gavriluk.txt ------------ ##INTERVIEW - Eric Gavriluk ##----------------------------------------------------------------------- Erik Gavriluk, currently aged 28 and living in Hollywood Hills, CA, co- authored with Greg Miller (see interview) the fabulous Color Max and Color Max Deluxe for the CoCo3. Erik's attention to Graphical User Interface detail contributed greatly to the high level of performance and profesionalism found in these programs. Web Page - www.bombfactory.com E-Mail - erik@bombfactory.com CoCo Programming Achievements Product Title Platform Year Product Description Rainboard CoCo1/2 1983/4 BBS program. (with Lane Lester, Phd) McPaint CoCo1/2 1985 Graphics program. (with Greg Miller) McCheckers CoCo1/2 1986 2 Play modem checkers.(with Greg Miller) ColorMax 3 CoCo3 1987 Graphics program. (with Greg Miller) ColorMax Deluxe CoCo3 1988 Graphics program. (with Greg Miller) #$How and when did you become interested in computers? At a friend's pool party in 1977, I came inside to discover a Commodore Business Machines PET computer. Sitting next to it was a box full of cassettes with games on them. Trying to destroy an ASCII alien, I accidentally hit the RUN/STOP (break) key and haven't had a tan since. #$What made you choose the Color Computer? Within walking distance of my home was a strip mall with a 7-11, a pizza joint, and Radio Shack. (I was ten years old, so location was everything!) The staff was kind enough to let me remove the "Dino Wars" cartridge and tinker with BASIC. I basically spent the entire summer there and a shiny grey CoCo appeared under the tree that Christmas. Apparently the massive Tandy Marketing Machine wasn't selling many CoCo's in rural Virginia, because my machine had a very low serial number and the earliest revision of the circuit board. This oddball PCB rev caused endless problems with upgrades, though I managed to keep it alive until the CoCo2's became available. #$What computers have you owned and currently own? CoCo1, CoCo2, CoCo3, Tandy Model 100 and Amiga 500. Currently I'm on a Dell PC, with a blue plastic Mac G3 to my left. #$What companies did you work for? * 1984-1988 Milliluk Partnership * 1989 - Brief attendance at School of Visual Arts, NYC * Microsoft 1989-1995 * DreamWorks SKG 1995-1996 * 1997-present Founder of Bomb Factory Digital, Inc. At Microsoft, I led teams in the systems division delivering major pieces of Windows NT and Windows 95. #$Umm. Your credibility has been shot out of the water. :) Note that I didn't specify exactly which pieces! There's some good ones in there. #$What are some of your favourite CoCo products of all time? An unsung hero is Telewriter-64. Years ahead of its time, even surpassing PC products of the era. Nothing matched it feature-for-feature until the Macintosh came out. My favorite CoCo game, from both a graphics and gameplay perspective, is Time Bandit. But the breakthrough CoCo product was the original CoCo Max. Tim Jenison single-handedly shattered any notion of what people should expect from a user interface, feature set, and performance on the CoCo. I fired it up and gasped at its beauty; I spent hours in the disassembler learning code tricks. (Tim, of course, went off to shatter many other preconceived notions at NewTek, especially with the Video Toaster.) To me, CoCo Max belongs in the time capsule to show what the CoCo could do and, if there's room enough for some CoCo3 stuff in there, I think Color Max Deluxe sets the high water mark. But more than the products, I fondly remember the people and the era. Greg and I were both young--and I was even younger than Greg. People throughout the CoCo Community were incredibly cordial and supportive, even when, uh, "youthful enthusiasm" got the better of me. As an example, the first copy of the McPaint program featured a ten page manual with a twenty page "glossary-of-terms annex" which existed solely to make fun of competing products and their use of technology. If I did that today competitors would forge press strategies, slap me around in periodicals, and sue me. But back then, I received phone calls from people who were more amused than offended. As a result, I got to meet Marty Goodman, Steve Bjork, Dave Dies (Diecom), Art Flexser, Eric White (Whitesmith/Graphicom 2) and countless other great people. About a year after Color Max shipped I met up with Tim Jenison at a trade show party, shared some CoCo history, then danced with Kiki Stockhammer - all on my 18th birthday! (With the possible exception of a boat ride involving a tall sixteen year old girl sometime in 1985, I still consider this the high point of my adolescence.) #$Tell us about "McPaint", what it does, how was it marketed? McPaint's main goal was to provide control over color artifacts within a GUI paint program. As every CoCo programmer knows, individual pixels set in "monochrome" 256x192 graphics mode glow red or blue. Two adjoining horizontal pixels glow off-white. So we built a paint program with algorithms designed to work in the effective 128x192 mode, giving people control over red, blue, black and white -- and, for flood fills, a wide range of dithered colors. Released as shareware under the "Milliluk Partnership / Starving Programmers" corporate umbrella, McPaint topped the download charts at CompuServe long enough to get us free sigop accounts there. Version 2.0, with development funded mostly through abusing our complimentary CompuServe sigop access, had an improved GUI, better support for the Tandy X-Pad graphics tablet, and a pleasant synthesized chime stolen from a Steve Bjork game. (It also featured proportional fonts, even more meticulously hand coded printer drivers, and, at personal request of Wayne Day, a more appropriate glossary in the manual.) Credit for the McPaint 2.0 look goes to Whitesmith Partnership principal Eric White. A very talented graphic artist, Eric designed Graphicom Part II, and, despite being mocked in the McPaint version 1.0 glossary and by our Milliluk Partnership moniker, was unbelievably, unexplainably kind to us. #$What are your feelings of CoCoMax3? It came after ColorMax and soon #$overshadowed it. Tandy announced the CoCo3 on July 30th, 1986, but Greg and I didn't receive advance machines. We got ours at Radio Shack sometime in October, the first day they arrived in the stores. That night, Greg had the GIME chip doing perverse things and I was busy seeing how fast I could make zoomed 2x ellipses fill in with different colors and patterns. On April 10th, 1987, we debuted Color Max 3 at the Chicago Rainbowfest. Color Max 3 was a thorough, complete, and surprisingly bug-free product. It beat CoCo Max 3 to market by six months! CoCo Max 3 and our followup, Color Max Deluxe, shipped simultaneously at the Princeton, NJ RainbowFest that October. I'm immensely proud of Color Max Deluxe and only recently built something that I feel exceeds its quality. Visit bombfactory.com for a peek--and we'll see how it's remembered in 13 years! #$Tell us about "Bomb Factory". Are you the owner? What is your role there? Bomb Factory is a music recording studio housing an immense collection of "old things that make noise" from 1888 to present. I'm the owner and studio manager, and I also use it as an R&D lab to develop music software that emulates and authentically recreates the sounds of vintage music equipment. The corporate line: "Housed in a 12,500 square foot music recording studio in Los Angeles, Bomb Factory develops audio and video processing software for Windows and Macintosh when the lead programmer isn't busy producing and engineering music recordings." #$How did you meet up with Greg Miller? In 1984, I stumbled upon a broken 2400 baud modem at a surplus shop, and, through luck, managed to get the power supply working. Excited, I fired up Teleterm, but kept getting garbage characters back. The modem worked fine on my trusty Model 100, yet Teleterm wouldn't work. So I hunted down the irresponsible maniac who foisted this untested piece of software garbage on the poor, unsuspecting market, and demanded that he fix the bug. When he refused (he didn't own a 2400 baud modem, you see, and had merely divided down his jury-rigged time constants by a factor of two), I demanded that he give me his code so I could fix it for him. A year later, Greg and I were still sending files back and forth to each other at 1200 baud. #$What was working with Greg like? Greg is both brilliant and stubborn, and he is always right. Really. I had a blast working with him, and I learned way more from him than I could ever offer him back. He had as deep an understanding of the CoCo as anyone I've ever met; I could catch up fast, but the initial low-level insights always came from him. #$How was he "stubborn"? Greg was incredibly protective of his interrupt driven I/O routines. Even after we'd worked together for months, he refused to let me see (or fix!) the code. Finally throwing up his arms in dismay, he gave me a BASIC program that let me tweak his time constants. Armed with this hint to the workings of the code, I immediately started disassembling it, breaking a prior vow. I sent him code fragments and he didn't talk to me for days. Eventually he sent me a fully commented copy. Brilliant stuff! Everything optimized perfectly like a little puzzle, tick counts in the comment field, code the way code is supposed to be. On the CoCo, you didn't have a UART for serial access, you had the "bit banger" port. And, of course, no programmable timer. Just the "fast" interrupt (horizontal retrace) and "slow" (vertical). At 1200 baud, the fast interrupt would fire every six and a fraction multiples of the best case of Greg's interrupt routine. At 2400 baud, it was some horrible, unmanageable fraction of three. So he'd try and grab the tail of the start bit. This left him lined up for the perfect start of the first bit. Then he'd take advantage of this time to preshift some stuff, ending just in time to grab the tail of the second bit. He'd grab the third bit right away, with time left over to save the register needed to make room to lookup the shift table for the next bit. And so on. Unless incoming data forced a different set of jigsaw interactions. Madness! To his credit, he got it working at 2400 baud -- but an out-of-spec modem or dying CoCo here and there still caused problems. Software that pushes the limits of the crystal clocking the CPU-- now that's real code! #$When I interviewed Greg, he said you were more the "User Interface" #$expert between the two of you. I was attracted to graphics because of the visual gratification; once I started doing that, it was hard to imagine spending days visualizing interrupt timing errors. I preferred to make stuff dance around on screen; Greg always seemed happiest making the interrupts line up. Our strengths complimented perfectly. I'd write a cursoring/sprite engine to handle flicker-free mouse cursor and menu feedback; meanwhile Greg would toil for hours getting the damn Hi-Res joystick interface not to jitter. Of course, the end result wouldn't be right for the user unless both pieces were perfect. We were able to achieve some very nice polish by splitting the tasks between us, and giving each other plenty of support (and criticism!) #$Greg also mentioned that you hardly ever physically met each other. How #$did collaborating on a project such Color Max go when you are both #$separated this way? Was it difficult? Surprisingly easy. The constraints--slow modems, glacial compiles, not sharing the same hallway Silicon-valley style to encourage those important 30-second meetings--helped us focus on organization and quality. We kept each other in check, we factored problems into solvable pieces with each person doing what he was best qualified to do, and we stayed focused. Working in large teams, it seems like you spend more time writing "scaffolding" (throwaway code) than you spend managing dependencies. Instead, we spent our time coding, and we really honed our skills. One of my best memories is developing a printer driver for an oddball printer owned by Kenny, one of the partners at our distributor. He photocopied and mailed me a copy of the graphics codes from the back of the printer manual, and I wrote and sent a print driver to him via modem. It didn't work right, and he tried to explain what the incorrect output looked like. I made the changes, assembled, and we "went modem" on the same call. It printed perfectly the second time--and I believe I still have an open job offer at Motorola as a result of this exercise. #$Have you any "nightmare stories" during the development stages? Our code eventually grew to the point where the assembler wouldn't assemble it any more. About thirty minutes in, it'd die with "PHASE ERROR" -- basically symbols from pass one weren't where pass two expected them to be. Was it the assembler? The computer? The rinky dink hard disk controller, or the cheesy hard drive? Perhaps it's the disk operating system we hacked into thinking the 10MB hard disk was really 78 128K floppy drives. File corruption lurking in one of the optimized, text-transmission-only, error- checking modem transfer protocols we threw together? Or maybe it's the bulging capacitor with the brown spot that appeared after the memory upgrade? After about two weeks of suffering through assembly nightmares, Greg went off and wrote an assembler. It worked flawlessly, until phasing errors started to appear -- but only on Greg's machine. By this point we were up over 100K of code, working mostly in 8K chunks which were relocated and swapped into the main address space when needed. But the main 32K core was contiguous, and simply refused to assemble in Michigan. Greg dictated text changes over the phone; I'd test, explain to him verbally what went wrong if anything broke, and eventually send back working binaries. By the very end, my machine wouldn't build the core but Greg's started working. The shipping build was cobbled together from both our machines. #$How did you get into Microsoft? I learned the PC's unique style of doing things using ASM and Turbo Pascal. Once I felt I knew enough, I decided to join the ranks of a "real company." To me, Microsoft's Macro Assembler was better behaved than Borland's Turbo Pascal, so I fought my way into Microsoft. The morning I was supposed to fly to Seattle for interviews, the recruiter called. She said I'd be interviewing in the systems division, where they mostly coded in C. "You know C, right?" she asked. "Sure do!" I promised. Much like learning 6809 assembly after promising Greg I already knew it, I taught myself C on the plane ride. Fortunately my occasional slippage into Pascal notation on the dreaded Microsoft whiteboards didn't cost me the job offer. #$Can you tell us any interesting "stories" of your days at Microsoft? An old friend of mine--who I'd originally met at that same rural Virginia Radio Shack--also worked at Microsoft. Turned out that he'd been handed the "honor" of fixing up the legendarily-crappy Microsoft Paint. In the early 90's, the entire "Windows 93" team (not a typo) was particularly proud of their work optimizing the Windows Graphics Device Interface. I thought they were out of their minds, and some of that youthful CoCo arrogance emerged. I brought in my CoCo3, fired up Color Max Deluxe, and showed them the "Fat Bits" mode running at a smooth 60 frames per second. We went to my prototype dual-Pentium 90 running Windows, and the equivalent operation ran less than a frame per second. To rub it in, I told them the original loop ran at 220 fps, and I had to add a SYNC to the vertical screen retrace interrupt so the video monitor could keep up. In what seems like a particularly surreal, impossible moment, Bill Gates later passed by my office and noticed the CoCo. Give the man credit: he stopped in and shared a coding trick he personally coded into Microsoft BASIC. So the last time I sat in front of a working CoCo3, Bill Gates was in my office talking about 6809 assembly, and that's a perfect final memory of such a fine machine. #$Ever talk to Bill about the Tandy/Microsoft relationship? No, usually I talked to Bill about why Windows was so lame. He listened, usually agreed with my suggestions, and gave me people, budget, and support to go fix things. As a sidenote, the only time I witnessed nonsensical, monopolistic, or manipulative behavior in the computer industry was when Tandy informed me they weren't interested in Color Max unless we exclusively did an OS/9 version. #$Did Bill Gates ever get annoyed at your negative comments? Microsoft people are very confrontational with one another, all due to the tone set from the top. As long as you've thought out your arguments, and can back them up, you'll meet no resistance. You can't "politely mention what might be a technical issue..." -- it gets lost amongst all the screaming. #$Why did you leave Microsoft? When I realized my energies could better be spent elsewhere, I left. So now I scream about other, even stupider things that other companies do. #$What are your feelings of "Windows" compared to what else is available on #$the market? I find the "look and feel" inexcusably horrible--Windows is downright irritating to use. But beyond that, Windows really does support an amazing amount of software and hardware, and it's the only practical solution for the 90% of people who are trying to make money in their real jobs, not diddle with computers all day. It's cheap and it does a reasonable job at most things. By design, it doesn't excel at anything--and Microsoft applications make the same tradeoff. Seems like there's plenty of opportunity for competition to me! For example, there's got to be a better way to edit certain types of documents than the way Microsoft Word does it. Attorneys have different needs, scientists have different needs, grandmothers have different needs. Yet we still have four big, dorky word processors for Windows and they all do the same thing, poorly. Sure, the icons are different colors, but when my attorney sends me a Corel document, it still shows up with little black rectangles instead of quotation marks in Microsoft Word. When purchasers finally get fed up and stop upgrading, things will improve. When people finally realize that they don't want all-in-one super-apps, and instead, ask for small, elegant, understandable tools that do one thing well, things will improve. But sitting around and waiting for the government to fix it is pretty pathetic. I don't want a new operating system. In fact, I don't care about operating systems at all. Like most people, what I really want is programs for Windows that work properly. Microsoft stuff is the best behaved of the bunch, and it's still unacceptable. But other companies are worse! I'm using 3D rendering software that crashes constantly; the default install of Adobe Illustrator puts so many fonts in the font directory that it crashes on loading until you manually remove some; and getting a Hewlett Packard scanner working is like trying to launch the space shuttle. I just bought a Mac G3 and tried to set it up for music production. A nightmare! Adaptec SCSI cards don't work with the new PCI bus, there's no floppy drive so software with install disks or copy protection won't load, and Apple even removed the reset key from the keyboard and made it a teeny little translucent button on the tower. It's bad everywhere. When even Apple isn't thinking from the user's perspective any more, you know it's gotten out of hand. #$Did you ever meet Steve Jobs? I met Steve at NeXT. He's incredibly charismatic and a natural leader. I respect his maniacal focus on design, and the fact that he's nearly ruined several companies overdoing it. Everything you hear about his "reality distortion field" is true. That said, I was smart enough not to accept his offer to join NeXT. #$What is your opinion on Software Piracy? Re-reading my list of favorite CoCo products, I didn't purchase a single one of them. I either wrote them, or I stole them. From a Karmic perspective, I'm sure more of my software has been stolen, or gone as unregistered shareware, than I could fit on my hard drive today. But that doesn't make it right. Countless niche markets and computer operating systems have been destroyed by piracy. Piracy drastically inflates the economy of scale that leads to the dreaded "Microsoft syndrome." Today I'm working on audio and graphic processing software. It's already a small market, made even smaller by the fact that the socially accepted manner of acquiring this software is copying it from your pals. In 1985, McPaint needed a manual but my allowance didn't cover the $59.95 cost of Telewriter-64. Likewise in 1999, somewhere some guy in his bedroom needs a guitar amplifier simulator but barely has enough money for new guitar strings. By the time my first Color Max royalty check came in, you couldn't even buy Telewriter any more. And it doesn't seem like enough bedroom guitarists are ever going to make it onto MTV! Fortunately the audio/visual market is expanding wildly -- and that expansion alone is what's keeping companies in business. There's some focus on China now, but until sales growth finally plateaus in traditional markets, you won't see much serious interest in fighting piracy from the big boys. #$What is your opinion of the CoCo2 and CoCo3 hardware platform? I think all the 8 bit micros were great machines to learn programming on. Visual BASIC and Java are good learning tools, too, but where do you learn to count cycles? Running at gigahertz clock speeds, where do you get that positive feedback after unrolling loops? Nobody even knows what goes on under the system API calls any more. Things are so abstracted from the actual hardware that it's almost impossible to design efficient algorithms. Every once in a while I get to write something that functions standalone within the OS environment, and I get to write what I consider real code. The rest of the time I basically just throw pointers around and handle error codes--not exactly an inspirational environment for the coders of tomorrow. No wonder hacking government websites has such appeal! #$Interview copyright by Nickolas Marentes - July 8, 1999 #EThe End _______________________________________________________________________________ ##INTERVIEW - Greg Miller ##----------------------------------------------------------------------- Greg Miller, currently aged 31 and living in Sunnyvale, California, USA, is the creator of the popular 'Greg-E-Term' terminal program. Later, working in partnership with Erik Gavriluk, he also created the first commercial 'MacPaint' style paint program for the Color Computer 3 titled 'Color Max' which sold successfully via a company called Computize. Web Page - http://www.northernembedded.com E-Mail - gregm@northernembedded.com Programming Achievements Product Title Platform Year Product Description Teleterm CoCo1/2 1983 Int-driven I/O based terminal program Blackbeard's Island CoCo1/2 1984 Graphics adventure McPaint CoCo1/2 1985 Paint Program Greg-E-Term CoCo1/2/3 1985 New terminal program, w/more features ColorMax CoCo3 1987 Paint program. (with Erik Gavriluk) ColorMax Deluxe CoCo3 1987/8 Paint program. (with Erik Gavriluk) #$How and when did you become interested in computers? I hadn't thought about this in years. I distinctly remember walking into a Radio Shack sometime around 1978 and seeing a TRS-80 Model I running some of the ASCIIfied games that the machine had. They were simple "hit the target" games, but as a ten-year-old kid in the early days of video games, they were pretty entrancing. My father, however, was not amused and had no intention of wasting so much money on "a game". #$What made you choose the Color Computer? Honestly, I don't recall if I chose or my parents chose. I suspect they did. The truth is the affordability. My parents were completely convinced that it was the same thing as a 'video game' and in a case of pragmatism chose to go for the cheapest thing they saw. The fact that it was cheap and didn't need an external monitor had to have been the "seller" to them. In fact, I recall distinctly that the 4K CoCo I was $399. (And I also remember that BASIC reported 2304 bytes free. ) #$What computers have you owned and currently own? Oh, my. I suppose it depends on what you call a 'computer'. I have tons and tons of single-board computers around here with different embedded processors, but I suspect you're talking more along the lines of desktop machines. Well, the CoCo1, of course, which underwent endless modifications and at one point had me stripping off the 'battleship grey' paint, leaving it pretty much black. I owned a couple of CoCo3's. I think I ended up with a CoCo2 in there somewhere, as a toy. Ummm ... I owned a TRS-80 Model III after awhile. I bought several Amiga systems. I had a couple of Macs stuffed in a storage room. And, of course, and endless number of PC clones. #$What companies did you work for? Sometimes it seems simpler to name the companies that I *haven't* worked for. - The System Software group at Commodore on the Amiga computers. - The System Software group at 3DO on the 3DO console machine. - Several startups, some for good, some for ill. - Contracted for Microsoft/WebTV for awhile. - And am currently employed by Nokia, the cell-phone people. On the side, I do contracting work and development on several outside projects, including some JTAG-based tools for embedded systems. #$Can you tell us a "story" from your time working with startup companies? Well, I'll give you a good story, without identifying the company or the people. This particular startup was an odd company. It was founded by a nice man who'd made a fair chunk of change in the early 1980s by working "out of his garage" (the early 80's were a time when you could do exactly that). He's a technical guy at heart, but after several different companies, decided that he wanted to play a more direct role in the guidance of a company and founded a Silicon Valley startup with an ex-employee of his from a previous company (also a really nice guy). I was hired as the 'fourth employee', but quickly became the 'third' when the guy they'd hired just before me was asked to leave. (The founders were #1 and #2, naturally.) The notion of the company was good, and I was bored at my current employer, so after some negotiating, I said "okay" to the offer to join them. The experience was . . . interesting. The CEO was an engineer at heart, and seemed pained that he didn't have the hands-on feeling as a CEO that you get as an engineer. So much so that about every two to three weeks, he'd assemble the entire company and announce that he felt the need to take a firm hand in the company's direction and would suddenly make drastic changes in the product that the company was pursuing. The strangest event, however, happened in my fourth month there. I had two people reporting to me, and I was responsible for 1/2 of the company's product. The CEO and I would walk around the industrial park once every couple of weeks and I'd give him an update on our progress, our hurdles, and our problems. He'd make suggestions or requests of me then. One day in particular, he said: CEO: 'Greg, would you do me a favor?' Greg: 'What's that?' CEO: 'I'm a little worried about how focused the guys are.' Greg: 'Okay . . .' CEO: 'Could you get your guys (referring to the two people who reported to me) to rearrange their desks so that their monitors face the office doors? That way, I can walk around and keep an eye on what everyone is doing.' I then spent five minutes trying to explain to him that what he was suggesting wasn't going to fuel 'corporate loyalty'. He slowly let the idea drop, but I don't think he understood why I was hesitant to follow up on it. #$What work did you do on the Amiga? I touched a handful of things, but primarily networking. Commodore was an absolutely fascinating place to work -- just the people, the situations, and the location (Commodore was based in an odd spot for a computer company -- just outside of Philadelphia, Pennsylvania). #$What work did you do on the 3DO? I worked a bit on 'Opera', the box that shipped in the stores, a bit more on M2 (the box that MEI bought for $100M then sat on), and spent most of my time working on the 3DO/Scientific Atlanta/US West Set-Top-Terminal. We did a STT trial in Omaha, Nebraska. Suffice to say that the event was interesting, and I spent quite a bit of time on airplanes flying back and forth between Scientific Atlanta's headquarters (in Atlanta, naturally) and California. I've an odd background; people tend to like to classify themselves as 'an expert in such-and-such' (trust me, there are very few real experts in nearly anything in this industry), and I confuse the hell out of recruiters when I tell them that I'm a 'System Software Generalist'. Strangely, 3DO decided to become an all-game company a few years back, and sold it's engineering division to Samsung. Samsung spun it off and immediately was hit by the financial troubles in Southeast Asia. Interested in cash, and less than convinced that the spin-off would win the contract that it sought, Samsung sold the assets to Microsoft, who purchased them for it's WebTV division. Through a very weird twist of fate, another ex-Commodore guy and I ended up with the rights to the embedded OS that had been written at the spin-off. We're hoping to give it away in some capacity soon as an embedded OS. #$What is your role at Nokia? I'm just a general-purpose System Software guy for their IPRG division -- their 'IP Routing Group' division. Pretty much whatever needs to be done, I'm usually willing to do it. At the moment, I'm doing something rather boring (adding disk mirroring to their architecture), but fun projects come and fun projects go. #$How did you meet and team up with Erik Gavriluk? When I wrote 'Teleterm', it had the advantage of using the 63Mhz Horizontal Sync interrupt on the CoCo1 to clock out each bit via the serial port in the background, while the terminal program ran. Most other terminals at the time worked purely by closed polling loops, and had problems sending and receiving data at the same time, not to mention scrolling the screen or any other type of action while data was flowing in and out. (Heck, I even remember that 1200 baud worked out to 6.5 interrupts/bit, forcing me to do 6 interrupts, then 7, then 6, and so on to approximate "6.5".) Anyway, I would get odd telephone calls from people who'd tracked me down and who had questions or comments. Most people were quite polite. However, one day I received a telephone call from a guy who'd found my home telephone number (I don't know how) and called me up *demanding* the "interrupt driven I/O routines" from 'Teleterm'. He wasn't asking, he was demanding. I distinctly remember thinking 'This guy must be some sort of nut.' It turns out that this particular nut was Erik, and that telephone call spawned a long period of collaboration on future projects. Interestingly enough, Erik and I had worked on several projects together over the phone (exchanging code via modem) before we ever met physically. In fact, I think I've met Erik in-the-flesh no more than four times total. #$Was it difficult working on projects via modem? Looking back, considering that Erik and I had no revision control between us, I'm amazed that we managed to get a product completed. That's probably because he and I had a pretty clear delineation between 'Erik's files' and 'Greg's files'. For the most part, I didn't touch things in his realm and he didn't touch things in mine. #$Any interesting techniques that were created to achieve the "polish" in #$Color Max Deluxe? I couldn't say. Erik definitely had a better eye than I for user interface. He caught subtleties in good UI's, such as careful use of proportional fonts, shadows and what-not, where I never noticed them. As a funny side comment on Erik -- he told me long after I'd known him that he'd agreed to work with me on some assembly-language project, claiming that he knew assembly. He didn't have a clue how it worked. He figured it out as we went. I never had a clue! #$CoCoMax III seemed to overshadow ColorMax when it was released. It was overshadowed primarily because of advertising. Erik and I worked with a couple of great guys who ran a company called 'Computize'. They made it clear to us from the beginning that they didn't want to play 'advertising war', and eluded more than once to the notion that Lonnie Falk (the publisher of Rainbow) was notorious for heavily promoting products of his friends, while offering substantial advertising discounts. I can't say if any of that is true, but Bruce and Kenny (the Computize folks) felt that it would be a losing battle. My experience in the CoCo days was generally great. The only time I was ever outright stunned was when people saw 'Color Max Deluxe' and said 'They're just copying CoCo Max!!' My reaction was one of stunned amazement. We weren't actively copying CoCo Max. The reality is that CoCo Max was copying the same User Interface that we were -- 'MacPaint' on the Macintosh. I have no complaints, incidently. Color Max Deluxe did *very* well, all things considered. #$What are some of your favourite CoCo products of all time? Color Max Deluxe. However, I'm clearly biased. I think it had a great 'feel' to it, and I was very proud of our work. Erik did a tremendous job giving the 'Deluxe' version a sense of polish that our 128K version simply didn't have. #$What is your opinion on Software Piracy? What can I say? Did I pirate software as a kid? Of course. Do I now? Hey, it's kind of hard to make your living in software and still rationalize piracy. In fact, I make a particular point of mailing off shareware registration fees to people these days. However, I can afford it today far more easily than I could as a teenager. #$What was your reason for leaving the CoCo market after Color Max Deluxe? Oh, I left because it was clear that technology was moving much faster than Tandy was, and that Tandy felt that they were best off in the PC market. The new 68K machines those days (like the ST and the Amiga) looked *very* cool. Erik, incidently, advised doing PC work instead. I didn't want anything to do with the yucky Intel processors, and refused. In retrospect, he was right (at least in a financial sense). #$What is your opinion of the CoCo1/2 and CoCo3 hardware platforms? Neat little boxes. They gave exposure to computers to a group of people at a level that isn't done any more. I meet few people these days with an inherent "to the metal" background, and even fewer still that didn't get that background from the IBM PC. Don't get me wrong, I've spent quite a bit of time working in System Software, and have often been a proponent for people to "Use the OS, don't try to go to the metal", but there certainly is something to be said for understanding a machine in it's entirety. #$What did you think of Tandy's role in the marketing of the CoCo3? What can I say? Tandy saw the volume in moving PC clones. Sales and marketing people love "the sure thing", and PC clones definitely looked like that. Compared to something where other people are doing your advertising and promoting for you, why would Tandy want to put much (if any?) effort into pushing a CoCo? Hence, their eventual move to only PC clones, and eventually to selling other people's systems instead. #$What features do you feel were lacking in the CoCo3? Oh, a faster CPU. The 6809 was a great little CPU, but it was seven years old by the time the CoCo3 came out. It should've had a better CPU, and preferably some custom hardware for graphics and sound. Perhaps something between what they did and the hardware and sound abilities of the Amiga. #$What was your opinion of OS-9? A nice idea, but poorly thought-out in some fields. Besides, unless it was shipping on every box, even had it been well-thought-out, it couldn't have been a major target for applications authors. #$Did you ever get into OS-9 on the CoCo? A little, but I was frustrated by the graphics model that Microware was pushing. OS-9 was kind of like a "poor man's Unix" at the time, and it shared the Unix view of graphics -- that is, very vector oriented. The OS-9 graphics methodology granted little access to bitmaps, and when it *did*, it prevented you from double-buffering frames. That alone made it of limited value to me. #$What is your opinion of today's PC's? What kind of opinion can one have any longer? As someone who learned the hard way that 'Technically Superior' and 'Success' are rarely linked (That is, if something is successful, don't presume that it's technically superior to the competition -- and conversely, if something is technically superior to the competition, don't presume that it will be successful.), I take most of it in stride. PC's won the battle for sure ten years ago. Having a single architecture with a single OS (hey, I've contracted for them, and I use and advocate WinNT, but let's be honest, they *are* a monopoly) is great for consumers in one sense (centralized, less confusion, industry focus is on one and only one target) -- it's horrific in another sense: Significantly less competition. As I said before, I live in Silicon Valley. People here are terrified of Microsoft, and with good reason. I've seen VCs (Venture Capitalists) be awfully interested in ideas until they hear that Microsoft might actually be remotely interested in something that's kind of, but not quite, similar to the same topic. At that point, they run for the hills. You truly cannot compete with Microsoft, so nobody is willing to even discuss trying any longer. Silicon Valley companies are either trying desperately to climb into bed with Microsoft or finding ever-shrinking niche/vertical markets. Suffice to say, I'll be interested to see what happens at the end of the current US Government vs. Microsoft trial. (I suppose, after the end of years and years of appeals.) #$Interview copyright by Nickolas Marentes - July 03, 1999 #EThe end