Re: svn commit: r23740 - branches/merge-tracking/subversion/tests/libsvn_subr

On Sun, 18 Mar 2007, David Glasser wrote:

> On 3/8/07, pburba@tigris.org <pburba@tigris.org> wrote:
> > static const char *mergeinfo1 = "/trunk: 5,7-9,10,11,13,14,3\n/fred:8-10";
> > static const char *mergeinfo2 = "/trunk: 1-4,6,3\n/fred:9-12";
> >-static const char *mergeinfo3 = "/trunk: 3-7, 13\n/fred:9";
> >-static const char *mergeinfo4 = "/trunk: 5-8, 13\n/fred:9";
> > static const char *mergeinfo5 = "/trunk: 15-25, 35-45, 55-65";
> > static const char *mergeinfo6 = "/trunk: 15-25, 35-45";
> > static const char *mergeinfo7 = "/trunk: 10-30, 35-45, 55-65";
>
> This seems a little odd --- shouldn't we renumber the larger-numbered
> variables to not leave a gap?

I'm not following this comment. (The screaming baby next to me might
have something to do with it.) Would you elaborate?

> (And is there a compelling reason to have these constants separated
> out instead of inline, closer to where they're actually used?)

These are DannyB's. IIRC, he defined some merge info constants and
re-used them across the tests he wrote for simplicity.