On Sat, Feb 21, 2009 at 3:35 PM, C. Michael Pilato
<cmpilato_at_red-bean.com> wrote:
> Mark Phippard wrote:
>>
>> Was this Python change intentional BTW? I thought we just wanted to
>> require it for our tests. It seems crazy to me that we would be
>> requiring it for our bindings and other things. How are we supposed
>> to support RHEL 4? A lot of enterprises still use that (and RHEL 3
>> for that matter).
>
> IIRC, it's only gen-make and friends that need the new Python, and you only
> call gen-make if you're building from a checkout. The source tarballs have
> already been autogen'd. I build from a working copy on svn.collab.net that
> I simply 'svn switch' from tag to tag as we release.
That is what I am trying to get clarity on. I just re-read the old
threads when this was proposed. I thought that the intent of the
change was to just require Python 2.4 for developers ... those that
are running gen-make.py. There was some discussion about whether the
test suite could also require it. I'd be fine with that too.
It just seems that someone ought to be able to take our release
tarball and still build SVN and the Python bindings with 2.2 or 2.3 of
Python on their box. If that can still be done, then I am fine with
it. That is NOT what we are currently seeing when building our
CollabNet package, but if it is supposed to work I can ask them to
take a second look. Maybe the build is being done from our tag
instead of the tarball etc.
--
Thanks
Mark Phippard
http://markphip.blogspot.com/
------------------------------------------------------
http://subversion.tigris.org/ds/viewMessage.do?dsForumId=462&dsMessageId=1204892
Received on 2009-02-21 21:40:55 CET