[...] Aha... That makes a lot more sense, and I hope I can make some useful comments. I hate to say this here, but I would _not_ recomend a classic computer (any classic computer) as the only system for a non-computer-literate user. IMHO such a user will quickly become disapointed when they find that the things that _they_ can do with the machine are somewhat limited, and that nothing from the local computer store will be of any use to them. I can have advanced video on one of my older machines because I am prepared to write drivers, solder up interfaces, read technical manuals, patch existing programs, etc. OK, I enjoy it. But to suggest that a PERQ 2 driving an I2S image processor and a homebrew GPIB -SCSI interface would be a good machine for the average home user to view his photo-CDs on would be nothing short of insane. Yes, it _could_ be done, but for such a user, a normal, boring, PC-clone would be much more suitable. Even I don't use a classic machine (or at least, what _I_ consider classic) all the time. Earlier this week I needed to write some letters. I could have used a classic-PERQ and printed the results on a Versatec V80. I could have used the paper tape editor on my PDP8/e to punch a tape containing the correct postscript commands and then printed it to my laserprinter using a serial-interfaces paper tape reader. In fact I used something non-classic - a much hacked PC/AT running Linux. I used LaTeX to format the text, and then printed it on said (fairly modern) laserprinter. There are plenty of other things that would have been just as good, but that's the one I chose. The people who (IMHO) we should be trying to attract to classic computers -- and in fact the ones I've had some success in attracting -- are those who are already computer literate to some extent. People who already can write simple programs and want to understand exactly how a system operates. I think it's a lot easier to understand many classic computers (minis and micros) than modern PCs, and the educational value of such machines should not be overlooked. There are (IMHO) at least 4 different types of tasks that can be done with computers : 1) A standard application for which software (commercial, GPL'd, freeware, shareware, whatever) already exists. In which case you pick whatever machine the software runs on. If it runs on more than one system, you choose between them using whatever criteria matter _to you_ (hence me picking a classic computer for its better documentation over a modern PC since both can carry out the tasks I want to do) 2) An existing embedded system. Although it's possible (and sometimes common) to 'modernise', say, a control system based on a minicomputer, there are still a lot of PDP8's and PDP11's running machinery, etc. Provided the old machine works reliably and support/spares are available, it's fine to keep on using it. 3) A new 'embedded' system. Although my example of the I2C chip tester was, I agree, slightly contrived, I was simply pointing out that if you are making a single-purpose machine _for youself (so you can maintain it)_ there's no reason not to use a classic. There's probably no reason not to use something modern either. 4) Education. This includes 'writing programs for fun', understanding how computers (hardware and software) works, etc. This is IMHO the main use of classic computers to the average person these days, and the way we'll attract more people into the hobby. I've said before that you can learn more about processor operation by spending a couple of days with a minicomputer CPU, the schematics for it and a 'scope than by sitting through most lecture courses on the subject. And, I can fully understand the joy that comes from getting a computer to do something that is widely claimed to be 'impossible'. When the CoCo was current, it was normally claimed that you couldn't have PMODE 4 graphics and Semigraphics 4 blocks on the screen at once. I did it, and totally amazed the staff in the local Tandy store. Since it's easy to do most things on a fast PC (you don't really have to save every last cycle, etc), it's more of a challenge to do them on an old home micro. Again, we may be able to attract people because of that. It may be worth looking at how (say) classic car clubs keep their membership. They may well have similar problems - why would you drive (say) a pre-war car when you could have a modern one. Rationally it probably makes very little sense... I will have to disagree with you there. The larger machines that I am involved with have very active user comunities who offer a lot of help and support to newcomers. To give a couple of examples : PERQ : When I was trying to add a printer to my PERQ, I had a few questions. The result was that other owners dismantled their own machines, 'buzzed out' cables, e-mailed me wirelists, and talked me through the whole thing. That's not an isolated incident - a person who was having monitor problems got other owners to pull the covers off their monitors and measure voltages, etc. This is just about the only machine where this level of support still occurs over 10 years after the entire line was discontinued. PDP11. The user community here is split (IMHO) into the professional users who need to keep the machine running, and who can afford DEC diagnostics, replacement boards, etc, and the home hackers who debug everything with a logic probe and printset, rebuild modules and even repair dead fans. I am certainly in the latter category here. But again, questions get answered. People will dig out manuals for obscure hardware and look up pinouts and jumper settings. Newcomers will get answers to introductory questions. BTW, I'm not flaming the support that other user communities give. I'm sure it's excellent, but I've never experienced it, since I don't (in the main) use those machines. The UK chapter of DECUS is pretty useless for home hackers. I am told that other chapters are a lot better in this respect. -tony