On Fri, Sep 08, 2000 at 03:17:56PM -0400, Bradley M. Kuhn wrote:
> Michael Salmon wrote:
>
> > | > It seems pertinent to ask if anyone knows whether changing the ctime
> > | > rather than the atime will make the mail agents any happier about the
> > | file > being unmodified? There's no point implementing it if it won't
> > | solve the > problem.
> > |
> > | I use mutt, and when I use my preserve-atime-patch, things work right.
> > | Without it, things don't (mutt thinks none of my new mailboxes have new
> > | mail after I rsync them).
> >
> > What happens when your system is backed up?
>
> Everything works fine; but I backup based on mtimes, and I use cpio, with
> the preserve atime option. ;)
>
> I think it is somewhat silly to worry about the clobbered inode data. If
> people are preserving atimes, they know the cost. Plus it's easy to
> document and even print a warning so people know what is lost if atime is
> preserved.
i try compiled rsync 2.4.6 with atime preserve patch
there is error:
patching file `./backup.c'
patching file `./compat.c'
patching file `./flist.c'
Hunk #2 succeeded at 230 (offset 52 lines).
Hunk #4 succeeded at 280 (offset 52 lines).
Hunk #6 succeeded at 339 (offset 52 lines).
Hunk #8 succeeded at 449 (offset 52 lines).
Hunk #9 succeeded at 512 (offset 14 lines).
Hunk #10 succeeded at 594 (offset 52 lines).
patching file `./generator.c'
Hunk #1 succeeded at 43 (offset 2 lines).
Hunk #2 succeeded at 61 with fuzz 2 (offset 9 lines).
patching file `./options.c'
Hunk #2 succeeded at 134 (offset 11 lines).
Hunk #3 succeeded at 183 with fuzz 2 (offset 6 lines).
Hunk #4 succeeded at 232 (offset 12 lines).
Hunk #5 succeeded at 459 (offset 14 lines).
Hunk #6 succeeded at 647 (offset 28 lines).
patching file `./proto.h'
Hunk #1 succeeded at 184 (offset 4 lines).
patching file `./receiver.c'
Hunk #1 succeeded at 312 (offset 1 line).
patching file `./rsync.1'
Hunk #2 succeeded at 493 (offset 12 lines).
patching file `./rsync.c'
Hunk #2 succeeded at 145 (offset -1 lines).
Hunk #3 FAILED at 165.
1 out of 4 hunks FAILED -- saving rejects to ./rsync.c.rej
!!!!!!!!!!!!!!!! ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
patching file `./rsync.h'
patching file `./sender.c'
Hunk #2 succeeded at 174 (offset -2 lines).
patching file `./util.c'
Hunk #1 succeeded at 212 (offset 58 lines).
rsync.c.rej:
***************
*** 161,180 ****
st = &st2;
}
- if (preserve_times && !S_ISLNK(st->st_mode) &&
- st->st_mtime != file->modtime) {
- /* don't complain about not setting times on directories
- because some filesystems can't do it */
- if (set_modtime(fname,file->modtime) != 0 &&
- !S_ISDIR(st->st_mode)) {
- rprintf(FERROR,"failed to set times on %s : %s\n",
- fname,strerror(errno));
- return 0;
- } else {
- updated = 1;
- }
- }
-
change_uid = am_root && preserve_uid && st->st_uid != file->uid;
change_gid = preserve_gid && file->gid != (gid_t) -1 && \
st->st_gid != file->gid;
--- 165,170 ----
st = &st2;
}
change_uid = am_root && preserve_uid && st->st_uid != file->uid;
change_gid = preserve_gid && file->gid != (gid_t) -1 && \
st->st_gid != file->gid;
i compiled this with debian potato and i have .deb package with atime preserve
... it's works fine with -s option
my question: is it good??? maybe my rsync has error elsewhere ...
is existing some updated patch???
--
5o
-------------------------------------------------------------
Ing. Peter Mann e-mail: [EMAIL PROTECTED]
UVT - TU Kosice ICQ # : 12491471 - KLFMANiK
PM2185-RIPE
