mktemp, tempfile & coreutils

Bryan Kadzban bryan at
Wed Oct 17 15:41:02 PDT 2007

Hash: RIPEMD160

Matthew Burgess wrote:
> That offer was declined[2], as upstream don't want to be seen to
> encourage the continued use of `tempfile'.  This is eminently 
> understandable, and I have to wonder if we shouldn't follow suit.

If it's not too difficult, then sure, I'd agree.

> Does anyone know how widespread the use of `tempfile' is these days 
> (for example, do _any_ LFS or BLFS packages call out to it)?

Without doing any testing or grepping or poking around in log files or
anything like that, I have no idea.  However, I'd say that if there are
only a few packages that use it, and if mktemp is a fairly easy drop-in
replacement for tempfile, then we should create patches (or seds?).

If there are tons of packages, then I don't think patches are a good fix
(although it'd be nice to have them all using mktemp, I think the cost
is probably too high in that case).

And I've done some grepping through log files: of the packages that I
used Matthias's pkg-user scripts for (which are the only ones I have
build logs for), none use the tempfile program during any of the steps
("./configure --args", "make", "make check" (or equivalent), or "make
install").  I don't know which packages may use it at runtime.  (But
this may not be every package in LFS, and it's certainly not everything
in BLFS.  It's almost everything that I've installed, though.)

> I think we've given plenty of time for any users of the `tempfile'
> binary to have been updated now, so any remaining users should be
> patched to use `mktemp'.

That's my first thought as well, but I don't know for sure how many
users there are.  It looks like "none during installation", but who
knows if that's actually true.
Version: GnuPG v1.4.7 (GNU/Linux)
Comment: Using GnuPG with Mozilla -


More information about the lfs-dev mailing list