On Mon, 2010-12-06, Blair Zajac wrote: 
> On 12/6/10 7:47 AM, Julian Foad wrote:
> > On Sun, 2010-12-05, Stefan Fuhrmann wrote:
> >> I've felt kind of uneasy about that issue as well.
> >> However, it would have been difficult to implement
> >> a deprecated svn_checksum_to_cstring() in terms
> >> of svn_checksum_to_cstring2().
> >
> > I don't think Blair was concerned about how the function is implemented,
> > but about the API promises.  (He can correct me if I'm wrong.)
> 
> Nope, that's correct.
> 
> > But, as I said in my previous reply in this thread, I think it's fine to
> > just change the existing API as you have done.
> 
> Well, I don't agree, for problems that it could cause later, but don't feel 
> that 
> strongly to argue about it :)

Ok.  FWIW, I'd also be very happy with a revved API, I just don't feel
strongly enough to do it myself.

- Julian







Reply via email to