Mail Archives: djgpp/1999/09/08/09:41:45
In article <NrGOGVAqER13EAvt AT lwithers DOT demon DOT co DOT uk>, Laurence Withers <lwithers AT lwithers DOT demon DOT co DOT uk> wrote:
>In article <37d4ce7c DOT 0 AT news DOT uni-bielefeld DOT de>, Manni Heumann
><manfred DOT heumann AT uni-bielefeld DOT de> writes
>
>[snip - no 'const' in Allegro]
[snip demonstration of -fpermisive]
>>It seems the problem really is in the varargs macros and Shawn is off the
>>hook, so the question seems to be what the C++ guys want to do about all this.
>>Suggestions, anybody?
>
>If you are using C++ string objects, you don't really need to use
>Allegro's varargs stuff, because you can use strstream objects, etc.
>Although I admit allegro_message() might be fundamental to a program.
>
You're right, but then again, I'm not that hardcore a c++ fan. I prefer
printf-style functions to strstreams, at least in those kind of situations,
where I use Allegro's textprintf, e.g. This will simply results in less lines
of code and decreased chances of making silly mistakes.
>If you are willing to wait, I'm trying to write a set of C++ wrappers
>for Allegro which will, at least, solve the lack of const problems.
That sounds good, if you need some help, just say so.
>Whether making the Allegro sources const-correct would be worthwhile, or
>even possible, I'll leave to somebody more experienced to decide.
>
No doubt, it would be possible. But there is a price to pay, and it wouldn't
make Allegro's clear cut code more beautifull.
--
Manni
- Raw text -