Paolo Bonzini wrote: > On 07/26/2011 12:18 PM, Jim Meyering wrote: >> Regarding NTFS, can you point to a real gnulib-using application that is >> misbehaving because of this? I've seen that some NTFS implementations >> *do* have usable inode support. Both cygwin and fuse-based ones do, >> so you must mean mingw. > > Yes, I mean mingw. I was quoting MS docs. > >> > However, we could fix that in gnulib. >> >> Is it possible to do that without a run-time penalty? > > We would need to reimplement stat on MinGW. So it would completely > bypass the C library and should have comparable performance. If not, > you pay the price on MinGW only. > > I'll run some tests on a real Windows machine.
I don't know about typical mingw file system usage. Do mingw users use NTFS often? google searches for "mingw ntfs" (716k) suggest it's more often written about than "mingw ext3" (318k).
