Discferret broken?

Exactly the same way that a user looking at a 36 pin contronics connector on a CP/M era machine would say "Thats a printer port". Or a 25-pin female 'D' connector on a DOS era machine would say "thats an LPT" port, or a 25-pin male 'D' connector and say "Thats a serial port". Or a 15-pin 'D' and say "thats a joystick port". Or a 5-pin DIN connector and say "thats a keyboard port" ... All of these are gone from modern PCs. Lots of "younger generation" people already don't know what a LPT port or a serial port is. They might know that some of their friends "really old" systems use a "round" keyboard or mouse connector - but they are speaking of the PS/2 connector (which is also disappearing). USB will fade - it will fade faster than parallel, serial, keyboard etc. because it's hard to talk to - as soon as something bigger-better-faster becomes mainstream, it will drop from "current" systems, and the support chips will go end-of-life shortly afterward. 5 years later, few people will know what it was, and 10+ years later it will be completely unknown (except to people like us :-). It's astounding how quickly "PC" technology is obsoleted (it's already getting hard to find IDE drives, PS/2 keyboards and mice - etc. and I still depend on systems that use these components). I wouldn't use USB for anything that I wanted to be usable 20 years in the future - I wouldn't even count on it for 10. For high-speed communications I use ethernet - much easier to talk to and lots of support. I think it has a better long-term viability. Still - it will also fade eventually (I don't think nearly as fast as USB) - If I really need to know I will be able to talk to it in the "far future" - I use serial. Hell, I can build level translators with transistors and bit-bash the serial data if I have to. Absolutely! Yeah, but it's a matter of cost .vs. return. Sure, if if some critical bit of historical information is stored on an ancient USB device, chances are some organization would fund the research and development needed to communicate with it. But that doesn't mean that anyone who might want to save a bit of data from an old non-functional system he found in a garage will find it worthwhile. The development to talk to USB devices from "bare metal" is substantial. Why is RS-232 "unusable"? It may not be fast, but it beats the hell out of not getting the data. I've imaged many disks from native systems over a serial port. Sometimes a slow serial port - Might take a few minites, but the images were recovered. And that is often older systems that often can't go above 9600 (or if you are lucky 19200). 115k is pretty standard these days -- about 11k/second. To transfer a meg would take about 1-1/2 minites. Even if you double that for overhead - I'd wait 3 mins to get an image from a disk I really wanted. Some people buy into the idea of support - sometimes it is real, sometimes it is an illusion. The idea of having a company "backing it up" appeals to them ... although I find I get much better support from individual authors whether or not they charge for their package. Another factor is that commercial products are motivated by the desire to make money - I agree that this does NOT mean they will have better code, but it usually means they are designed to appeal and "perform in the showroom". Dancing bears and beating drums - many people will base their decision on what it looks like in a first impression and how many "features" it has, not how well it was designed or will ultimately work. I've been watching this thread, and I'm surprised nobody has mentioned another very good reason for a casual developer to charge a modest fee for his product/service instead of giving it away - even it he is not out to make money.... "Nothing dispells enthusiam like a small admission fee" Sometimes a small fee is the best way to filter out those who would happily avail themselves of your time but have no serious interest. Offer to email <insert something interesting that you have created> to anyone who asks and you will be flooded with requests. Offer to email to anyone who sends you $5 and the load will be far more manageable... -- dave12 (at) Dave Dunfield dunfield Firmware development services & tools: www.dunfield.com (dot) com Classic computers: http://www.classiccmp.org/dunfield/