blarson@dianne.usc.edu (bob larson) (12/03/89)
In article <89Dec2.021206est.2151@neat.cs.toronto.edu> moraes@cs.toronto.edu (Mark Moraes) writes: >[Geoff Collyer also warns that you had better use gcc -traditional -- > he staunchly denies that C News is written in ANSI C, and insists > that compiling C News with an ANSI C compiler (alleged or otherwise) > will void your warranty...] Does this mean than C news won't work on a C compiler without the "traditional" bugs, (such as using comments as token pasting, and substituting macro paramaters in quoted strings) or just that it doesn't take advantage of the new features in ANSI C. If the former, it means I have a bigger job in front of me, I'm contemplating porting C news to os9/68k, which has a C compiler that doesn't have these bugs. (ANSI version will be done, but isn't yet available.) (Yes, I realize this means translating all the shell scripts to C.) Bob Larson blarson@dianne.usc.edu usc!dianne!blarson --** To join Prime computer mailing list **--- info-prime-request@ais1.usc.edu usc!ais1!info-prime-request
henry@utzoo.uucp (Henry Spencer) (12/03/89)
In article <21546@usc.edu> blarson@dianne.usc.edu (bob larson) writes: >> ...compiling C News with an ANSI C compiler (alleged or otherwise) >> will void your warranty...] > >Does this mean than C news won't work on a C compiler without the >"traditional" bugs, (such as using comments as token pasting, and >substituting macro paramaters in quoted strings)... We don't commit such vilenesses. Our attitude towards Reiserisms like this resembles our attitude towards cockroaches. > ...or just that it >doesn't take advantage of the new features in ANSI C... The main problem is areas like function prototyping. Personally, I think Geoff exaggerates the non-ANSIness of the code, but there are problems here and there with making things acceptable to *both* current compilers and ANSI compilers, and these have generally been resolved in favor of current compilers. >... I'm contemplating porting C news to os9/68k... >(Yes, I realize this means translating all the shell scripts to C.) Between the shell scripts and the awk scripts, this will be a fair-sized job, but the C code should not be a major problem. -- Mars can wait: we've barely | Henry Spencer at U of Toronto Zoology started exploring the Moon. | uunet!attcan!utzoo!henry henry@zoo.toronto.edu
davidsen@sungod.crd.ge.com (William Davidsen) (12/05/89)
In article <1989Dec2.235748.7304@utzoo.uucp> henry@utzoo.uucp (Henry Spencer) writes: | The main problem is areas like function prototyping. Personally, I think | Geoff exaggerates the non-ANSIness of the code, but there are problems | here and there with making things acceptable to *both* current compilers | and ANSI compilers, and these have generally been resolved in favor of | current compilers. I don't know if the problems are exagerated or not, but I gave up trying to get it to compile under the latest, most ANSIfied, C compiler for Xenix. The problems were in the procedure types among other things. If I had a good reason to change I would, but I spent about four hours hacking at the sources, and finally got it to compile only to find that I had broken something. I suspect that it would compile fine with pcc or even the old xenix compiler, but that's the price you pay for having the latest version. bill davidsen (davidsen@crdos1.crd.GE.COM) {uunet | philabs}!crdgw1!crdos1!davidsen "Stupidity, like virtue, is its own reward" -me