Olin Lathrop wrote: > My biggest gripe is that the USB organization doesn't seem to care > about the little guys. Yes, that's probably true, although of course I don't believe that they're deliberately trying to keep the "little guys" out, as a few people have suggested. > $1,500 is nothing for a large company, but despite what Andrew > Warren said (who works for a large company that produces USB > chips), $1,500 can be a real issue for a little company. Of course, I haven't ALWAYS worked for Cypress... Before I became a sellout and started working for The Man, I owned a small company much like yours, so I think I'm pretty familiar with the financial constraints under which you work. When I say that $1500 should be affordable if someone's serious enough to be successful making USB devices, I'm not saying "Let them eat cake"; I'm helping that person decide whether what he's doing is a hobby or not. > The vendor IDs should be much cheaper, since these obviously don't > cost much more than a little bookeeping. If they want to provide > plugfests, make those pay their own way by charging what it takes > to produce them. As I said in a previous message, I believe that the ID-only fee, which used to be $200, was raised to $1500 so as to narrow the gap between it and the ID-with-testing fee of $2500. Narrowing the gap encourages people to pay the $2500 and actually test their devices before unleashing them on the world. Testing is a good thing. Charging less for Vendor IDs and making people "pay their own way" by charging more for testing would DISCOURAGE people from testing their devices. That would be a bad thing. > Or, at least have a sliding scale for vendor IDs based on company > size or number of units shipped. Device programmers are expensive. Is yours priced on a sliding scale? Ok, maybe that's an unfair analogy. How about this instead: Every once in a while, someone on the piclist asks about receiving a sliding-scale royalty for his firmware, based on the number of units his customer ships. The response is almost always "Don't do it; it's too easy for your customer to cheat you." If the USB-IF had to verify company size and sales volume, I wouldn't be surprised if the extra overhead raised even the minimum license fee above $1500. The USB-IF is a small organization; they don't already have a bunch of accountants sitting around looking for something to do. > why not have the USB organization sell individual device IDs within > a general purpose vendor ID for $50 or something, just enough to > cover the registration costs. Because then the devices built by those people wouldn't get tested and certified, and the introduction of those untested devices into the USB "ecosystem" would probably disrupt what is currently a very pleasant user experience. > Little guys could be accomodated if there was only the will to do > so, but this is not in the interest of the big guys that are the > influential "members" of the USB organization. Besides, they all > got there with the existing rules, so why would they want to let > others in more cheaply? Think. If we really wanted to keep you out, wouldn't we have conspired with the USB-IF to charge a lot more than $1500? > The other dissappointment has been the lack of a really accessable > USB microcontroller for quick one-off or low volume projects. I > looked at the Cypress line a year or two ago, and there wasn't the > combination of what I really wanted, in addition to the whole > company presenting a strong attitude against small volume designs. Well, you can't really blame us... If you're planning to sell so few USB devices that you won't be able to afford the $1500 VID fee, how much money can we expect to make from you on sales of our microcontrollers? Much as I'd like for everyone in the world to be using our parts, it just isn't cost-effective to support thousands of 10- to 100-unit/year USB customers when the same revenue can be made from one medium-sized customer. > In one case the chip looked reasonable, but it was OTP only with > no flash or UV erasable parts. The only way to do software > development was to buy a case of parts and use a new one each > retry. Yeah, I've always thought it would be nice to have reprogrammable low-speed micros in addition to our reprogrammable full- and high-speed parts. Personally, though, I think your programming skills are good enough that it wouldn't take you more than a few dozen attempts before you got your code working... But if you think it would take more than 300 tries, it would be cheaper to buy our $500 emulator. -Andy === Andrew Warren -- aiw@cypress.com === Principal Design Engineer === Cypress Semiconductor Corporation === === Opinions expressed above do not === necessarily represent those of === Cypress Semiconductor Corporation _______________________________________________ http://www.piclist.com PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist