[ros-kernel] WineConf update

Vizzini vizzini at plasmic.com
Mon Feb 2 11:54:46 CET 2004


After a slippery drive home (12 stressful hours from Minneapolis ->
Kansas City, rather than the six it took me to get there on Friday), I
finally have time to report a bit on WineConf.

First of all, I'd like to thank Codeweavers on behalf of those of us
from ReactOS that attended the conference.  It was really run well, with
ample free food (always a requirement).  I didn't even mind -5F as much
as I thought!

There were about 35 Wine and ReactOS hackers there, and two days worth
of presentations.  I started getting serious questions about ReactOS the
instant people figured out who I was.  There was about a 50/50 split
between people wanting to know what it was and people wondering why we
are bothering to do this.  There were a lot of interesting (and
passionate) conversations about the role of Linux on the desktop, the
stupidity of trying to clone the windows kernel, intellectual property
considerations, etc.  I can say that in all cases, the discussions were
very positive and open.  I'm certain that I didn't convince everyone,
but just the opportunity to make the case was worthwhile.

The ReactOS presentation on Sunday morning went as well as can be
expected, given the state of our code.  I spoke from a prepared deck
(http://plasmic.com/~vizzini) for about 15 or 20 minutes, and then we
demonstrated various aspects of the system.  Steven ran the
demonstrations, which included showing off Explorer, Windows XP Notepad,
and our installation system.  Then, the four of us (Steven, Art, Mark,
and me) took questions.  Most of the questions this time around were
very technical in nature, and there was at least one question (from
Gavriel State, regarding our DIB engine), that pretty much stumped us
(can one of the UI guys please comment?).

We had two crashes due to the floppy driver (well, also due to my
stupidity at not configuring VMWare with a blank floppy disk image
before starting), and one non-detected mouse incident.  There was also a
bit of a repainting problem, but on the whole, the system represented
itself well.  

There was a very interesting question brought up on Saturday morning by
one of the Wine guys (Mike perhaps?) regarding running ReactOS inside
Linux, like Win4Lin.  I was skeptical at first, but having slept on it a
day or two, I think this is a really cool idea.  The proposal is that we
adapt our own ntoskrnl.exe and hal.dll to run correctly in user mode,
just like user mode linux does, and use them to run the rest of a native
Windows XP system.  In other words, it replaces Win4Lin with free
software, and it would be compatible with the current Windows OSes. 
Much investigating remains to be done, of course - I can't really even
guess at the level of difficulty or time required at this point - but it
is worth pursuing in my opinion.

Jan Kratochvil also joined us to present about his Captive NTFS
project.  For those that don't remember what this is, Jan got the
ntfs.sys driver from Windows XP running in Linux, in part by using
chunks of ReactOS code.  Jan and I had some offline discussions as well
about bringing Captive's improvements back into our tree, and about
other captive-like projects that could be implemented.  I expect a lot
of focus will be applied to this area of Quasi-Wine, Quasi-ReactOS
hacking over the next year.

All in all, it was a really great conference, and I'll definitely try to
attend again next year.  Presentations were good, conversations were
interesting, people were friendly, and Minnesota at -5 was beautiful!
;-)

 -Vizzini




More information about the Ros-kernel mailing list