On Sun, Feb 08, 2004 at 04:55:27PM EST, Thomas Dickey wrote: > On Sun, 8 Feb 2004, David Yitzchak Cohen wrote: > > On Sun, Feb 08, 2004 at 04:36:06PM EST, Thomas Dickey wrote: > > > On Sun, 8 Feb 2004, Thomas Glanzmann wrote: > > > > > > > The access time wouldn't be useful for cygwin. > > > > > > > > This isn't an issue because we can fallback on modification time for > > > > cygwin. > > > > > > But the point was that access time is not portable (and I seem to recall > > > noting that it is a feature of the filesystem rather than the operating > > > environment). > > > > Well, Mutt can check at runtime whether or not the FS is mounted noatime > > by simply creating a temp file, waiting a second, then reading it, > > more than a second, actually. sleep() won't necessarily wait a whole > second, and there are (no surprise) filesystems that don't record times > in seconds. Well, it can nanosleep(2) or usleep(2) (or even select(2)) a second or two, instead, but I'd hesitate to rely on atimes for a filesystem with worse precision than a couple of seconds. . . - Dave -- Uncle Cosmo, why do they call this a word processor? It's simple, Skyler. You've seen what food processors do to food, right? Please visit this link: http://rotter.net/israel
Attachment:
pgplMBlXFgnVj.pgp
Description: PGP signature