enag, programming harness?

[FallNAngel]

We're specifically talking about the E/Xmag Atmel and it's specific set of instructions. This is done in assembly and is simply done far more efficient when 'written' directly than using a compiler.This is the primary benefit of writing to that language than some of the more 'advanced' others.If you want to smallest most efficient code possible,assembly is one of the best choices and is used in this case precisely for that purpose.

I agree that coding in ASM leads to small, efficient code. However, you can't just take that code and burn it to the chip. You *will* have to compile it, which *will* require a compiler. Any program, in any compiled language you write (ASM, C, C++, etc) *will* require you to compile it, no matter what platform it's for.

The program is not written into the EEPROM. It is loaded into the program flash on the AT90S2313.

You are correct, I misspoke when I said EEPROM.

I got my Xmag used... No idea what the original software was. It has 3.2 now. I do not believe the memory protection bits are set. Anyway, they just prevent writes as I understand the datasheet. The fuse bits can block data downloading but they can be changed without modifying the EEPROM or the program memory.

Ahh, that would explain things. Honestly, I'm not very familiar with AVR's, as I've only really messed with programming PIC16F84's in my spare time. With PIC chips, you can lock the chip from being read, any attemps to read from it will just return 0's. I find it a bit odd that AVR's would allow you to do what you say, but I suppose that's how they were designed.

[athomas]

If the lock bits are programmed as they are supposed to be then you cannot read the flash memory or the EEprom memory of an AVR. If the lock bits weren't programmed, then it is quite easy to read the program. I suspect that there may be a few emags/xmags kicking around that didn't have the fuse bits properly programmed. I don't think its a big issue personally.