Re: cygwin/mingw experiences

[Ralf Wildenhues]

* Gary V. Vaughan wrote on Fri, Feb 04, 2005 at 12:29:40PM CET:
> Ralf Wildenhues wrote:
> > - something is broken in the Makefile rules that causes regeneration
> >   of toplevel aclocal.m4 during `make check', after
> >     reconfdirs=. ./bootstrap
> >   (on HEAD/branch-2-0).
> >   I suspect this is a timestamp resolution issue.
> 
> ISTR that Windows file systems have only 2 second resolution, so it is
> very easy to get into race conditions.  This problem is getting worse
> as systems get faster:  when I last used cygwin regularly, about 5 years
> ago, the average machine running cygwin was so slow that it didn't bite
> too often.  A strategic 'sleep 4' is almost certainly the fix...

I knew this was the case for FAT.  But I thought NTFS was better?
(Did I mention this is on NTFS?  I actually did this test on a newer
machine, not my old win98.)

> unfortunately, finding the right spot is likely to be a real PITA :-(

I won't be the one searching for it.

Regards,
Ralf