Mail Archives: djgpp/1997/05/22/05:18:42
On Wed, 21 May 1997 22:17:58 GMT, dbrotherhood AT geocities DOT com (Michael
D. Ryan) wrote:
>I think DJGPP is plenty fast, a lot smaller than my version of Borland
>and a whole lot easier to use. I have never received a memory error
>that I could not deal with using DJGPP. With Borland, I scrapped many
>programs because I couldn't find what was causing it to crash...
I've been working with Turbo C V2 for the last ten (!) years and am
new to DJGPP. I never upgraded Turbo C for the reasons you've just
mentioned. Later versions were bulky and had some strange bugs in
them. I patched my Turbo C a long time ago and found no errors since.
I must say DJGCC is bulkier than Turbo C, but the opportunity to get
rid of these segments was too good to leave alone. It causes headaches
where there shouldn't be any. Furthermore, I love that it catches
segment violations. I used to have a complete Unix OS to catch just
these things. There rarely show up in DOS. Sometimes I use the
debugger in Turbo C, since it is that easy. But rarely.. I usually
debug the old-fashioned way ;)
>IMHO, DJGPP is the best 32-BIT C compiler I have found, and like DJ
>says, it's free.. I have used a lot of compilers and this one is the
>best hands down.
I quick-compiled my Forth bytecode VM with -O3 and it performed twice
as fast.. Love it!
>I also encourage everyone to use it. It is hard to use at first
>(since there is no IDE type interface), but once you learn, there is
>no going back...
I usually work witout one. Once you get the hang of C, you don't need
one I guess.. So there is no problem. The only time I startup the
Turbo C IDE is to consult the HELP quickly.
Hans.
- Raw text -