Re: [AD] END_OF_MAIN removal patch for msvc

["Trent Gamblin" <trent@xxxxxxxxxx>]

On Tue, November 18, 2008 4:56 pm, Stefanos A. said:
> On Tue, 2008-11-18 at 16:41 -0700, Thomas Fjellstrom wrote:
>> On November 18, 2008, Stefanos A. wrote:
>> >
>> > Would this solve the problem? Is this worth it when compared to
>> the
>> > END_OF_MAIN hack? Any catches?
>>
>> Thats what END_OF_MAIN does on OSX.. But that api might not play
>> nice with X
>> or Windows...
>
> Any reason to hide this from the user? This is how modern operating
> systems work, the single process / single thread / no reentrancy
> model
> of DOS no longer holds.
>
> This code works fine on X11 and Win32 (simply raise the callback
> whenever you hit an idle event). Unless you were referring to
> something
> else entirely...?

The current api has been fairly well thought out and tested.
Changing to some kind of GUI toolkit-alike API at the last minute,
removing a ton of features from Allegro just for this is ridiculous.
END_OF_MAIN has worked with A4 for a long time and there haven't
been any major problems with it. Binaries don't run across operating
systems anyway. And leaving out END_OF_MAIN at least in my
experience always results in a predictable compiler error. There
will be far less confusion if we just keep it if it's needed for
OSX.

Trent :{)>