Apple or Mac? Both. Actually Apple II, IIgs and Mac. Nearly all the IIgs and Mac titles can be fast copied or at most have one protected track (which I'll skip). I use Apple's built-in disk imager for the Mac and Asimov (http://www.ninjaforce.com/html/products.html) for the IIgs. I'll just sector-edit out the protection call on the image. Due to time constraints I'll skip format-protected titles for either (there aren't many I've seen). For the Apple II at least 50% aren't format-protected, so same strategy (Asimov images 5.25" or 3.5"). I'll use Super IOB (ftp.apple.asimov.net) to normalize everything else, many times w/the swap controller (capture the program's DOS and use it to read the protected disks). Dos 3.3-based stuff is far more reliable to get a solid image out of than 3.2. Might be better error checking, might also be that 3.2 is so old the media is WAY past its shelf life. For those "oh god I will not be complete without this" titles, I will dissect (boot trace, disassemble, etc.) the loader for however long as need be (and no I don't have a cookbook from Computist for these) until I have decoded data in normalized form. These would be stuff like "Caverns of Callisto", "Sherwood Forest", "Elite"... All of these have stories that would go for pages :) And last there are titles that are so heavily protected I just don't try (almost anything Broderbund and definitely anything Sirius). They have been beaten 15+ years ago by The Disk Jockey and the Apple Bandit, and because the formats are so far out there, I see no value in repeating these feats to only get the identical results (yes real men rewrite loaders, if you have 2 months to work 1 title). Oh yeah, SST will make nibble copy IMAGES of many protected titles (as long as there is no nibble count). I've used this as a last resort, but because its so hard to tell if a sector was misread using this tool, I try to avoid it where possible. Stuff I have in this form includes: Choplifter, Russki Duck, Wings of Fury and the original CP Ultima I. I wish I knew more, I wish I had more time, but I do what I can with what I have (and I would give my right eye to have the read routines from Copy II+ as a callable library... those things are just so good). And Copy II+ 9.1 to do the 'refresh' copy (make a copy first, THEN normalize from a fresh copy greatly enhances reliability).
Pushing hardware beyond its limits is so cool: Hit the nail on the head. I pity the people who had to PORT titles like that :) Unlikely we'll ever see that pushing again for that reason (portability, diverse hardware platforms). Wonder if we'll see 2Ghz PC emulators in cell phones someday (we never expected to see 2600, Nintendo and Intellivisions run in such small packages, who knows).
Sid tunes: I have SidPlay2 on my Mac and the HVSC library in it. I've listened to it for hours and burned a non-trivial number of CD's from it. Of course I also bought the Mac version of all the Doom II episodes to get all the music files (they're in QuickTime MIDI format... iTunes burns them straight to CD :)
I wish I could make a site w/the 'minimal crack' images I've made (along w/the original bytes if you have a better way) and not get sued from here to hell. I just *know* I've got images of stuff no one else has, and clean copies of stuff (like Rescue Raiders) that's been covered in 'graffiti' everywhere else, mods to images to make them work w/the real hardware AND emulators, etc. I think in 2004 I'm going to start scanning in all my issues of Computist (much like the Andover project) into PDF files. Tedious but its at least a known quantity of work (probably 1 hour per issue to assemble). I thought I had them all. I'm shy about 9 (mostly the really, really early ones.) Oops.
I have (no kidding) at least 10 copies of 'Omega', probably more. I still can't bring myself to destroy just *one* copy of the massive manual to scan it :)
On Dec 29, 2003, at 5:57 PM, Jim Leonard wrote:
Dan Chisarick wrote:
When I was done w/this batch (61 disk images) and waiting for the lot to transfer over the serial AppleTalk network (think 10.5MB over a 115KB/s
These are Apple II images or Mac images? Protected or not? If protected, what program are you using?
In addition to all the wonderful things that vintage games had going for them, many of them had that "wow, how'd they do that" sort of feeling. Even "Mode 0x13" for the PC (320x200 256 colors, and I think it was a linear frame buffer too instead of interlaced), even though it was more or less just poorly documented, was a "big thing".
MCGA didn't need to be documented, it was brainless to program for. 64000 bytes of video RAM at 0xA000 that you could linearly address; no paging or funny business, and each byte was exactly one pixel. Talk about simple! The neato modes were the tweaked ones, like unchaining video memory on a VGA card so that you could not only tweak to get 320x240 (square pixel aspect ratio) but also three video pages. John R. can attest to the advantages of triple buffering :) More on this later.
Now some of the advances you've seen in years past: deformable surfaces, colored lighting, inverse kinematics, more colors, more polygons, better frame rates, full-screen anti-aliasing, hi-res textures, etc. while visually impressive, don't seem to have the same impact.
Agreed. This is because they aren't exceeding the design parameters of the hardware. The truly cool stuff of yesteryear was produced by clever programming and/or hardware exploitation that accomplished things the hardware was *never intended to do*. All the stuff you mentioned above -- full-screen AA, compressed textures, pixel shaders, etc. -- are *supposed* to be used; they are in the design specs of the hardware and software. But when you produce digitized sound out of a computer with NO SOUND HARDWARE, or produce realtively fluid 3D graphics on a machine with no mathco or dedicated graphics displays running at sub-MHz speeds, or coax six voices out of a four-voice sound chip... that's impressive!
I am with you 110% on this. It is the same reason I got involved with the demoscene, which, for a very brief period around 1990-1995, almost always exceeded game companies in the ability of pushing hardware past its specified limits. Not to toot my own horn, but if you haven't heard of the demoscene, you should buy a copy of MindCandy (www.mindcandydvd.com) and watch the "oldskool" side and the featurette on the "new" side for 2+ hours of this kind of nostalgia :)
So two questions: Can you think of any technical innovation in games in the past few years that really jumped out and made you say "gee whiz"?
Yes, actually I can (with links to screenshots):
Outcast: The entire game was software-rendered in an era where 3D hardware was taking over the industry, but it was done so entirely with voxels. It was designed for 500MHz machines, but if you have a 2GHz or faster machine, you can run it with every single option turned on at a high resolution and it looks stunning. Smoothly-sloped mountains, uneven ground, true reflections... very impressive.
( http://www.mobygames.com/game/shots/p,3/gameId,358/ )
Tron 2.0: The "glow" effect (every lightsource has a hazy glow) was perfectly recreated from the movie. It is not a hard effect to do (depending on implementation it doesn't even need pixel shaders) but it was extremely effective. I love that game.
( http://www.mobygames.com/game/shots/p,3/gameId,10153/ )
Any classic favorites whose technology at the time was just awe-inspiring?
Well, unfortunately for anyone who doesn't want to read this, you have touched on the very nerve that has drawn me to computers since I was a boy. I am not saying that to be melodramatic -- it's the honest truth. So here's some stuff off the top of my head that truly impressed me by either being clever with existing hardware or exceeding the hardware's design limits:
C64 SID tunes: There are a few tunes that exploit very distinct behaviors of the C64 SID that produce sound that theoretically should not be possible given the waveforms and operators you had available. It is said that the designer of the SID is continually amazed at what has been produced with the chip, stuff that he never intended -- sounds that should not even be possible.
C64 SID music with digitized sound: A good (but not the best) example of this is the title theme to Turbo Outrun. Getting digitized sound out of a C64 is amazing enough, but at the same time all three channels are maxed out playing music?!? One of these days I'll figure out how it was done...
Digitized sound on any machine without dedicated sound hardware: Apple II, IBM PC, C64, and many (if not most) 8-bit computers did not have dedicated hardware to produce digitized sound (music and/or sound effects, yes, but not dedicated sampled sound like a Sound Blaster). So any title that *did* do it automatically impressed me.
MOD players on the PC. Armed with a Sound Blaster, the PC only had one channel of digitized sound output. So how do you output four channels at the same time, as needed by Amiga music MODule files? You build and mix four channels into one via the CPU, in real time, and output through the Sound Blaster. On a 386 this was a neat trick; on a 286 it was impressive; on a 4.77MHz 8088 it was downright mind-blowing (I know of exactly *one* program that can do this). For extra credit, output through the IBM PC speaker, sucking up even MORE cpu time! (more on PC speaker later)
MOD players for the Apple IIGS: The Apple IIGS had an Ensoniq synth capable of 15 voices of stereo digitized music. Unfortunately, it was limited by Apple's architecture, and could only access a 64K block of samples. This severely limited the variety of sounds you could play during a program; 64K was about enough to hold maybe 2 or 4 instruments or sound effects. However, there exist today some MOD players for the Apple IIGS that can indeed load more than 64K of samples into RAM and have the ensoniq chip play the tune, and it sounds just like any other decent MODplayer. I believe it does this through a fast bank-switching hack -- John R., do you know more about this? (Real-time mixing was not possible on the IIGS because it just wasn't fast enough for decent quality results.)
Digitized sound on the IBM PC that did *not* halt the computer and had *no* distortion: Digitized sound on the IBM PC was accomplished by pulsing the speaker on and off faster than it could physically respond; the result was anywhere between 6-8 bits of resolution (for comparison, the original Sound Blaster was 8-bit). Needless to say this was very timing-intensive, and typical methods either halted interrupts to play the sound (entire computer froze while sound was playing) or did NOT halt interrupts but the sound was scratchy and halting... Except Realsound. Realsound was a patented process created by Access that somehow managed to play digitized sound through the IBM PC speaker that was not only loud without distortion but also played "in the background" without halting or skipping. So you had real sampled sound, through your IBM PC speaker, that didn't halt the game dead every time a sound was played. Their early titles Echelon, World Class Leader Board, and Mean Streets showed this off and I bought every title they ever made that had Realsound on the box. One of these days I'll disassemble the code and see how they did it. I interviewed the patent holder in 1997 but haven't written up the interview yet.
Atari 2600 hacks (Multiple sprites without flickering / 3-voice sound / "textured" or "paralax" scrolling backgrounds): With only 76 cycles per scanline, all Atari 2600 hacks were cool. The 3-voice sound in Pitfall II was expecially cool because the Atari 2600 simply could not do it, so the Pitfall II cartridge has a custom sound chip on it!
Unchained video modes on VGA: Around one year after its release, Michael Abrash, RIX (remember ColoRIX?), and others discovered you could "unchain" the [EMAIL PROTECTED] mode of the VGA so that you could make it do all sorts of awesome stuff. I say "discovered", but that wasn't technically true -- it was written, more or less, in the programming manual for the card -- but what they *did* with that knowledge bordered on cool hacks and discovery: Different resolutions ranging from 320x240 (square pixel aspect ratio) to 360x480 (almost three times the resolution); having multiple video pages for artifact-free animation at the full refresh rate of the card; using spare sections of video RAM to store sprites that could be blitted much quicker than through the ISA bus from system RAM; filling polygons at almost four times normal speed by enabling a mode that could write four pixels with one write instruction... I could go on and on. I myself wrote a series of articles on how to display more than 256 colors using a stock VGA card without any 15-/16-/24-bit VESA modes. And all this was possible on a stock VGA card.
Rescue on Fractalus and Captain Blood: Fractal landscapes on underpowered hardware. Enough said.
Realtime Phong-shaded torus on a C64: This is slightly cheating because it was created in the late 1990s, but impressive all the same. Same technique as conventional tricks (it's not really Phong shading, but an environment-mapped highlight that is alpha-blended onto a textured surface, for lack of a better explanation), and running at about one frame a second, for a C64 it is very impressive.
Scaled sprites on a Genesis/Mega-drive: The Sega Genesis was a Motorola 68000 running at 7.16 MHz. Some enterprising programmer realized that this was just enough power to show a couple of scaled sprites in realtime, and used it in an early motorcycle racing game for the Genesis (name escapes me at the moment, it's not Road Rash though).
Six/Eight voices out of a four-voice chip: Ed Bogas' famous Studio Session and Jam Session programs coaxed six channels out of the original Macintosh's 4-channel sound chip (later, Super Studio Session did eight). This was a bit of software trickery, as the extra channels were mixed via software in realtime with the existing hardware channels. But it was still very cool. You can download both programs for free here: http://www.madcapps.com/
Extremely optimized IBM PC CGA graphics engines: There are a few games I can think of which impressed the hell out of me when it came to the speed and fluidity of their graphics on an 8088 IBM PC clone with CGA. CGA was not pretty to program for: It packed four pixels into a single byte, had no sprite or blit routines, and the memory was arranged in an interlaced fashion. Most significantly hurting all CGA graphics was the lack of video pages -- you had to either draw right on the screen and hope it didn't flicker or look bad, or you had to build your new video page in system RAM and then copy it to video RAM (which was very slow). And yet, some games had fantastic bitmap routines that ran as fast as consoles or 8-bit computers of the same era. Karnov had a fast tiled-background routine so that the game played nearly as fast as it did at the arcades; Outrun/The Cycles/Grand Prix/Test Drive (all from same development house) had a very fast pseudo-textured asphalt racetrack drawing routine; Robotron displayed way more sprites onscreen than I thought possible without slowdown. Essentially, raw CPU was used to emulate tiles, sprites, scrolling video pages, etc. through clever programming.
Extremely optimized IBM PC 3D engines: 3D on any platform was difficult, and the IBM PC's 8088 running at 4.77Mhz was no exception. Yet there are some games with astonishingly fast/fluid 3D graphics: Starglider and Elite had fast hidden-surface-removal 3D wireframe graphics; Powerdrome had full-screen full-resolution (ie. no "cheating" with "thick" lines) filled 3D graphics that ran at a decent clip at 8MHz; Stunt Track Racer displayed incredibly fast 3D *filled* polygon racetracks even at 4.77MHz; 4-D Boxing had motion-captured boxers in 3D (which played on an 8088 if you switched to wireframe mode) at least five years before motion capture and 3D actors were commonplace in games.
Amiga self-booting demos: Some of these "trackmos" start playing music and displaying graphics within SECONDS of booting. Most notable is "State of the Art" by Spaceballs -- you have video and music within 2 seconds after drive starts booting off of the diskette.
Using co-processors for things they were not intended to do: Anything that used hardware for things it was NOT intended for was equally cool. Things like copying RAM around on a Pentium (not sure about other chips) was actually FASTER using the MATH COPROCESSOR (hint: qwords) than using the regular 32-bit MOV opcode! There is also a demo (Facts of Life by Witan) that uses VIDEO RAM to store PROGRAM DATA to get around the 640K system RAM DOS limit! I would love to hear more examples of hardware misuse for other platforms, if anyone has stories :)
Sorry for the rambling, but technical innovation/cleverness/hackery is one of the driving factors of my hobby :)
--
Jim Leonard ([EMAIL PROTECTED])
World's largest electronic gaming project: http://www.MobyGames.com/
A delicious slice of the demoscene: http://www.MindCandyDVD.com/
Various oldskool PC rants and ramblings: http://www.oldskool.org/
----------------------------------------------------------------------
This message was sent to you because you are currently subscribed to
the swcollect mailing list. To unsubscribe, send mail to [EMAIL PROTECTED] with a subject of 'unsubscribe swcollect'
Archives are available at: http://www.mail-archive.com/[EMAIL PROTECTED]/
----------------------------------------------------------------------
This message was sent to you because you are currently subscribed to
the swcollect mailing list. To unsubscribe, send mail to [EMAIL PROTECTED] with a subject of 'unsubscribe swcollect'
Archives are available at: http://www.mail-archive.com/[EMAIL PROTECTED]/
