Re: svn commit: r35900 - in trunk: . build/ac-macros subversion/libsvn_ra

Justin Erenkrantz <justin_at_erenkrantz.com> writes:

> On Wed, Feb 25, 2009 at 11:04 PM, Eric Gillespie <epg_at_pretzelnet.org> wrote:
> > Anyone get serf working with the Python bindings yet?  I've
> > posted about it before.  The onus is on serf supporters.
>
> With r36174, "it works for me"(TM). All r36174 does is work around
> the fact that the tests don't provide a full configuration which
> shouldn't matter for 'real' users of the bindings...anyway...see below
> for my interaction transcript.
>
> In the absence of specific reports, I'm not sure what else you're
> seeing. -- justin

I was quite specific in those earlier posts, if I recall
correctly. I even shared a rough draft to running the swig-py
test suite over multiple ra layers; sadly, I never got had time
to polish it up, but it's there if you want to see the failures.

The swig-py test suite is woefully incomplete, however.
gvn's own test suite highlights more serf failures (and is
already easy to run over serf), if you want to try that.

It's not really a bindings issue. It's actually "serf blows up
when called by anything but svn_client". It's just easiest to
see this with our own binding and gvn's test suites. I provided
some simple C programs demonstrating some of the problems last
time I brough this up, and Lieven (I think) fixed those issues.

I say again: the burden of proof is on serf advocates, not on me.

-- 
Eric Gillespie <*> epg_at_pretzelnet.org
------------------------------------------------------
http://subversion.tigris.org/ds/viewMessage.do?dsForumId=462&dsMessageId=1243391