File: изображение.png (175 KB, 1062x642)

175 KB PNG

>>78635286
>>78635338
shit, prebuil doesnt help me even i start build from clean razzle
Any ideas?

>>78634554
What if you can actually use it for everything with out botnets or bloat?

File: 2020-11-11-182224_226x175(...).png (7 KB, 226x175)

7 KB PNG

>>78633656
what about doggo??

>>78636341
Make sure the v10 files copied over properly, if you still get this result try using a fresh extract of the source..

Easy-Build is still very in testing sadly I had error aswell, but I *think* mine was caused with using older Winlogon Source

>>78636341
>>78637318
Also with it's the version of Easy-Build that's got the patcher, then try manually copying the prepatch files, could be something didn't get placed

>>78637318
The fact is that I took clean sources without any patches from the hidden git, unzipped the prepared v10 there, replacing all the files, and copied the easy-script from the github. What could I have done wrong?
If a clean build doesn't work now, I'll try unzipping notrepacked and the previous version of the patch

>>78637383
I didn't see this version, there is only one file on github. I'll look again later

>>78637611
Hmm strange, I wish I knew what I did to sort when I had that issue, I am doing a build with Easy-Build to see if I get that error still..
If you still no luck then follow the rentry.co guide for now while I fix a few things for next update

Btw I will add a prebuild.cmd check on first launch of Easy-Build, don't know why I didn't add one anyway lol, thanks for the suggestion

>>78629166
>>78629252
>>78629543
>>78629944
>>78630191
WSUS Offline (download updates manually, can limit to just security ones)
https://download.wsusoffline.net/

Adobe Photoshop CS2
Windows Serial number: 1045-1412-5685-1654-6343-1431
http://download.adobe.com/pub/adobe/magic/creativesuite/CS2_EOL/PHSP/PhSp_CS2_English.exe
https://www.adobe.com/gr_en/downloads/other-downloads.html

Microsoft Live Essentials 2012
https://archive.org/details/wlsetup-all_201802
https://answers.microsoft.com/en-us/windowslive/forum/all/where-to-download-windows-essentials-2012/7665b354-4f2b-445d-811e-9d8c98e8f872?auth=1

https://www.nirsoft.net/utils/zipinst.html
https://www.nirsoft.net/utils/cports.html
https://www.nirsoft.net/utils/uninstall_view.html
https://www.nirsoft.net/utils/ipnetinfo.html
https://www.nirsoft.net/utils/cprocess.html
ShellMenuView (remove context menu entries)
https://www.nirsoft.net/utils/shell_menu_view.html


https://live.sysinternals.com/
https://docs.microsoft.com/en-us/sysinternals/downloads/contig
https://docs.microsoft.com/en-gb/sysinternals/downloads/

Processhacker
https://processhacker.sourceforge.io/

Resource Hacker
http://www.angusj.com/resourcehacker/

halite
http://www.binarynotions.com/halite-bittorrent-client/
https://www.fosshub.com/Halite.html

paint.net
https://www.getpaint.net/
https://forums.getpaint.net/forum/44-plugin-packs/#elSortByMenu_menu

clipx
http://bluemars.org/clipx/

7zip
https://www.7-zip.org/download.html

copyhandler
https://www.copyhandler.com/

RichCopy
https://docs.microsoft.com/en-us/previous-versions/technet-magazine/dd547088(v=msdn.10)

Fastcopy
https://fastcopy.jp/en/

Sequoiaview
http://www.win.tue.nl/sequoiaview/

TestDisk (data recovery)
https://www.cgsecurity.org/wiki/TestDisk_Download

Bbar
https://www.uwe-sieber.de/bbar_e.html

>>78637880
tinywall
https://tinywall.pados.hu/

simplewall
https://www.henrypp.org/product/simplewall

hexchat
https://hexchat.github.io/index.html

qdir
https://www.softwareok.com/?seite=Freeware/Q-Dir

rufus
https://rufus.akeo.ie/

1by1
http://mpesch3.de1.cc/1by1.html

mpc
https://mpc-hc.org/
https://codecguide.com/download_kl.htm

irfanview + plugins
https://www.irfanview.com/
https://www.irfanview.com/64bit.htm
https://www.fosshub.com/IrfanView.html

conemu + clink +far
https://conemu.github.io/
https://mridgers.github.io/clink/
https://www.farmanager.com/index.php?l=en
https://cmder.net/

thunderbird + lightning +enigmail
https://www.thunderbird.net/en-GB/download/
https://addons.thunderbird.net/en-US/thunderbird/addon/lightning/?src=ss
https://addons.thunderbird.net/en-US/thunderbird/addon/enigmail/?src=ss

Slypheed
https://sylpheed.sraoss.jp/en/download.html
https://sylpheed.sraoss.jp/en/plugin.html

Claws Mail
https://www.claws-mail.org/win32/
https://www.claws-mail.org/plugins.php

youtube-dl
https://github.com/rg3/youtube-dl
https://ytdl-org.github.io/youtube-dl/index.html

notepad2
http://www.flos-freeware.ch/notepad2.html
https://xhmikosr.github.io/notepad2-mod/

metapad
https://liquidninja.com/metapad/

everything
https://www.voidtools.com/

locate32
https://locate32.cogit.net/

edxor and other ds software
http://members.ozemail.com.au/~nulifetv/freezip/freeware/index.html

ripme
https://github.com/RipMeApp/ripme

jdownloader (Ad-Free)
https://board.jdownloader.org/showthread.php?t=54725
https://lzcom.wordpress.com/2014/06/09/disable-annoying-behavior-in-jdownloader-2/
https://www.ps2-home.com/forum/viewtopic.php?t=512

>>78637896
clamwin
http://www.clamwin.com/
clam-sentinel
http://clamsentinel.sourceforge.net/index.php?Lang=en

Bleachbit + winapp2.ini
https://www.bleachbit.org/download
http://www.winapp2.com/Winapp2.ini
https://github.com/az0/winapp2.ini

Desktops 2.0
https://docs.microsoft.com/en-gb/sysinternals/downloads/desktops

wipefile
https://www.gaijin.at/en/software/wipefile

npop
https://www.nakka.com/soft/npop/index_eng.html

discryptor
https://diskcryptor.net/wiki/Main_Page

Virtual Clone Drive
https://www.elby.ch/en/products/vcd.html

wincdemu
http://wincdemu.sysprogs.org/

scanner2
http://www.steffengerlach.de/freeware/

Windows Privacy Dashboard
https://wpd.app/

Microsoft Mathematics
https://www.microsoft.com/en-us/download/details.aspx?id=15702

ERUNT & NTREGOPT
https://www.bleepingcomputer.com/download/publisher/lars-hederer/
http://www.larshederer.homepage.t-online.de/erunt/

azpainter
https://osdn.net/projects/azpainter/
https://www.deviantart.com/frankqbe/art/AzPainter-212-English-Version-Download-488404806
http://hp.vector.co.jp/authors/VA033749/soft/azdrawing2.html
http://azsky2.html.xdomain.jp/linux/man/azpainter.html
https://cosbyart.blogspot.com/2018/04/art-review-azdrawing-and-azpainter.html
https://launchpad.net/~alex-p/+archive/ubuntu/azpainter/+packages

Torifier
http://www.torifier.com/index.html
7TMU-QVWK-RMPU-YAIA

utorrent 2.2.1
http://www.oldversion.com/windows/utorrent-2-2-1-25110

Bitpump
https://www.softpedia.com/get/Internet/File-Sharing/BitPump.shtml#download

Hammerhead Rhythm Station
https://www.freewarefiles.com/HammerHead-Rhythm-Station_program_20888.html

>>78637896
>>78637880
>>78637903
But where's the pornhub downloader?

>>78629166
Teracopy is nice for verifying shit copied/moved correctly, replaces windows default copy function

>>78629166
>>78638088
Also, hashtab to add file hashes to properties dialog

How to make Schannel SSP support TLS 1.2?

bump

Why the hell people post link for software? It so fucking difficult to understand that here we speak about the leak code? Open other thread to speak of these thing!

File: blackcomb.png (71 KB, 967x651)

71 KB PNG

Wow Windows 7(Blackcomb) in 2003

>>78639646
Very interesting!

>>78635242
Remember when Windows didn't look like absolute dogshit? What the hell happened?

>>78639891
Is it bad that I miss the basic aesthetics of the Win9x shell?

>>78639975
Not at all. Windows 2000 looked amazing.

>>78639646
makes sense, given the history here
https://betawiki.net/wiki/Windows_Blackcomb

File: os2nt.png (51 KB, 892x895)

51 KB PNG

looks like you can enable OS/2 support. This is a part of the CreateProcessInternalW API

>>78641028
Shame it's just executable support in 16 bit mode

bump

>>78639891
Someone started getting paid for the redesign

>>78637809
I forgot to mention that when I wrote this to you, I also tried to run the build manually via bz and bcz.
Now I completely deleted everything, unzipped it from notrepacked and patched it with the same v10. And everything came together perfectly without errors.
Maybe your script is messing up files somewhere, but I can't be sure, maybe its my faulth

>>78641996
If you ran without prebuild the build would have deleted some files it needs, only way to fix would be to re-extract with fresh code, like >>78635286 said

>>78641996
I think maybe something did mess it up I am not sure either but glad its resolved! but I have updated the script now so should be okay to use again

>>78642087
I didn't notice that he said to run from fresh sources, maybe it would save me time. Anyway, thank you

does anyone know how to compile netfx from windows server 2003 source code?

>>78642974
=>
>>78639612

>>78643019
My mistake, Thank you.

Anyone willing to recite their experiences compiling on Windows 10 x64? It works with razzle64.cmd provided on the build-doc yeah?

>>78643162
All I build on is Windows 10 x64, just follow the guide yeah :)

>>78643162
my experience - a couple of errors that no one here knows about. Other people didn't seem to complain

>>78643187
>>78643201
Awesome, sounds like things are running smoothly.
Thanks, Anons!

>>78622400
I think the difference here is that "repackfag" criticizes people who lie. And because you're likely a manipulative liar yourself, you don't like being called out like that.

File: f3d9446600f2392f45acb491c(...).png (139 KB, 480x466)

139 KB PNG

>>78588834

>>78598879
no, it's chad med

>>78637933
>youtube-dl
>https://github.com/rg3/youtube-dl
>https://ytdl-org.github.io/youtube-dl/index.html

>>78639612
>>78629166
>what are some based windows programs that are must have?

>>78646040
MORON!

>>78588834
you will never be a woman xDDD

File: image.jpg (157 KB, 1024x678)

157 KB JPG

>>78646109

File: stexbar-screenshot.jpg (73 KB, 1020x764)

73 KB JPG

>>78629166
StExBar is extremely powerful, maybe too powerful for /g/
https://tools.stefankueng.com/StExBar.html

>>78637054
Clippy > Bob

>>78639891
They turned it into flatshit so they could get better battery life on their shitty Windows phones.

>>78646744
Microsoft Bob was the god damn titties

bump for ms bob source code

Slightly off topic, but when the Windiws 10 source leaked, was it just the 'Shared Source' or did other 10 source get leaked too?

>>78648639
nothing leak a part from a small stuff regarding driver but was nothing

Bump

File: file.png (80 KB, 644x920)

80 KB PNG

Any chance someone can test out my exinit.c?

Managed to get pretty much every function to match with leak exinit.obj, sections/segments all match up too.
Only things missing are some license check crap I didn't bother reversing, besides that everything should match up (in fact the first 0x670 bytes up to the first license func I didn't add seem to match perfect - haven't actually compared each byte yet though)

BinDiff gives pretty much everything a 1.00 similarity besides 1 function that had license crap I didn't add, pic related.

Haven't actually tried a full build with it yet though, need to clear some space, would appreciate it if someone can test it for me: https://pastebin.com/eraZcuQY

(if you want to build it to compare with the leak exinit.obj you'll need to remove PRECOMPILED_ stuff from ex\sources.inc file first, don't do this if you're building it to actually run the code though)

>>78650955
If you're using prepatched v10 you'll need to delete these two files first:
>base\ntos\ex\mp\obj\i386\exinit.obj
>base\ntos\ex\up\obj\i386\exinit.obj
Then just save the pastebin as base\ntos\ex\exinit.c, now build should create exinit.obj from the .c file.

>>78650955
Whwn you have time I guess having the license check would be fine to have 100/ swappable file. Because you know something can than use that functionality for no reason ahah. This ms afterall

>>78639891
Indians happened. Useless curryniggers pieces of shit.

So did anyone ever got to compile and run a full fledged XP?

Could you just like slap win32 into a little box and point .exes to it in gnu/Linux without virtualization after some r&d?

>>78651343
Yes from first week after src leak we can compile and boot XP from server 2003 sources

File: Снимок.png (97 KB, 1243x784)

97 KB PNG

>>78650955

>>78651579
Weird, is that with the v10 prepatched pack?

Try adding
>ERESOURCE ExpKeyManipLock;
somewhere after
>#pragma alloc_text(INIT, ExInitSystem)

File: Снимок.png (68 KB, 1243x784)

68 KB PNG

>>78651621
Now it's compiles!

File: file.png (96 KB, 1039x761)

96 KB PNG

Made some progress running Windows 10 software on Windows XP:)

I already got around the "".exe is not a valid Win32 application." error message when trying to start a Windows 10 .exe, now it is actually complaining about missing in function APIs in KERNEL32.DLL, ADVAPI32.DLL, etc. When I implemented one, it complained about another one, so it is definitely recognizing my change:)

Commented out the check in the screenshot
run bsz in \base\win32\client
replaced kernel32.dll in the C:\WINDOWS\SYSTEM32 folder on the installed Windows XP OS. (not when booted, just mounted the VirtualBox VDI using ImDisk)

I am now working with Dependency Walker to add some missing function APIs using the Wine code base, and the headers from the Windows 10 SDK.
I have successfully added a missing function by adding in one of the existing .c files and adding the name to the .def/.src file

Some Win32 API DLLs live here:
KERNEL32.DLL - G:\NT\base\win32\client
ADVAPI32.DLL - G:\NT\mergedcomponents\advapi32
NTDLL.DLL - G:\NT\base\ntdll
MSVCRT.DLL - G:\NT\base\crts\crtw32

>>78651914
you might want to checkout projects that exist for extending the XP kernel - github()com/MovAX0xDEAD/NTOSKRNL_Emu

>>78651914
It'd be nice to find out what the difference is between Visual C's "xp" toolset vs the normal one (eg. v140 vs v140_xp), afaik a huge amount of apps are only incompatible because of this toolset difference, if the non-XP one could be made to work under XP it'd probably help get a lot of apps working

File: Снимок.png (201 KB, 802x674)

201 KB PNG

I Just found method to build working WinPE iso.
Do "bz" in "base\ntsetup\opktools\wpebins"
in root folder execute "tools\postbuildscripts\makewinpeimg.cmd"
And build iso! "tools\oscdimg.cmd winpe"

There is still some missing files in build_logs\winpebuild.log (Search for "Error in copying")

File: Снимок.png (159 KB, 802x674)

159 KB PNG

Another pic

Also it's seems that in base\ntsetup\dirs building of "opktools" folder removed

File: Снимок.png (149 KB, 802x674)

149 KB PNG

All WinPE build errors fixed by adding "opktools" to "base\ntsetup\dirs" and killing "base\ntsetup\opktools\factory\pid.c" file.

"fixed" pid.c if someone needs https://pastebin.com/raw/HwE5A7rU

Is the rumor that there’s a lisp engine buried deep inside the windows kernel true?

>>78652648
Whoa nice, any idea if that could work with amd64?

Anyone have a link to the latest systime.c?

File: Снимок.png (165 KB, 802x674)

165 KB PNG

>>78653129
It's should work if amd64 has working kernel and winlogon.

Anyone have a link to the win 10 enterprise?

new thread when

New thread: >>78654807

>>78651914
What did you change to get rid of not valid app error?

Bump