Todd Vierling wrote:

> Why does bsd.pkg.mk warn you when a patch was unsuccessful, and then *retry*
> the patch?

  It retries the patch w/o being so picky about fuzz.  This works if all
hunks fail due to a fuzzy patch, but not if only a subset of hunks fail
...

> (filename.orig, filename, filename.rej) reflect a usable set:  the original,
> the file with successful patched hunks, and only the failed hunks?

  IMHO, this is isn't too much of a problem, because patch backs up the
first set of files, i.e., you get a filename.rej.orig that contains the
original hunks that failed ...

> I don't see any point to reapply a patch that partly failed.  After all,
> patch(1) will bomb anyway because some of the hunks may already be applied.

  This is only true for patches that partly fail.  For patches that fail
as a whole, a warning is issued, but it's possible to continue 'make'. 
That was the original intention of this behaviour, I think ...

  Cheers
      ,
   Rene