Interview with
Ray/tSCc
cxt: Hi ray, I guess you are already a VIP inside the Atari
scene and perhaps everybody already knows your real name, which
is Reimund Dratwa if I am not completely wrong. ;) Hmmm, perhaps
I asked the wrong guy for an interview :) You were born in 1983
and are famous for coding and hmmm drugs? :) You have also been
interviewed by that lamer :) ST Survivor in the very first issue
of Alive and also big Evil and the vanished Moondog added their
share. It seems there are not much questions left unasked :),
anyway what else can you tell our audience about you?
ray: hi there cxt, first of all you're right about my real name
(you may even add in "Marvin" as my second forename) and my
birthyear. Calling me a VIP inside the scene is probably
something I wouldn't underscore as I try to remain humble
whenever possible ;) so deciding is left up to the reader. And
yep that drugs-thing was a hoax made up for the dhs IRCgallery
as you obviously seem to have noticed. Usually I am not quite
into drugs; at least I'm used to strictly keeping away from
chemical substances. So typing "coding, drugs" into dhs' doing-
field was basically because it's called "doing drugs" and it was
the first thing that popped up into my mind at that time. I just
felt to add something more than just "coding" like most others
did, but being a coder was the only true statement regarding my
Atari aspects, I could also have typed "girls" or whatever ;).
As you already said it, there aren't many unasked questions left
and it's hard for me to make up something as I'm known as a
rather elliptical and at most two-dimensional person but I will
try, though. Sadly there isn't much left to be said when it
comes to my Atari activities. I still got lots of projects going
and too few time left so that my machines' harddisks are full of
many unfinished frameworks which might be a common problem to
coders. It had been worse in the past as sometimes I stayed up
most of the night since I just couldn't let go trying to realize
dozens of ideas floating through my mind which resulted in
overstraining myself. Becoming organized, sticking to ToDo-lists
and setting priorities is one thing I had to learn in the past
few years. I believe that is the sort of work getting you
somewhere even if you have to curb yourself all the time and
abolish lots of ideas, at least it's the way working out best
for myself at the moment. I also like to share my ideas a lot
and that's what I had been writing those articles on my website
for. To me this is a creative way of resorbing my ideas
somewhere instead of just throwing them away, it's something I
can finish without consuming very much time as opposed to for
instance a large coding project. Enough said about me? :) The
last thing to mention is that I wouldn't see why to call STS a
lamer, but I guess you were kidding
cxt: You have finished your military service some time ago and
it seems you are living in the city of Aachen now, what are you
doing there?
ray: Yes that's right, I'm currently living in a far west German
city which is called Aachen and famous for "Charlemagne" and
nowadays its university especially for being good at teaching
technical subjects, which is why I went there. I'm busy trying
to obtain a degree in technical computer science which is
closely interconnected with electrical engineering so it goes
together with my interest for computer hardware and low-level
programming well and I'm quite okay with that. Besides that
Aachen is a nice place to be and it isn't far to Belgium and the
Netherlands. (gets himself a cup of coffee…)
cxt: When did you start to work with computers? What was your
first computer?
ray: I think I mentioned this during another interview. To
answer your question it must have been at the age of 12 when I
got my first machine which was my uncle's Mega ST he gave off
after he had brought himself his first peecee. He had been using
it as music-equipment along with "creator/notator" which never
was my cup of tea. The first things for me to do with my Atari
back then was mainly about playing PD games like "Ballerburg
ST", "Arka" and so on or Lucas Art adventures like "Loom" or
"Monkey Island" since they ran on my unmodified SM124 (I had a
colour emulator but that one - I think it was called Panther -
never appeared to be very useful, especially with games as one
would expect). The actual "work" started after I first started
looking into the subject of my Omikron Basic manual, which came
along with the Mega's language-disk when I started to try
programming some most likely not very useful things. This is the
reason why I had become more familiar with OM-Basic rather than
GFA in the first place because I simply didn't have any GFA
interpreter/compiler :).
cxt: It seems you are still quite young and missed the so called
golden Atari days, why did you start coding on a dead platform?
ray: idealism, dunno? And maybe it was due to the Atari being my
first and only computer which to me always made it appear more
decent than any peecee or Amiga some of my friends had even if
Atari had yet "died". Call it devoted or whatever, I think it
had been merely the spirit that keeps the rest of our scene
alive nowadays. I also thought that 68k would be a better and
easier assembler language to learn than for instance 80x86 even
when I already had my own peecee. My dream of porting
Wolfenstein 3d to the ST which aroused maybe in '97 or something
must have been another important point in keeping this notion.
When I first discovered that an active scene of Atari-
enthusiasts still existed everything seemed clear to me: like
earx would put it, "stay cool, stay Atari!".
cxt: Hmm, these words have become quite popular. I guess Edo was
the first one to use them but I might be wrong about that. You
joined tSCc around 2000 and the story about you becoming a
member leaves me with the impression that the communication
inside tSCc is not the best sometimes :) What do you do to keep
all members up to date? Is there still a scheduled meeting or
something?
ray: Yes, you might be right. In my opinion this is a generic
symptom to our scene rather than something very unique to
.tSCc., though. To stay in touch and keep each other up to date
about crew-internals we have our own mailing list. We used to
hold a weekly IRC team meeting, too which has in fact died out,
eventually.
cxt: I heard that you also play drums and guitar so you have at
least some background when it comes to music. In the past you
also seemed to have a Mad Max alike haircut :) Have you ever
thought about doing modules or chiptracks?
ray: In case of interest you may check some of a band's songs I
play drums in over here: http://www.carpet-records.de *big
advertisement* ;). Now believe me or not, I even thought about
trying to compose some YM stuff since I had been challenged with
"programming" a drum machine and typing various drum tablatures
several times, so I might be at least basically familiar with a
chip-/module-tracker's layout and functionality only that in a
tablature-editor tracks are arranged horizontally instead of
vertically. Call me ignorant and lazy but the last time I tried
I got completely lost in the editor of Gwem's definitely
excellent MaxYMizer (maybe I should just read the fucking manual
before, next time :) ), nonetheless. What I've come to find is
that just stepping behind the drums and play a certain pattern,
groove or whatever is something pretty different from composing
a chiptrack it's all a matter of feeling into it which I can't
quite manage in the latter case, so I'm quite pessimistic about
me ever becoming a chip musician ;). Please feel free to take
this conclusion as my big respect I have for the great
chip/module composers that I know.
cxt: You started coding on a Mega ST with Omikron Basic like you
mentioned above and went on to Turbo Assembler later. What's
your favourite Atari development system and which assembler do
you use nowadays?
ray: I eventually went from Turbo Assembler, which provides
incredibly fast assembling speeds and a comfortable and very
stable testing environment (i.e. in my opinion the still best
68k debugger, bugaboo), to devpac due to technical reasons.
First of all I always deplored that even despite its Part-
feature TurboAss doesn't provide any decent ability to easily
and reasonably split up your code into modules which is bad
especially when it comes to coding greater projects such as
Wolfenstein 3D. Secondly, TurboAss does neither support FPU code
nor macros which I wanted or needed to use sometimes (Beams,
some TT-stuff…). Additionally, bugaboo refused to run on my TT I
got back in 2000, so I finally grew into using devpac's gen.ttp
with q.e.d. as an excellent editing package with syntax
highlighting and other useful features. I recently played around
with gcc a bit which seems cool and awfully slow on an unboosted
Falcon at the same time, but oh well, you asked for development
systems.
cxt: It seems your revived tSCc when you joined em, how do you
feel about the echo on beams and the fame this demo has gathered
even beyond the Atari horizon?
ray: Even if it isn't up to me to evaluate if I revived .tSCc.,
I am in fact very satisfied with the feedback we got for "beams"
(even if I wouldn't exactly call myself satisfied with the
result codewise) even in spite of that long delay since the
first presentation at EIL 2003 (the 3d engine and first screens
were coded in late 2001 and the demo was actually planned as a
Mekka&Symposium 2002 release) I wouldn't have expected the
amount of appreciation we got….maybe this is a point to spell
out my thanks the scene, if you want to put it that way. I'd
also like to take this as a chance to encourage people to finish
their stuff (reproachlessly waving hello to ultra and cream)
after years, as it had obviously turned out as such a great
relieve, after all. I'd even say to feel a bit proud of our
achievement.
cxt: You have worked several years on beams, and it was just
released when lots of people started to call it vapourware :),
what beside your military service time caused this huge delay?
ray: After my military service I was a bit out of practice as I
hadn't done much coding during that time. I also had to dig
through my own old source code trying to understand the lines I
had written down once which is always a problem when returning
to work on "old" projects which didn't exactly increase my yet
low motivation at that time. That way I got stuck several times.
I guess there were also various other personal issues such as a
general low that I had been encountering during spring/summer
2004.
cxt: It seem you believe some screens in beams could have been
done faster in retrospective, so I guess it's not optimized to
death?
ray: No, beams surely isn't optimized to death, at all. That
wasn't my main goal anyway; finishing the demo had gotten
highest priority. I would have probably have had to rewrite
everything in order to achieve the best performance, visual
results and on most of the screens, but that way beams would
more likely have ended up a death project - this is why I said
I'm not quite satisfied with the final product code-wise as lots
of improvements in the implementation still could have been made
but I didn't want to push the release date even further. On the
other hand there are things that suffer from the falcons own
technical boundaries (e.g. a creepingly slow 16bit bus) such as
the fullscreen 3d scenes - keeping in mind the full demo runs in
320*200 pixels and the fact that those scenes are being mapped
by the BLiTTER much faster than a cpu routine in the standard
Falcon could draw gouraud-shaded polygons, their speed gets
primarily limited by fillrate than anything else.
cxt: Beams runs on a standard Falcon with only 4 MB which is not
the smallest but perhaps the most common configuration. Was
there a reason to stick to that configuration? Why not go for at
least 14 MB? You could have avoided some problems in that case,
like realtime streaming the MP2 from the floppy for example.
ray: well, idealism and challenge are two words I could mention
regarding this question. I believe there are still some people
who would appreciate to watch demos on their real machines
instead of watching recorded mpeg video of those or who plainly
don't have any 14MB board, so this is where the idealism taps
in. one of the concepts for beams was that we would try to
refrain from non standard features like an FPU, accelerator
boards and just those 14MB of RAM, which is also stated during
the initialization ASCII-screen. So I hope our demo basically
manages to suit those ideals. As sort of a by-product I had to
repatch nocrew's mp2 replayer to support d2d streaming replay as
known from toys' 2001 EIL entry "wait" (but their sources had
gotten destroyed during a hdd crash). Maybe one can use it for
future productions or something; you can download it at
http://dhs.nu/files_source.php
cxt: In the past you haven't been too fond about accelerators.
Since you own a ct60 yourself now, why did you change your mind?
What do you think about the ct60? Is it a good thing that allows
us to do create decent effects in demos or is it a bad thing
splitting an already small scene into even smaller sub-scenes?
ray: that's a hard one, at first, I finally had changed my mind
once I moved from the TT to the Falcon realizing that even if
this machine is capable of rendering stunning demo effects (also
without any accelerator) it tends to not to be too responsive
when it comes to more general works such as coding (especially
with the yet quite slow devpac assembler), driving a word-
processor, image processing etc. - I just feels very slow and
sometimes too unstable for tasks like these. As mentioned, it
badly suffers from its 16bit bus. So buying a ct60 merely helped
improving "everyday-work" on that computer. To refer to demo
coding type arguments, I believe today that it is as well
challenging to exploit a powerful (as opposed to the standard
falcon) '060 accelerator's features with new upper boundaries
set. I'm curious what the future might bring. And without
contradicting to what I've said before I don't think that it
will split our scene into small sub-scenes since nearly everyone
is able to watch most demos recorded as mpeg video. It doesn't
change the fact that I still think there should be separate
compos, i.e. I'm one of the guys contending it to be unfair to
put '030 and ct60 entries into a single compo.
cxt: You seemed quite interested in the TT some time ago. AFAIR
you were also in league with some other guys trying to open
borders on that machine. Did you get any useable results in that
direction?
ray: Actually, I'm still quite interested in doing stuff on my
TT, but time is rare and atm. I am quite busy with finishing
Wolf3d. Hence I'm rather keeping it in the background a bit.
Regarding your question upon opening borders on the TT the only
things I could discover where that performing resolution instead
of sync switches (which aren't possible with the TT-shifters
fixed vsync frequency anyways) at critical raster positions
didn't lead to any success and that there are 2 undocumented and
very strange display modes (modes 0x03 and 0x07 otherwise marked
as "reserved") that produced some sort of bitplane repetitions
(due to the lack of a TTM I couldn't verify the result on a ECL
screen). Another promising trick that had be common on the STE
which used the hscroll-registers to open the left border
unfortunately didn't produce the expected result on the TT,
either. If you find any trick that opens one of the TT's borders
let me know. I am quite dubious though, since the TT-shifter
internally seems to be pretty diffrent from the ST/STE's
shifter.
cxt: Besides your Mega ST, the Falcon (ct60) and obviously the
TT do you have any other Atari machines? Are you a collector
like many other sceners?
ray: Nope, but with a bit of luck I'm gonna get hands on a BJL-
equipped Jaguar soon, which would be could since I would also
love to develop Jaguar-stuff. Its hardware looks so promising
and very interesting from a coders point of view (2 strong RISC
processors, a very powerful blitter, a multiprocessor
architecture in general - lots of hardware bugs on the other
hand). I'm very curious at what one could achieve on that "box".
Besides that I don't prefer collecting vintage items too much as
for the space it takes. I'm already quite well served with the
Ataris I got. I still have a PowerPC (604e) based workstation I
can't quite use due to lack of any development environment (I
couldn't yet get Linux to run on it). Atm. it runs the PPC
version of Windows NT 4.0 which doesn't appear very useful
either. I'd consider this as the only far Atari/Motorola related
collector's computer-item I got atm...
cxt: Is there a specific piece of Atari hardware you would like
to own? I think I saw you looking for a Crazy Dots VME card some
time ago.
ray: I actually swapped my Crazy Dots with DBug against a Nova
VME which didn't suit my needs too much. Disliking the external
casing too I finally got rid of it. What I could actually use in
my TT would be a Riebl VME card with RJ-45 converter but apart
from that, there isn't anything in particular I would like to
own, no (A Supervidel, maybe *nature-nag-nag*).
cxt: What do you think about Atari emulators? Do you consider
the current ones like STEem or SainT useful or are they killing
the true Atari spirit? Would you like to have a full scale
Falcon emulator?
ray: Sure, a full featured and useable Falcon emulator would be
cool stuff as no one seems to have yet actually taken the
challenge. But I am one of the guys who prefer to work on the
real machine and because I feel I'm with many others who stick
to their beloved computers I do not fear emulators could once
kill the spirit, even if I'm avid at what for instance STEem is
yet up to.
cxt: The Wolfenstein 3D port is one of your current projects,
which should hopefully be finished when this issue of Alive is
released :) How much time did you use to make that conversion
happen?
ray: I'd estimate it to be more than 5 years of hard on and off
work as I started the first vague conversion attempts in 1999
(still under that rd-development label), dunno if I still can
dig out the first pre alpha v0.1 engines…I think moondog could
still have a copy of it. It must have looked awful, hehe ;).
cxt: What were the most difficult parts in that port?
ray: I guess keeping at the game at all must have been the
hardest task, since I was the only coder involved in it.
Optimizing most of the time critical main loop "atoms" and
especially extending the game's 3d engine feature by feature
without sacrificing a playable speed on a humble 8MHz ST, that
was very hard work too, yeah.
cxt: So what's your next project? Do you plan to add textured
floors and ceilings like in Doom or is Wolfenstein 3D a dead
end?
ray: You should rather consider Wolfenstein 3d as a dead end.
After such a long development period I wouldn't ever want to do
something like that again ;). I don't plan to extend it by
textured floors and ceilings at that would simply drop the speed
to an unplayable rate on the ST. However, mods could be a
solution in order to reuse the game engine. Some of the 3d
engine's features (the sprite and wall-slice scaling routines,
sound mixer etc. in particular) could prolly be reused, so I
don't hold a ST Doom conversion to be impossible (without
textured floors or ceilings though). But I surely wouldn't be
the one for such a project as the amount of development and
conversion tasks would be even greater. If anyone wants to
volunteer, I could provide my source code at least ;).
cxt: In the past you published some nice articles about various
coding topics. Now you seem to aim at digging deeper into dsp
coding. Does that mean we can expect more articles for your site
or perhaps even for a more or less famous diskmag? :)
ray: Sure, always expect more articles as long as I find new
things to work on. As I said, I like to share things I've
learned/figured out or whatever, so I don't see anything that
would argue against releasing new coding articles or gems, via
diskmag or my website or whatever except for projects occupying
too much of my time to write articles.
cxt: Last year you mentioned a 3D API for 060 based machines,
what happened to that project? Will it be compatible to any
other 3D API on Linux or Windows machines?
ray: That one is still in the pipe, yeah. It is going to be
called SirGL and meant to be a more general hardware abstraction
layer providing 3D primitive outputs, a geometry and a material
library for Falcons and TTs also exploiting the CT60s
possibilities. It isn't meant to be compatible to any yet known
3D APIs and my goal is to keep it as "lean and mean" as
possible, though. I can't tell about any release date yet.
cxt: As you probably know Alive is quite famous for silly
questions and especially for the unavoidable Alive brainstorming
test, since you have already passed it once, I will make this
one a bit shorter ;)
ray:
R:ays
A:tari
Y:achting-Club
O:riginally
F:orbids
T:ransatlantic
S:hipping of
C:ommodore
C:omputers
cxt: Seems to be the fastest performance to far :) If you have
any final words, something that needs to be scribbled onto your
tombstone or if you just want greet your girlfriend feel free to
add anything you like.
ray: Well even if this is without any chances of success I feel
in a funny mood: unfortunately, I'm without any girlfriend at
present, so if you are a nice, between 20-25 years old, sexy and
perhaps redheaded girl who feels capable of caring a bit for a
depressed and lonely crabber feel free to drop me a line via the
usual way ;).
cxt: Ok ray, thanks for all the time you sacrificed to answer my
questions and even more for the time you spent on beams and
Wolfenstein 3D. Keep up the cool spirit and stay Atari...
ray: It was a pleasure to me.
Cyclone / X-Troll for Alive, 2005-10-04
Appendix A
Web
References
Interviews:
Alive 1 - http://alive.atari.org/alive1/ray.php
DHS - http://dhs.nu/scenenews_2001.php
UCM 25 - http://www.tscc.de/ucm25/ray.html
Websites:
Rays Homepage http://people.freenet.de/ray.tscc/
tSCc Homepage http://tscc.atari.org/
|
