(changing my email to avoid the silly nonsense I've been having the my
@aim.com address and mailing lists, pardon for probably breaking threading)
> I'm concerned this is a api/abi breakage. It looks like you've
completely removed the old set/get functions. Are they only in 5.1 and
can actually be removed like this?
These functions are actually one of the first things added in 5.1,
they've been there for years. Other than that, yes, they were never
back-ported into the 5.0 branch.
> Would it break too much if they still existed, but behaved similarly,
and only acted on the backbuffer's local projection? Just so people who
use them don't immediately break?
I think the behavior is different enough that I think it would be worse
to have it change silently. I think having it be a compile error would
be better at alerting people something is wrong. Obviously I can only do
this because it's 5.1-only. I do want to make it easy to switch though
by making a tiny migration guide (probably more meaty than what I wrote
in the email).
I'll repost the OP on allegro.cc to gauge the users' reaction there
about this sometime later too.
> We can always tag them as deprecated (__attribute__(deprecated) ftw)
and slated for future removal.
Yeah, I was planning on using this functionality for 5.0 functions that
we one day might remove post 5.2 (hah!). I don't think it's necessary
for 5.1-only functions though.
-SL