Concerned about win32k rewrite...
Moderator: Moderator Team
Concerned about win32k rewrite...
I'm concerned that yet another implement win32k from scratch effort is underway. I'm concerned that
this will set ReactOS back potentially. Hopefully, the existing implementations where hacks aren't
a problem will be used as a reference. How much of ReactOS is tied up in win32k? Why is it easier
to start over than it is to remove hacks? Aren't there coding standards such that hacks are commented
or something similar so that someone later can spot the cheat and reimplement the functionality properly?
The developers can understand and read each other's code, right? This was talked about in newsletter 62,
but what are the ramifications of yet another win32k from scratch implementation?
this will set ReactOS back potentially. Hopefully, the existing implementations where hacks aren't
a problem will be used as a reference. How much of ReactOS is tied up in win32k? Why is it easier
to start over than it is to remove hacks? Aren't there coding standards such that hacks are commented
or something similar so that someone later can spot the cheat and reimplement the functionality properly?
The developers can understand and read each other's code, right? This was talked about in newsletter 62,
but what are the ramifications of yet another win32k from scratch implementation?
Re: Concerned about win32k rewrite...
There remains to be seen some finished rewrite, so maybe this can be considered as another attempt at first rewrite
Re: Concerned about win32k rewrite...
Its just a side project. The trunk remains with original win32 code. Whenever the arwinss rewrite effects will eventually end up in trunk or not, its still open case. Even if its not gonna end up in trunk, its still potentially profitable as a research project/reference that might help us locating and fixing bugs in trunk.
I cant understand what do you mean by tied up in win32k? Win32k contains kernel-mode part of win32 susbsytem code.
Often whole implementation of specific functions depends on one or more hacks. Removing hacks would require to rewrite those functions, and in effect also other functions that would depend on them. Its like with the house made of cards, when you have to change one card at the bottom, without pulling down whole construction.
I cant understand what do you mean by tied up in win32k? Win32k contains kernel-mode part of win32 susbsytem code.
Often whole implementation of specific functions depends on one or more hacks. Removing hacks would require to rewrite those functions, and in effect also other functions that would depend on them. Its like with the house made of cards, when you have to change one card at the bottom, without pulling down whole construction.
-
- Posts: 103
- Joined: Thu Apr 23, 2009 9:45 am
Re: Concerned about win32k rewrite...
I've also got a question
Does Windows have that? With this module, will ReactOS be able to run GNOME/KDE apps? Wow, it would be so cool if it would. No need for the Linux Unified Kerneldll/win32/winex11.drv - implements X11 X-windows graphics and user drivers
-
- Posts: 103
- Joined: Thu Apr 23, 2009 9:45 am
Re: Concerned about win32k rewrite...
How come no? Then what's the matter with this file?
-
- Posts: 581
- Joined: Sat Nov 10, 2007 10:42 pm
Re: Concerned about win32k rewrite...
To run gnome apps, you have to compile and execute a GNOME environment for windows. Such as cygwin.PetruDimitriu wrote:How come no? Then what's the matter with this file?
-
- Posts: 103
- Joined: Thu Apr 23, 2009 9:45 am
Re: Concerned about win32k rewrite...
Then what is this component supposed to do?
- EmuandCo
- Developer
- Posts: 4723
- Joined: Sun Nov 28, 2004 7:52 pm
- Location: Germany, Bavaria, Steinfeld
- Contact:
Re: Concerned about win32k rewrite...
Wrapping enough X11 Crap to more or less compatible API calls to get it run wine's Win32 Subsys which depends on this son of Lucifer. (I still dislike this way, but I prefer the recent Subsys anyway, because arwinss does not even have a working mouse yet )
ReactOS is still in alpha stage, meaning it is not feature-complete and is recommended only for evaluation and testing purposes.
If my post/reply offends or insults you, be sure that you know what sarcasm is...
If my post/reply offends or insults you, be sure that you know what sarcasm is...
Re: Concerned about win32k rewrite...
Hi,
is it possible to take the wine gdi32+user32 in arwinss and rewrite some parts of them to use win32k real window calls (using as base the other win32k rewrite)?
the development could be done in Windows...
Yes, i know its more easy to say it than to do it, but what is the opinion of some developer about this idea?
Its like the current win32k in trunk, but removing the horrible legacy hacks done over wine code in the past, now that some more information about the internal structures exists...
In this way the development could be divided in user mode and kernel mode parts.
Thanks,
Regards
is it possible to take the wine gdi32+user32 in arwinss and rewrite some parts of them to use win32k real window calls (using as base the other win32k rewrite)?
the development could be done in Windows...
Yes, i know its more easy to say it than to do it, but what is the opinion of some developer about this idea?
Its like the current win32k in trunk, but removing the horrible legacy hacks done over wine code in the past, now that some more information about the internal structures exists...
In this way the development could be divided in user mode and kernel mode parts.
Thanks,
Regards
Re: Concerned about win32k rewrite...
Uh, the arwinss stuff are using "real" win32 syscalls.
Re: Concerned about win32k rewrite...
Yup, i mean real windows 100% compatible in parameters and behavior win32k syscalls + windows 100% compatible exports.
Thanks for the reply
Thanks for the reply
Re: Concerned about win32k rewrite...
that requires a lot of research. even trunk's win32k isn't 100% compatible with windows' oneinf_loop wrote:Yup, i mean real windows 100% compatible in parameters and behavior win32k syscalls + windows 100% compatible exports.
Thanks for the reply
-
- Posts: 581
- Joined: Sat Nov 10, 2007 10:42 pm
Re: Concerned about win32k rewrite...
a lot of hacks are in it though, Fireball wants to remove those hacks by doing a re-write and make it implemented the way it was supposed to be.Smiley wrote:that requires a lot of research. even trunk's win32k isn't 100% compatible with windows' oneinf_loop wrote:Yup, i mean real windows 100% compatible in parameters and behavior win32k syscalls + windows 100% compatible exports.
Thanks for the reply
Re: Concerned about win32k rewrite...
It seems I am going to have to provide more information in the next newsletter.
Who is online
Users browsing this forum: Ahrefs [Bot], Google [Bot] and 37 guests