In message <[EMAIL PROTECTED]>, Andrew Lunn writes: >You have old sources. That was moved into the POSIX package in >October.
Yes. I am running the version distributed with the TAMS MOAB. I can move to a new version, but doing so makes the article harder to explain. In this case, I think, it's worth it though. >gettimeofday is a funny function anyway. It is only recently added to >the POSIX standard, which is why it was not part of the POSIX package >in the first place. The POSIX package was based on an older version of >the standard which did not have gettimeofday. Ahh, that makes sense. >Once you do have gettimeofday() and if you have a wallclock driver, >the time returned by time and gettimeoday should be "correct", >assuming your battery is not flat etc. You might have to take care of >timezones though. As it happens, I don't even care whether it's correct, because the only thing it's used for is checking whether input should have timed out. *sigh* But, thanks to your patience and explanations, I now know what's going on well enough to have a working port of frotz to eCos that's complete enough to, for instance, let me save and restore games to flash. Took some figuring, although the most serious bug I had (restore never changed anything) was in itself a beautiful bug; the save format records only changes to the original story data, and I had casually assumed I didn't need two copies of it, so I was very carefully recording a record of all changes between a chunk of memory and the exact same chunk of memory. Heh. -s -- Before posting, please read the FAQ: http://ecos.sourceware.org/fom/ecos and search the list archive: http://ecos.sourceware.org/ml/ecos-discuss
