I understand your nightmare, but all hope is not lost. For example, how about incorporating show_mouse(NULL) in the destroy_bitmap() function? If mouse_init() was never called, and instead of crashing or erroring out, just gracefully handle that exception and ignore it, backwards compatibility would be retained and the API will make more sense in that regard.
On 5/10/2016 at 8:24 AM, Thomas Fjellstrom <[email protected]> wrote: >On Tue 10 May 2016 08:08:33 AM Andrew Robinson wrote: >> PS -- The http://www.cppgameprogramming.com mouse example is a perfectly >> good example, since if I leave off destroy_bitmap() as they did in their >> example, no crash will occur. >> I've looked at many examples of mouse programming by Allegro users, and I >> have yet to see any that use show_mouse(NULL) before using >> destroy_bitmap(). This is the community's way of inadvertently giving >> feedback to the Allegro developers that the usage of show_mouse(NULL) >> before destroy_bitmap() doesn't make sense. No other toolkit does this type >> of thing, so the Allegro developers should eliminate it. So my question is, >> are the Allegro developers listening to their users? >They do, but it wasn't actually brought up as an actual issue in the past when >Allegro 4 was still in development at a stage where incompatible changes could >be made. >Allegro 4 and some Allegro 5 releases made some pretty crazy forwards and >backwards compatibility guarantees. We can't just go in and change apis willy- >nilly, lest it break existing applications. _______________________________________________ Allegro-developers mailing list [email protected] https://mail.gna.org/listinfo/allegro-developers
