On Wed, Jan 30, 2013 at 5:58 PM, Jordan Rose <[email protected]> wrote:
> It does seem unfortunate that this has to bump the major version number, 
> though. Someone being very safe with their code would then assume none of the 
> past APIs can be trusted.
>
> Maybe we can mark that this particular number is known to be volatile across 
> minor versions?

That's an option.

Does anybody actually use this number though? The only purpose I can
think of is to do some kind of introspection on the actual cursors,
and I think that is going to be problematic for compatibility no
matter what (and this use case assumes that the cursor enum values
have no gaps, which I'm not sure is something that we guarantee).

-- Sean Silva
_______________________________________________
cfe-commits mailing list
[email protected]
http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits

Reply via email to