On Wed, Jun 14, 2006 at 10:23:30PM -0400, Benji York wrote:
> Brian Sutherland wrote:
> > Log message for revision 68639:
> >   A basic makefile, the ZPL and copyright headers.
> Is there consensus about adding makefiles like this (diff below)?  I 
> personally don't like the idea, but I wanted to bring it up to the list.
> I don't like them for a couple of reasons. First, they're make files; 
> that's an automatic strike :).  Second, they're not useful 
> cross-platform; if what they're doing is important, then they should be 
> Python scripts, if not, then they shouldn't exist.  Third, /I/ don't 
> value what they're doing for me.
> Other opinions?

Firstly, being the one that added these, i'll take them out/replace them
according to the consensus.

But here's my objective:

The reason I put them there in the first place is to nudge committers to
create both the tarball and egg (for perhaps more than one python
version) when creating a new release. The makefiles are not required and
are trivially replacable by whatever type of script you want.

Why do I need a tarball and not just an egg?

Well, Debian (and I presume other) packaging systems always build from
tarballs. There are also automated tools to check for the presence of
new upstream tarballs. So if I am packaging zc.table for example and a
committer makes a new egg-only release I have to take a number of extra
    * Actually notice that there is a new release.
    * Download the tag, make a tarball, and upload it.

Where to from here?

I guess I could replace them with a make-a-release.py script which just
invokes the setup.py script for the tarball and a list of python
versions. Or we could enforce this within the setup.py (Too invasive in
my opinion).

Brian Sutherland

Metropolis - "it's the first movie with a robot. And she's a woman.
              And she's EVIL!!"
Zope3-dev mailing list
Unsub: http://mail.zope.org/mailman/options/zope3-dev/archive%40mail-archive.com

Reply via email to