Re: More error leaks

On Thu, 26 Jan 2006, Garrett Rooney wrote:

> On 1/25/06, Julian Foad <julianfoad@btopenworld.com> wrote:
>
> > 2. SVN_NO_IGNORE(svn_error_t *)
> > function_name(param one,
> > param two);
>
> I prefer this method, FWIW, since it implies that what you're not
> supposed to ignore is the svn_error_t * you get from the function, and
> more importantly because it avoids obscuring the return value.

I lean towards not marking up the function decls at all, but if
consensus goes the other way, I also prefere this one. As Garrett
mentions, it obscures what's going on the least.

--
Daniel Rall