Harrison
doesnt know if that hurts since that thing works all the way from XP to 10
Harrison
Harrison
GB2312 issues, ehh, this is wine btw
Tobiyo
fonts i guess
Harrison
it's GB2312 decoded as ISO8859
Tobiyo
oh my
Chungy
Isn't Windows supposed to be UCS-2 native?
Harrison
Windows uses native character set by default
Harrison
you'll have go to region settings in control panel to enable full OS Unicode in Windows 10
rosbridge
<hbelusca (Gmunu=8PiG/c4Tmunu)> For interested people, here's a made up documentation for the very early PE format used in MS NT v1.196 . https://cdn.discordapp.com/attachments/480317303339089930/828441453495189544/NTPEConv_doc_format.md.html
Harrison
I know not much about these details but if my computer complains about missing VC++ runtime this AIO installer fixes everything
Harrison
Because some programs that used non Unicode will break
Harrison
Yeah what I exactly meant is that MS will probably get rid of those C programs that invokes NT APIs and NT internals gradually and everyone should use .NET WPF Winforms etc wrappings from now on. Didnt express that one clear enough btw this message didn't make it through telegram bridge.
Harrison
This is what it says
Harrison
It's really just installing a pack of VC++ runtime redist
Harrison
Harrison
Id like to know where to find one
Harrison
overwrite them in ReactOS?
Harrison
Okay, ReactOS bugchecks when I copy
Harrison
That application requires to be installed
Harrison
It's got daemons monitoring for safety as well(QQProtect.exe)
Harrison
Nope, it wont work even after replacing them in ReactOS\System32
Harrison
Harrison
no definitely didnt
Harrison
oh i did actually
Harrison
I replaced all ReactOS\system32 api-ms-* with windows 10 syswow64\downlevel ones
Harrison
including ucrtbase.dll
Harrison
yes
Harrison
i even put them into C:\Windows\System32
Harrison
yes, that's what i'm thinking about. it's got problems dealing with the dlls we got
Harrison
The only way we can find out would be debugging
Harrison
Which I have no idea how to do
Harrison
I do and I did custom build mine
Harrison
For that task manager mod
Harrison
Received dozens of reviews the minute I created PR
Harrison
will try it another day
Harrison
it's really late at night here. see you
Harrison
Oh, by the way, is there an effective way of debugging ROS?
Harrison
I got a docking station for my old ThinkPad so with a null modem cable I think I can do live debugging on real machine
Harrison
Seems it's a topic for me to discover another day
Harrison
I read the wiki pages and that's pretty vague though
Harrison
I read the wiki pages and that's pretty vague though
Like how do I breakpoint in GDB when I have my sources
Harrison
What to do to make GDB look for symbols
Harrison
WinDbg is fine, but I'll have to install VS2017 then because of stupid MSVC
Harrison
that doesnt sound good
Harrison
I thought you meant that the kernel isn't capable of handling the serail port communications so the kernel crashed when data comes in 😂
Harrison
Someday I'll try to do it
Harrison
I have Haiku on that ThinkPad currently
Harrison
Just trying different things
Harrison
IDK I think I'll go with WinDbg anyways
Harrison
There's much much more information on debugging Windows with WinDbg
Harrison
Not with GDB
Harrison
user apps will be as easy as doing it on Windows
Harrison
OS level stuff is what i'm talking about
Harrison
Like weird crashes, stuff like that
Anonymous
"getting rid of those C programs that invoked NT APIs and NT internals" so now you're telling that MS will probably get rid of the kernel mode drivers?
Anonymous
You're mixing NT and Win32 as usual
rosbridge
<hbelusca (Gmunu=8PiG/c4Tmunu)> That's completely idiotic: your über-beloved loosenix program, when using these POSIX apis, is doing direct syscalls to your loosenix kernal.
rosbridge
<hbelusca (Gmunu=8PiG/c4Tmunu)> That's exactly the same level of "internal stuff" as a user-mode NT program using NT apis.
Victor
go on, we accept patches :)
Victor
the attached link describes the bug, actually. Need to figure out if it's our failure or microsofts
Victor
there are not so many of us, I honestly don't want to bother for now. Left my old version of MSVC 2019 not updated for now)
Victor
I wouldn't be so confident here, maybe we should better write our ASSUME:whatever stuff now
Victor
the previous bug was worked around)
Victor
well, gdb is working actually
Victor
you need a build with KGDB debugger and DWARF symbols
Victor
no
rosbridge
<hbelusca (Gmunu=8PiG/c4Tmunu)> @extravert34 If the discusso is related to the mov "long" stuff I reported, the ASSUME don't fix anything.
Victor
<hbelusca (Gmunu=8PiG/c4Tmunu)> @extravert34 If the discusso is related to the mov "long" stuff I reported, the ASSUME don't fix anything.
maybe, those assembly stuff is hard to keep in the head :) I don't remember the rules how long and short jumps should be generated
Victor
by default on gcc builds, our custom debug data format is used along with a custom debugger (KDBG)
Kumamon
now that April's fool is over, isn't time to change the name back to usual already?
Chungy
https://www.supremecourt.gov/opinions/20pdf/18-956_d18f.pdf
Chungy
tl;dr: Google did nothing wrong by copying the Java API and reimplementing it in Android: no copyright violations.
Chungy
This also has the side-effect of protecting projects like Wine and LXSS
Chungy
also protects ReactOS for doing it :P
Anonymous
No
Anonymous
Using WinDBG as a debugger tool to debug GCC builds will print garbage to the output outright
Anonymous
For obvious reasons
Victor
when there will be a way to produce PDB with rosbe