Den 2007-11-24 06:59:18 skrev Bob Blick : > IMHO, Microchip is doing too many things at once and doing them badly. We once turned down Atmel AVR because at that time they had delivery problems (they seem to have focused on delivering memory chips instead) Another reason we do not even consider them now is they stop producing chips in only a few years, while we plan product life cycles in more than 10 years. Pity, as AVR and is a nice core design. Microchip still deliver the PIC14000 after more than 10 years although it is a pretty bad chip i love them for that as it is in the first product i made with a microcontroller and it is still in small production. Othe nice cores besides AVR i think are ARM, MIPS and dsPIC... Our main pain with PIC18 currently is we choosed to use C only for programming which makes it impossible to cram up interrupt response anywhere near a hand coded assembly. For main loop C is nice but for hardware near perfection C is a pain; we have not yet figured out how to insert instruction for quick response before context save, also it insists on not using MOVFF to write PortB - thus destroys interrupt on change. Now you can blame it on PIC being a bad design for C, or C being a bad choice for PIC... Next project I will write the I/O system in pure assembly! On any core for that matter... as I think i will never truly trust the libs, I want to code any pheripheral acess myself anyway, and i can do it in assembly more cleraly and deterministic than in C. > They can't even do a datasheet for the 16F series parts > that isn't preliminary(until it's obsolete). That is a pain especially combined with undocumented erratas, bus as have been discussed here also othe companies ship chips wiht half-known issues... /Morgan Olsson -- http://www.piclist.com PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist