[ros-kernel] Q2 Bug (210 in bugzilla) Fix or bigger bug?

[Waldo Alvarez Cañizares]

Sure:
 
That could not be right, that's true and wasn't meant to be final, that was to call the attention and to show that the problem seems to be in fact a lot bigger, take a look at the subject (a bigger bug?). Provably lot of functions rendering graphics need to check if the device adapter does display images the other way. Imagine if the driver implements some of them one way and win32k the other way. That would be a funny mix of garbage on the screen. The problem is seems things were not implemented with that in mind (I'm in new ground when it comes to Win32k) beleive me, it was shocking to realize that dibs could be upside down. And in my opinion it will probably will take some time to fix that. Now that we know, we could put the wrong fix for a while (properly commenting a warning in that code) and meanwhile the routines that should check for it could be fixed. That way there will be people showing Quake II to others on top of reactos. It is a common way to see people that knows about reactos programs running on top of reactos. I've been talking with some people here about reactos and the main questions and reactions are all similar.
 
- People that knows little about computers: Does it have windows and buttons like in Windows? :) It is graphical? Do I have to spend time learning to use it?
 
- Non Fans (people open to new things): What is the size of applications that run on top of it? When I say Quake II almost runs. They say "Ohhh seems that things are going ok and it is advancing very fast"
 
- Microsoft fans: What! a windows NT clone? That's a sh... When I say, yes but it is open source, it will contain less bugs and probably will be faster than Windows and if it is not fast enought you can change it to be faster. (a silence, no more replies)
 
- Linux Fans: Ahh, but a windows clone will contain the same design flaws as Windows. And I say. Yes ok tell me wich flaws? They just say "flaws, flaws"
 
The thing is. It will do the trick if it takes long to fix.
 
Best Regards
Waldo Alvarez

________________________________

From: ros-kernel-bounces at reactos.com on behalf of Filip Navara
Sent: Wed 3/3/2004 5:33 AM
To: ReactOS Kernel List
Subject: Re: [ros-kernel] Q2 Bug (210 in bugzilla) Fix or bigger bug?



Royce Mitchell III wrote:

>>  bm.bmHeight = abs(bi->biHeight);     <------- Change by this
>>  bi->biHeight = -bm.bmHeight;         <------- + add this to signal
>> programs the way we render
>
> This change can't be right. It is throwing away information that the
> DIB is upside-down. We have no choice but to handle it in the DIB
> rendering code. Any other approach is doomed to failure, IMHO :)

Yeah, this fix isn't correct, but I don't agree that we need to handle
it directly in the DIB code. The only thing that we need to do is to
pass a correct values to the DIB routines. This means correctly
calculating the surface "start" pointer (SurfObj->pvScan0) and
positive/negative "pitch" (SurfObj->lDelta), the code itself then can
use the same pointer arithmetic for traversing through top-down and
bottom-up bitmaps. AFAIK one of the routines that needs to be fixed is
EngCreateBitmap.

- Filip

_______________________________________________
Ros-kernel mailing list
Ros-kernel at reactos.com
http://reactos.com/mailman/listinfo/ros-kernel


-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/ms-tnef
Size: 6637 bytes
Desc: not available
Url : http://reactos.com:8080/pipermail/ros-kernel/attachments/20040303/a4ab36d4/attachment.bin