reactos-kernel

Re: [ros-kernel] snapshots

[Casper H. <ch...@us...>]

l=F8r, 2002-06-01 kl. 22:33 skrev Rex Jolliff:
> Hi all,
>=20
> I noticed that old snapshots are eating up a huge amount of the time
> it takes to run backups.  Anyone mind if I delete those older than a
> year.  Also, I noticed that the firewall is currently blocking access to
> FTP on mok.  Does anyone find the snapshots useful, or should I
> just nuke em all?

I find only releases important to keep; a sentimental thing I guess ;-)

Nuke'em all.

- Casper

[ros-kernel] snapshots

[Rex J. <re...@lv...>]

Hi all,

I noticed that old snapshots are eating up a huge amount of the time
it takes to run backups.  Anyone mind if I delete those older than a
year.  Also, I noticed that the firewall is currently blocking access to
FTP on mok.  Does anyone find the snapshots useful, or should I
just nuke em all?


Rex Jolliff
re...@lv...

Re: [ros-kernel] native vs. Win32 - again

[Casper H. <ch...@us...>]

l=F8r, 2002-06-01 kl. 18:10 skrev Yuri Prokushev:
> On 31 May 2002 23:27:55 +0200, Casper Hornstrup wrote:
> >Yes, the graphics APIs of other subsystems can be written as thin
> >wrappers to gdi32.dll and user32.dll. Since these modules perform most
> >of the win32k.sys interfacing, the wrappers should be small compared to
> >the subsystems using win32k.sys directly.
> Well. Now I again don't understand goal of subsystems. Why then don't use=
 ODIN approach? Just make wrappers from pm* to *32 dlls? And how win32 apps=
 will interact=20
> with gdi32?

That is exactly what subsystem developers have the option to do if we do
as I have described earlier.

Or all win32 will be moved to native part?

Only lowest-level Win32 components. The rest can be shared with WINE.

If done as described, then lowest level Win32 components will be tied
closer to the ReactOS kernel than to the Win32 subsystem.

- Casper

Re: [ros-kernel] native vs. Win32 - again

[Yuri P. <pro...@fr...>]

On 31 May 2002 23:27:55 +0200, Casper Hornstrup wrote:

>l r, 2002-06-01 kl. 00:16 skrev David Welch:
>> On Fri, May 31, 2002 at 02:49:05PM +0200, Casper Hornstrup wrote:
>> > Maybe you don't like it because you still see user32.dll and gdi32.dll
>> > as Win32 components. Well that all changed when we decided to make them
>> > native. Even such a small change can make a big difference to how you
>> > percieve things. Now, these components have been adopted as _the_
>> > standard user mode interface to the window manager and GDI and they are
>> > no more part of the Win32 subsystem than any other subsystem. They are
>> > subsystem support code. Using these DLLs in subsystems does not mean
>> > that the subsystems cannot have their own look and feel. It gives you
>> > the _option_ of not having to duplicate the user32.dll for each and
>> > every subsystem and you can then focus on getting your subsystem to do
>> > what it is supposed to do; run applications that expect certain APIs.
>> > You also have the option of reimplementing user32.dll if you feel
>> > strongly about your applications look and feel differently from others
>> > on the desktop.
>> > 
>> I did not realise we had decided this! In my view, user32, gdi32 and 
>> win32k.sys should implement the user and gdi parts of the Win32 API and 
>> nothing else. Of course other subsystems can use these to emulate whatever 
>> APIs they want just as they can on Windows NT itself.
>
>Yes, the graphics APIs of other subsystems can be written as thin
>wrappers to gdi32.dll and user32.dll. Since these modules perform most
>of the win32k.sys interfacing, the wrappers should be small compared to
>the subsystems using win32k.sys directly.
Well. Now I again don't understand goal of subsystems. Why then don't use ODIN approach? Just make wrappers from pm* to *32 dlls? And how win32 apps will interact 
with gdi32? Or all win32 will be moved to native part?

Re: [ros-kernel] native vs. Win32 - again

[Casper H. <ch...@us...>]

l=F8r, 2002-06-01 kl. 00:55 skrev KJK::Hyperion:
> At 00.16 01/06/2002, you wrote:
> >On Fri, May 31, 2002 at 02:49:05PM +0200, Casper Hornstrup wrote:
> > > Maybe you don't like it because you still see user32.dll and gdi32.dl=
l
> > > as Win32 components. Well that all changed when we decided to make th=
em
> > > native.
> >[...]
> >I did not realise we had decided this!
>=20
> "we" didn't decide, Casper did. The others accepted for lack of better=20
> ideas, or they just don't seem to follow the discussions anymore. Speakin=
g=20
> of myself, despite I really care for ReactOS, I chose not to comment=20
> anymore on the topic. I'm losing interest in this sort of things

Maybe it was a mistake to assume that this discussion was over, but
there has not been any messages in the thread "Fwd: Reorganizing the cvs
repository before starting reactos.org" where someone argued against
this for two weeks. I would have provided some links to the list
archieve, but the sourceforge list archieve is screwed up.

I initally wanted to seperate subsystems to their own CVS modules and
have subsystems not depend on each other (ie. put the Win32 subsystem in
it's own module too just like all other subsystems). David then said
that Win32 was inseperable from the ReactOS kernel and that the Win32
subsystem should stay in the reactos module. I did not like the idea of
having modules marked win32 only in the required reactos module, so I
suggested that we make these "inseperable" Win32 modules native so they
don't belong to a specific subsystem and can thus optionally be used in
any subsystem that needs the functionality they provide. Then I was okay
with putting them in the reactos module. David agreed to this compomize
and nobody else has argued against this for about two weeks now.

So please present your arguments if you have some; so we can take vote
or something.

- Casper

Re: [ros-kernel] native vs. Win32 - again

[KJK::Hyperion <no...@li...>]

At 00.16 01/06/2002, you wrote:
>On Fri, May 31, 2002 at 02:49:05PM +0200, Casper Hornstrup wrote:
> > Maybe you don't like it because you still see user32.dll and gdi32.dll
> > as Win32 components. Well that all changed when we decided to make them
> > native.
>[...]
>I did not realise we had decided this!

"we" didn't decide, Casper did. The others accepted for lack of better 
ideas, or they just don't seem to follow the discussions anymore. Speaking 
of myself, despite I really care for ReactOS, I chose not to comment 
anymore on the topic. I'm losing interest in this sort of things

Re: [ros-kernel] Forward: "ReactOS is too limited"

[Jason F. <jas...@ya...>]

Sorry, I meant flame wars as in GUI vs Command Line - which involves
concepts and not people/projects (which tend to happen every now and
again). OS flame wars aren't acceptable, of course. There's no point,
no good comes of it and it can only do harm. If someone doesn't like
Linux they shouldn't tear it down here. There's no official condoning
of that kind of behaviour.

- Jason

--- Casper Hornstrup <ch...@us...> wrote:
> fre, 2002-05-31 kl. 15:48 skrev Jason Filby:
> > Heh, I tried to douse the flames - seems it didn't work! I think
> > ros-general is where the flames belong, but that's down right
> now. It
> > could be back next week though.
> > 
> > - Jason
> 
> Actually, I wouldn't want to see flamewars on ros-general either
> because
> it is still a ReactOS related list and ros-general is great for
> people
> interrested in ReactOS (but don't have the time to participate in
> development) to monitor progress. IMO, flaming other OS's should
> not be
> something this project should officially condone.
> 
> - Casper
> 
> 
> _______________________________________________________________
> 
> Don't miss the 2002 Sprint PCS Application Developer's Conference
> August 25-28 in Las Vegas --
> http://devcon.sprintpcs.com/adp/index.cfm
> 
> _______________________________________________
> reactos-kernel mailing list
> rea...@li...
> https://lists.sourceforge.net/lists/listinfo/reactos-kernel


__________________________________________________
Do You Yahoo!?
Yahoo! - Official partner of 2002 FIFA World Cup
http://fifaworldcup.yahoo.com

RE: [ros-kernel] Forward: "ReactOS is too limited"

[Steven E. <ste...@ya...>]

> So when writing "a new" shell (command interpreter
> in Winspeak) careful
> consideration is needed - what parts of "Unixism"
> have to be thrown out
> to get some consistent result under Windows.

We have the best shell in the world. Its called
rosapps/cmd. We have some of the best from Windows and
UNIX in our cmd.exe and if you think it needs more
then please send a diff and I will apply it.

 The fact that Microsoft cannot make a decent cli
should tell you how well they write software =P

__________________________________________________
Do You Yahoo!?
Yahoo! - Official partner of 2002 FIFA World Cup
http://fifaworldcup.yahoo.com

Re: [ros-kernel] native vs. Win32 - again

[Casper H. <ch...@us...>]

l=F8r, 2002-06-01 kl. 00:16 skrev David Welch:
> On Fri, May 31, 2002 at 02:49:05PM +0200, Casper Hornstrup wrote:
> > Maybe you don't like it because you still see user32.dll and gdi32.dll
> > as Win32 components. Well that all changed when we decided to make them
> > native. Even such a small change can make a big difference to how you
> > percieve things. Now, these components have been adopted as _the_
> > standard user mode interface to the window manager and GDI and they are
> > no more part of the Win32 subsystem than any other subsystem. They are
> > subsystem support code. Using these DLLs in subsystems does not mean
> > that the subsystems cannot have their own look and feel. It gives you
> > the _option_ of not having to duplicate the user32.dll for each and
> > every subsystem and you can then focus on getting your subsystem to do
> > what it is supposed to do; run applications that expect certain APIs.
> > You also have the option of reimplementing user32.dll if you feel
> > strongly about your applications look and feel differently from others
> > on the desktop.
> >=20
> I did not realise we had decided this! In my view, user32, gdi32 and=20
> win32k.sys should implement the user and gdi parts of the Win32 API and=20
> nothing else. Of course other subsystems can use these to emulate whateve=
r=20
> APIs they want just as they can on Windows NT itself.

Yes, the graphics APIs of other subsystems can be written as thin
wrappers to gdi32.dll and user32.dll. Since these modules perform most
of the win32k.sys interfacing, the wrappers should be small compared to
the subsystems using win32k.sys directly.

- Casper

[ros-kernel] ZwCreateToken implementation...

[Joseph G. <gal...@va...>]

Hello everyone,

I've been working on an implementation of ZwCreateToken(),
and I'm having some trouble that I'm not sure how to tackle
at this point.

Attached is my version of token.c, as well as a program
called tokentest.c to exercise it.

The system crashes when ZwCreateToken() calls ObCreateObject():

Bug detected code: 0x1D
Divide Error Exception: 0(fff8)

I put a DbgPrint() in ObCreateObject before the assert_irql()
call and I don't see the output from it -- so it seems to
be crashing making the call, which seems weird.

I can't figure out what is going on -- it seems
like a calling convention problem or something.

Can anyone see what I've messed up here?

Does anyone have any hints for how to get started
setting up a debugger here?  I'm using the bochs
emulator to run things, and haven't got a clue
on where to look for information about setting
up a debugger.

Thanks,

Joseph

Re: [ros-kernel] native vs. Win32 - again

[David W. <we...@cw...>]

On Fri, May 31, 2002 at 02:49:05PM +0200, Casper Hornstrup wrote:
> Maybe you don't like it because you still see user32.dll and gdi32.dll
> as Win32 components. Well that all changed when we decided to make them
> native. Even such a small change can make a big difference to how you
> percieve things. Now, these components have been adopted as _the_
> standard user mode interface to the window manager and GDI and they are
> no more part of the Win32 subsystem than any other subsystem. They are
> subsystem support code. Using these DLLs in subsystems does not mean
> that the subsystems cannot have their own look and feel. It gives you
> the _option_ of not having to duplicate the user32.dll for each and
> every subsystem and you can then focus on getting your subsystem to do
> what it is supposed to do; run applications that expect certain APIs.
> You also have the option of reimplementing user32.dll if you feel
> strongly about your applications look and feel differently from others
> on the desktop.
> 
I did not realise we had decided this! In my view, user32, gdi32 and 
win32k.sys should implement the user and gdi parts of the Win32 API and 
nothing else. Of course other subsystems can use these to emulate whatever 
APIs they want just as they can on Windows NT itself.

Re: [ros-kernel] native vs. Win32 - again

[KJK::Hyperion <no...@li...>]

At 23.44 30/05/2002, you wrote:
> > > With other words: Has csrss.exe much to do with windowing.
> > No. Csrss.exe handles consoles and access to the blue screen device.
>I propose to rename that server "w32ss.exe". It is no more a c/s run-time 
>and our implementation is not plug-in based (=DLL based).

Emanuele, but I liked the plugin-based approach. Don't you think it's 
better? especially when several subsystems are installed at once?

[ros-kernel] Next import of WINE.

[Steven E. <Ste...@ya...>]

Hola,
This is in response to emails a few regarding creating a shell or perl
script to create the ReactOS Makefiles for the WINE tree. I have lost
some of my email in all of the noise as of late because I'm now on the
mingw,wine,cgywin and reactos mailing lists. Anyone that is interested
in helping or has a question needs to email me again.

If anyone wants to hack on wine under Mingw you now can using MSYS with
great results, however there still isnt a good way to intergrate a
current WINE snapshot in the ReactOS build system. We need to make a few
changes to the ReactOS makefile for building WINE as our Win32
subsystem.

1. Auto generation of the dllname.def's using winebuild.
2. Intergration of the Wine Resouces Compiler in to the build system.
WINE resources must first be compiled with WRC then fed to windres.

Now for the ReactOS to Wine Makefile Converter we need to do the
following things.

1. run ./configure with the mingw options.
2. Generate the Makefile.ros for each dll and program we want to build
from the wine tree. The current Makefile.ros's are in our WINE fork on
mok.

Casper, Rex, Jason: When are you guys going to start on the CVS
restructure? This can wait untill you guys are done and you can go ahead
and blow away the WINE tree on mok. I have a clean copy of the ReactOS
WINE tree with Makefiles. After the script is done and CVS is clean we
can do another import.

Thanks
Steven

"Every revolution was once a thought in one man's mind"
- Ralph Waldo Emerson 

"Every revolution was once a thought in one man's mind"
- Ralph Waldo Emerson 

Re: [ros-kernel] Cleanup

[KJK::Hyperion <no...@li...>]

At 19.51 30/05/2002, you wrote:
>We have several advantages when using/supporting only gcc/binutils. they
>are present on a lot of platforms. This means less work for us when we
>expand to these platforms.
>By supporting only one compiler we don't have to be so strict about
>about following every standard out there or have workarounds for missing
>features of other compilers. [...]

Dangerous attitude. While it's ok to have compiler-specific *code*, at 
least the headers shouldn't be

Re: [ros-kernel] Use or not use

[Frank D. E. Jr. <fd...@ya...>]

Uh...

In a country with no copyright laws, the GPL would have no affect, as
it itself requires the material being protected to be copyrighted.  It
is a copyright license.

In addition, would location in such a country not render the foundation
useless when trying to defend developers in the U.S., or some other
country where copyright laws exist?

Because of this, placing the foundation in a country like that would
eliminate the protection offered by the GPL to the project, and anyone
could take the system and copyright it in some other country or
something, turning ROS into an essentially closed-source system there,
could they not?

--- "Robert K." <ro...@ko...> wrote:
> There are in fact countrys that have no copy right. Especially in the
> near east.
> Some arabian countrys. And I think Taiwan.
> This is also what I suggested. Locate the foundation in souch a
> country, where one
> can buy DatacenterServer for 4 US$.


=====
=======
Frank D. Engel, Jr.
Please note my new address: fd...@ya...

__________________________________________________
Do You Yahoo!?
Yahoo! - Official partner of 2002 FIFA World Cup
http://fifaworldcup.yahoo.com

Re: [ros-kernel] bug in alloca ?

[Casper H. <ch...@us...>]

fre, 2002-05-31 kl. 17:05 skrev Hartmut Birr:
> Hi,
> 
> some programs, which use msvcrt.dll, crash on ros. The crash occurs within 
> alloca. If there is a call to alloca, the size parameter is pushed to the 
> stack, but alloca needs this parameter in eax. What's wrong ?

Could it be because our msvcrt headers define _alloca to use cdecl
calling convention?

- Casper

Re: [ros-kernel] Forward: "ReactOS is too limited"

[Casper H. <ch...@us...>]

fre, 2002-05-31 kl. 16:31 skrev Steven Edwards:
> I would be more then happy to moderate if more of our
> developers/subscribers used #reactos on openprojects.

I would like to be more on IRC, but I don't know when there is a chance
of anyone else beeing on this channel. Could we publish such information
on the website?

- Casper

Re: [ros-kernel] Forward: "ReactOS is too limited"

[Patrick M. <ox...@st...>]

On Fri, May 31, 2002 at 08:40:40AM -0600, Adrian  Sloan wrote:
> I've always thought that win32 isn't trying to achieve network
> transparency in the
> gdi, that seems like an odd place to do it.  Rather win32 is becoming
see win2k terminal server (aka citrix server), doing exactly that

> network transparent
> at the component level first through COM/DCOM and now through .NET.
that would suggest that each and every single control is exported over
COM/DCOM - with every other path (like direct calls) purged or at least
declared "obsolete"...
oh, and btw. - it'll break backward compatibility...

patrick mauritz

[ros-kernel] bug in alloca ?

[Hartmut B. <har...@te...>]

Hi,

some programs, which use msvcrt.dll, crash on ros. The crash occurs within 
alloca. If there is a call to alloca, the size parameter is pushed to the 
stack, but alloca needs this parameter in eax. What's wrong ?

- Hartmut

78017464 <_fcvtbuf>:
...
78017485:	68 78 02 00 00       	push   $0x278
7801748a:	e8 b9 ab 00 00       	call   78022048 <__alloca>
...
78022048 <__alloca>:
78022048:	51                   	push   %ecx
78022049:	89 e1                	mov    %esp,%ecx
7802204b:	83 c1 08             	add    $0x8,%ecx
7802204e <probe>:
7802204e:	3d 00 10 00 00       	cmp    $0x1000,%eax
78022053:	72 10                	jb     78022065 <done>
...

RE: [ros-kernel] Forward: "ReactOS is too limited"

[Adrian S. <aa...@ha...>]

This is my first message to any ReactOS list, so please don't kill me
For its point being so minor.

On Thu, May 30, 2002 11:25 PM, Patrick Mauritz wrote:
>but I'm getting off-topic...

>the win32 GDI is only partially better, it still lacks:
> - network transparency (please in a better, more highlevel way than
this
>   X mess)
I've always thought that win32 isn't trying to achieve network
transparency in the
gdi, that seems like an odd place to do it.  Rather win32 is becoming
network transparent
at the component level first through COM/DCOM and now through .NET.

-- 
Adrian Sloan
aa...@ha...

Re: [ros-kernel] Forward: "ReactOS is too limited"

[Steven E. <ste...@ya...>]

Lets no provide a forum to waste bandwith. While I
would subscribe to it also =), I would suggest instead
we make it part of our guidelines that if you get in
to a heated disscussion on the mailling lists a
moderator can slap you down and make you move the
debate to IRC. 

I would be more then happy to moderate if more of our
developers/subscribers used #reactos on openprojects.

Thanks
Steven

--- Casper Hornstrup <ch...@us...>
wrote:
> This thread is turning into a flamewar. Jason, how
> about setting up
> os...@re... mailing list for these
> discussions? They pop up once
> in a while, and as the number of subscribers
> increase, I expect we'll
> see them more often. I would subscribe to this list,
> but then I can
> automagically route them to the
> not-so-important-reactos-stuff folder
> ;o)
> 
> - Casper
> 
> 
> 
>
_______________________________________________________________
> 
> Don't miss the 2002 Sprint PCS Application
> Developer's Conference
> August 25-28 in Las Vegas --
> http://devcon.sprintpcs.com/adp/index.cfm
> 
> _______________________________________________
> reactos-kernel mailing list
> rea...@li...
>
https://lists.sourceforge.net/lists/listinfo/reactos-kernel


__________________________________________________
Do You Yahoo!?
Yahoo! - Official partner of 2002 FIFA World Cup
http://fifaworldcup.yahoo.com

Re: [ros-kernel] Forward: "ReactOS is too limited"

[Casper H. <ch...@us...>]

fre, 2002-05-31 kl. 15:48 skrev Jason Filby:
> Heh, I tried to douse the flames - seems it didn't work! I think
> ros-general is where the flames belong, but that's down right now. It
> could be back next week though.
> 
> - Jason

Actually, I wouldn't want to see flamewars on ros-general either because
it is still a ReactOS related list and ros-general is great for people
interrested in ReactOS (but don't have the time to participate in
development) to monitor progress. IMO, flaming other OS's should not be
something this project should officially condone.

- Casper

Re: [ros-kernel] Forward: "ReactOS is too limited"

[Jason F. <jas...@ya...>]

Heh, I tried to douse the flames - seems it didn't work! I think
ros-general is where the flames belong, but that's down right now. It
could be back next week though.

- Jason

--- Casper Hornstrup <ch...@us...> wrote:
> This thread is turning into a flamewar. Jason, how about setting up
> os...@re... mailing list for these discussions? They pop up
> once
> in a while, and as the number of subscribers increase, I expect
> we'll
> see them more often. I would subscribe to this list, but then I can
> automagically route them to the not-so-important-reactos-stuff
> folder
> ;o)
> 
> - Casper
> 
> 
> 
> _______________________________________________________________
> 
> Don't miss the 2002 Sprint PCS Application Developer's Conference
> August 25-28 in Las Vegas --
> http://devcon.sprintpcs.com/adp/index.cfm
> 
> _______________________________________________
> reactos-kernel mailing list
> rea...@li...
> https://lists.sourceforge.net/lists/listinfo/reactos-kernel


__________________________________________________
Do You Yahoo!?
Yahoo! - Official partner of 2002 FIFA World Cup
http://fifaworldcup.yahoo.com

RE: [ros-kernel] Forward: "ReactOS is too limited"

[Frohwalt E. <ma...@fr...>]

> If that is the case, could you spell the thing properly, please? How
> would you like it if the fine people at Microsoft started exclusively
> using the term `Weenux' to describe something so many people put time
> and effort into?

Sorry, old habits die slowly ... and, I still strongly dislike Windows for the
company that stands behind it. And if you'll read the Cryptonomicon, you'll
see that Stevenson spelled Windows "Windows" but Linux "Finux". ;-)

> Yup. Note, however, that (imho) X window as a GUI _implementation_ is
> inferior to Windows NT as a GUI _implementation_ (network transparency
> notwithstanding); and I think this has some relevance to the relative
> proportions of `GUI people' and `command line' people among the two
> camps. 

Just nitpicking: X11 should be compared to GDI, since X11 itself is not
a desktop. The correct "rivals" would be KDE and Gnome - and which one
is better KDE, Gnome or Windows' desktop is even more a matter of personal
taste. I'm truly amazed what the KDE and Gnome people have come up with
in the last few years.

Said that, there is the exception of Window management in Windows. One
thing I find very very annoying is the fact, that an application which crashed
or is busy is hard to move "out of the way" under Windows. (Up to NT4 / Win95 
I would have said "impossible", but now the "show desktop" button often
saves the day ...)

As far as I know, this is due to the fact, that a process has to process
"resize / move / hide" messages itself. I prefer the X11 model where the
window manager moves the window without the associated process having a
chance to object. (However, nowadays there a protocols where an application
can ask the Window manager "Would you please ... ?")


>  > - cmd.exe is REALLY REALLY REALLY weak compared with bash or zsh. I
>  > know bash/zsh do run on Windows, but I'd like something "native"
>  > with the same power (programmable command line completion, good
>  > scripting, etc.) No big issue, just a lot of work to write a really
>  > really cool cmd.exe.
> 
> Zsh runs native in win32; it can be configured to be
> comfortable. Integrating it into the system is worth considering--it's
> X11 licenced, as I recall, and less buggy than bash :-) . 

My main gripe about running a Unix shell in Windows is the "\" and "/"
problem: Unix shells use "\" as an escape character. Some Windows programs
accept "/", others dont. And while I prefer the unix model of the shell
extending wildcards many windows programs do not like this. (For example
they accept exactly one argument - which accepts a wild card.)

So when writing "a new" shell (command interpreter in Winspeak) careful
consideration is needed - what parts of "Unixism" have to be thrown out
to get some consistent result under Windows.

(What I mean is: I want the feel of cmd.exe, since this is Windows not
Unix, but I want programmable completions, working pipes, (nearly)
unlimited command line lengths, scripting, whatever ...)

> I don't think a lack of command line interfaces will really be a
> problem :-) , given that the compiler of the system is gcc, not VC++. 

I don't think so either, just because Reactos includes source :-)


Cheers,
   Froh

Re: [ros-kernel] Forward: "ReactOS is too limited"

[Casper H. <ch...@us...>]

This thread is turning into a flamewar. Jason, how about setting up
os...@re... mailing list for these discussions? They pop up once
in a while, and as the number of subscribers increase, I expect we'll
see them more often. I would subscribe to this list, but then I can
automagically route them to the not-so-important-reactos-stuff folder
;o)

- Casper