SvOlli
SvOlli was a German Dreamcast developer and he released DCStella. We have an interview with him on our site.
This interview was by DarthEmuFan on 3rd March 2002 for ConsoleVision.com.
Consolevision: Where were you born, raised, personal details, etc?
SvOlli: Born and raised in a small town called Lehrte near Hannover, Germany (the town of the Expo 2000 and the world's biggest computer fair, the CeBit). Moved to Hannover in May 2000. Only child, born '72, zodiac sign pisces, not married.
Consolevision: What qualifications do you have? College degree's?
SvOlli: Right now I'm writing my diploma thesis, but I already have several years of professional experience from jobbing which kept me off my diploma. Jobs includes ones from administration stuff to coding at Germany leg of the OS manufacturer QNX.
Consolevision: What got you interested in computers?
SvOlli: Hard to say. I've been hooked on tin-soldering by watching my dad. But I'm not nearly as good as him, though. After getting an Atari 2600 VCS, I was fascinated by the idea of the "virtual worlds" though they weren't called like that back then, in the C64 I saw the perfect combination of playing games and toying with technique behind it. That facination for the technique faded a bit away when the C64 went way out of state of the art. The last time I really used mine was 1992. From there I went to AMIGA, but couldn't really make good friends like I did with the C64, so my interest faded away another bit. It came back, when I got first contact with Unix (SunOS 4.1, to be more precise) on my studies at the univerity. Mostly the whole networking / remote stuff I got hooked on. I learned that there is a widely source code compatible free Unix for installation on PCs, called Linux (1.0.9 was my first self-compiled kernel). That got me into buying a PC, which I hated back then for MS-DOS and Windows 3.1.
Consolevision: What project(s) have you done in the past?
SvOlli: I'm still listed as a developer of sorcerer, a Linux distribution that's based on compiling source code when installing a package instead of using precompiled binaries. I put my work there on hold 'till my diploma's finished. In general you could say that I'd rather like contributing to other peoples projects, than beginning (and finishing) my own. If I think the software I use is lacking a feature, I'll take a look if I can hack it in there myself instead of asking the author to do so. The times I succeeded I sent the patches of my work to the original authors. Isn't that just a wonderful thing about open source software? I did that on several Unix utilities, including wmx, an X11 window manager.
Consolevision: What made you get into the Dreamcast Homebrew scene?
SvOlli: For gaming I got myself a PSX when the prices got reasonable. That one I built in an arcade cabinet once, running Namco and other museums. But then came the time I noticed that the PSX is too limited for a good allround emulator. My attempts on putting a PC in there resulted in two wrecked keyboards, so I looked for an alternative. I was playing Soul Calibur and other DC titles with a friend of mine, and he told me Mame is also available for DC. Since DCs are on the runout, they are cheap, so I bought one with Soul Calibur. Looking for Mamed I stumbled over the terms "coders cable" and other stuff concerning homebrewing software. DCStella was the very first thing I did on homebrewing after toying a bit around with KOS and dc-load and stuff. Seeing [d]cs2600 I was disappointed. I knew Stella for Linux and I liked that one much better, so I assumed that someone must have already ported it. Much to my surprise, my search was fruitless. And like every qualified villian states in the movies: "If you want get something done, you got to do it yourself." So I started to analyse the source on how to replace the screen output by the one on the DC. By just using the frame buffer it was not a very hard thing to do. The prove of concept was made. There was the birthday of the guy who got me into the DC in the first place was about 3 weeks after of my first sight of a running emulation. So I hacked the v0.1 (called alpha1) back then together quickly, and on his birthday I presented him a rather unimpressive 3"-CDR with the Atari Logo as the title saying: "This is the Stella Emulator for Dreamcast. I just ported it. This one here in my hand is the very first copy of it. Happy Birthday." - "Just a moment. Does this mean I can finally play the old Atari games in rgb quality like I always wanted?" - "Yep." - (then there was a twinkle in his eyes) "Cool."
Consolevision: What made you choose to do an emulator for the Atari 2600 and how complex is it a machine to emulate?
SvOlli: It was always the wish to have a DC version of Stella. I wanted to play some games of my childhood again, and I knew Stella for several years, I already poked around in the source once in a while when I got the time and was curious about some details of one or another 2600 game. So I knew the source was well organized and highly portable. And that's the point: I just did a port not a complete emulator. So I don't know a lot about emulation, the whole emulation just looks like a precompiled library to me that's called once for initialization and again once for every frame to process.
Consolevision: Would you say that it is an easier to port emus over from linux rather than rewrite them specifically for the dreamcast?
SvOlli: Definitively, yes. Graphics, sound and input need to be rewritten, but the core is still the same. This also can turn to a downside: I'm using some rather generic and not DC optimized source, so it'll be slower than a tweaked one. Have a look at the DreamSNES emulator: the machine they emulate is much more complex and it runs at roughly the same speed as DCStella. Awesome job done there.
Consolevision: What programs have helped you bring Stella to the Dreamcast?
SvOlli: About the same all others help to bring their software to Dreamcast: the gcc-toolchain done by M.R.Brown, AndrewK's dc-load, Dan Potter's KOS and libdream. A texteditor for writing / modifying the source, nothing far apart from the tools I use for writing Unix sources.
Consolevision: Do you have plans for updating DCStella or are you happy with what is an exceedingly good emulator already?
SvOlli: Though I'm quite pleased with it like it is now, I still have some plans, but most of them I hold of until I finished my diploma thesis, since they could prove to be rather time consuming. First of all I want the VMU image to be also themeable, probably even themeable for each ROM loaded. Second would be a monitor to go searching for cheats, because I want to play Solaris to the end someday. There are also open issues to be solved, like the paddles and the sound that's just not like the original Stella. I'm still curious of the feature requests that will come in the future, like the one ROM autostart feature would never have gone in, if nobody asked. But there are already some requests that I had to turn down, like support for 7800 games, since I only do (and only want to do) the port of Stella, there's no real ambition to supercede the emulation capabilities of the original Stella. Well, I've implemented an illegal opcode in the original Stella sources though, but that's an exception to the rule and was also rather easy to do. But these changes have already been submitted for inclusion on the original Stella.
Consolevision: The themeable VMU images, surely that is a massive undertaking, would you seek help from the dreamcast scene to make them?
SvOlli: Not really. That's something I'd like to hack in there myself. On the other hand if someone reads this, likes the idea and sends me a patch, it will go in there and save me work. But that's not help I seek out for. Where I really could use help is the darned paddles routine. It refuses to work and after several days of bug-hunting I still have no idea what's broken. Maybe I'm blind for the obvious, maybe it's some typecasting or endian problem, I just don't know. Finally one could say that porting Stella has is small enough for a project where someone could do all by himself, but it's still big enough for others to offer their ideas. That's one of the ideas in releasing the source: if someone has a cool idea and some skills, he can download the source, see if his idea you work, without the need to ask me for anything. That's something that bugs me about most of the DC homebrew releases: I have an idea for a feature but fail to see if it's really doable, since I can't take a look at the sources. If it wouldn't be for Dan Potter, each homebrewer would be forced to reinventing the wheel.
Consolevision: Whats your favourite Atari 2600 game and why? :)
SvOlli: I don't have that one game, since my taste changes quite often in details, but it's always something that's unique to Atari 2600 in idea and gameplay like Adventure or Haunted House. Or maybe some of these cool party two player games like Combat, Outlaws or Pele Soccer. The game I will be spending on most time in 2002 is going to be Solaris, since I want to beat that one with all it includes: drawing maps and stuff :)
Consolevision: Would you consider doing any other projects for the dreamcast or for that matter the Game Boy Advance for that matter?
SvOlli: Yes, absolutly. I want to do my own mp3 player using KOS and the mp3/oggvorbis libraries, since I want one that's also usable without the need for a monitor, so all output will be displayed on the VMU. Linux on DC is also something very tempting for doing a quick hack like porting FreeCiv, for example. But setting up a cross compiling suite for Linux is a bit more difficult than for a bare cross compilation kit, like it's needed for KOS and libdream. And Linux would be also a quite nice platform to see if porting of software would make sence. I'm thinking about VICE for example. If that Commodore emulator runs painfully slow on the DC one can skip the try in porting it to KOS. The other ideas I have are rather the result of sheer need, like I need some software for backing up the VMU, so if no one does it until summer I'll hack a tool together that allows you to save and load your vmu over dc-load. About the GBA I'm not so sure, I guess if I want something portable to toy around with, I'd rather go for a Linux-PDA.
Consolevision: Wow, you have some great projects and we wish you luck :) What's your opinion of the Dreamcast scene and the fact that through people such as your self we the fans can play great vintage games on a state of the art console?
SvOlli: Thanks. When Sega chose the name Dreamcast I'm they didn't think of it as the "Dream" of the homebrew scene. But imho it is, it's the platform with the best access for homebrewers and people using their work: no modifications needed for booting homebrew CDs, developer's cable for $25 and enough infos on the web to get started without bothering anybody. That's what I did, that makes it a dream. About the fans enjoying my work I'd like to say that you're the one pushing the project forward. Over 1300 diffrent machines (ips) from the internet downloaded DCStella V0.2, V0.2a and V0.2b from my site in February. That's encouraging. Overall it's a nice community to participate in, though some more homebrewers wouldn't hurt. The only sad thing is that at the end of 2002 it will be dead on the commercial scene like any other discontinued system. History repeating itself there: bad concepts seem to sell better, just take a look at VHS, Microsoft and even Sony's going to get in line with them. Someone once said that the Dreamcast is like the Amiga of the console world. There's no comparison on the Dreamcast that could be more true and that gives me hope, since no one dies harder than an Amiga die hard, or a DC die hard for that matter :-).
Consolevision: For the last question: What would be the emulator/homebrew software etc you would most like to see appear on the Dreamcast?
SvOlli: VICE! What's lacking now is a good C64 emulator, and VICE is the best portable of them all, just one single assembler line (and that one's in the DOS port), and already ported to six platforms, have a look at http://viceteam.bei.t-online.de/ There's also some other stuff that I would like to see more evolving, like ScummVM for example. If I take a look at the site promoting open source projects like freshmeat.net or sourceforge.net, I feel like a child in the candystore not knowing what to pick first. And there's also one thing I'd like to see more happening on the DC homebrew scene and that's more sharing of the source.