Steve Langasek <[EMAIL PROTECTED]> writes:
> This looks reasonable to me, and indeed I've run into this same problem.
> Russ, you added the current dependency on the .PHONY target, can you
> think of any reason that depending on debian/stamp-patched instead would
> be wrong?
No, although I'd argue that the fundamental problem is that the whole
structure is wrong. I've been switching all of my packages over to using:
configure: configure-stamp
configure-stamp: patch
# all the configure stuff
touch configure-stamp
instead of depending on $(builddir)/Makefile or config.status or any of
the other variations on that I've seen precisely because it avoids
problems like this.
Alternately, you can change patch to $(QUILT_STAMPFN), which is probably
safer than directly using debian/stamp-patched, but it isn't really
necessary provided that the debian/rules rule that depends on patch itself
uses a stamp file. It's probably a matter of taste whether to do the
configure-stamp thing or to use a $(QUILT_STAMPFN) dependency.
--
Russ Allbery ([EMAIL PROTECTED]) <http://www.eyrie.org/~eagle/>
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
