On Sep 22, 2011 8:01 AM, "Daniel Shahaf" <danielsh_at_elego.de> wrote:
> Greg Stein wrote on Thu, Sep 22, 2011 at 06:16:07 -0400:
> > On Thu, Sep 22, 2011 at 00:34, Daniel Shahaf <d.s_at_daniel.shahaf.name>
> > > gstein_at_apache.org wrote on Wed, Sep 21, 2011 at 17:02:35 -0000:
> > >> Author: gstein
> > >> Date: Wed Sep 21 17:02:34 2011
> > >> New Revision: 1173759
> > >>
> > >> URL: http://svn.apache.org/viewvc?rev=1173759&view=rev
> > >> Log:
> > >> Rename the private APIs to include a double-underscore, per
> > >> have no solid convention for typedefs (which are not exposed by the
> > >> library callpoints), so they remain unchanged.
> > >>
> > >> Suggested by: philip
> > >>
> > >> * subversion/include/private/svn_subr_private.h:
> > >> * subversion/libsvn_subr/spillbuf.c:
> > >> * subversion/libsvn_repos/reporter.c:
> > >> * subversion/tests/libsvn_subr/spillbuf-test.c:
> > >> (...): rename the functions
> > >
> > > Do you intend to nominate the spillbuf code for backport to 1.7.x?
> > > not exposed in the public API so we can do that (once we feel the code
> > > is sufficiently mature).
> > Nope. I extracted it from ra_serf in order to use it in the Ev2 shims
> > that Hyrum has been working on. One of my next steps is to rewrite the
> > ra_serf to use the new spillbuf code now.
> > But there wasn't anything that I intended for 1.7.x backport.
> Okay, thanks. The libsvn_repos/reporter.c work doesn't appear to be
> part of the Ev2 work, but I don't know if it warrants backporting by
> itself. (Haven't reviewed it yet)
C.Mike asked for a concrete example, offering up the reporter as a
candidate. It was a very good one, so I wrote the code.
But... it is a simple improvement and does not warrant a backport.
Received on 2011-09-22 19:38:18 CEST