pyle@ut-ngp.UUCP (02/06/87)
We've all seen sources posted to the net when a comment such as "ported to
the PC using the Acme C compiler". (Substitute your favorite compiler for
Acme :-).) Since I only have experience with Microsoft C, my question is
"how different are these various implementations of C?"
What I would like to have is a compilation of the differences between
compilers X and Y which encompasses the more common C compilers for the PC.
In this group would be Microsoft C, Lattice C, DeSmet C, etc. Please don't
flame at me if your favorite isn't named specifically. If those of you on
the net who have knowledge of the differences between two compilers will
document and mail me the information, I will attempt to prepare a summary
and post it to the net. In general, I think the things that would be of
interest are: differences in names of library functions that perform
identical tasks, presence/absence of common/useful functions in a particular
compiler (e.g., string functions, etc.), differences in calling arguments
or return values/types, and differences in capabilites of functions (e.g.,
signal trapping). Please don't be limited by these examples, however.
Please feel free to send whatever you have. I would rather be deludged
with responses than to have everyone say "Somebody else has probably
already sent this". Note that if you don't see my summary in the next
couple of months, then I probably didn't get any or many responses.
Keith Pyle
UUCP: . . .{ihnp4,seismo,ctvax,noao,gatech}!ut-sally!ut-ngp!pyle
ARPA: pyle@ut-ngpjeff@spp2.UUCP (02/10/87)
In article <4631@ut-ngp.UUCP> (Keith Pyle) pyle@ut-ngp.UUCP writes: > >We've all seen sources posted to the net with a comment such as "ported to >the PC using the Acme C compiler". >"how different are these various implementations of C?" > We sure do need such a list. We also need a good compiler validation suite. One that tries all the features of a language and reports on how the compiler handled them. And the results from that when run against all the popular compilers. Are you listening BYTE? How about it C language gurus?
thomps@gitpyr.UUCP (02/12/87)
In article <1236@spp2.UUCP>, jeff@spp2.UUCP (Jeff Hull) writes: > In article <4631@ut-ngp.UUCP> (Keith Pyle) pyle@ut-ngp.UUCP writes: > > > We sure do need such a list. We also need a good compiler validation > suite. One that tries all the features of a language and reports on how > the compiler handled them. And the results from that when run against all > the popular compilers. > A validation suite requires something to validate against. Currently there exists no standard C language. Kernighan and Ritchie might be called a standard by some but it was not intended to be and does not rigorously define the language. I expect C validation suites will become common after the ANSI standard is published. As it is, what would you validate against. -- Ken Thompson (No not that Ken Thompson) Georgia Tech Research Institute Georgia Insitute of Technology, Atlanta Georgia, 30332 ...!{akgua,allegra,amd,hplabs,ihnp4,seismo,ut-ngp}!gatech!gitpyr!thomps
jeff@spp2.UUCP (02/15/87)
In article <3085@gitpyr.gatech.EDU> thomps@gitpyr.gatech.EDU (Ken Thompson (No not that Ken Thompson) ) writes: >A validation suite requires something to validate against. >... >I expect C validation suites will become common >after the ANSI standard is published. > >As it is, what would you validate against. I think a validation suite is any set of programs that evaluate specified features of a language/compiler. True, most validation suites test features specified for a language by some standard, but not all. I think it is none too soon to start developing a validation suite for C even though the ANSI standard is not yet finalized. I also think it would be a good idea if our validation suite reported on BOTH Kernighan & Ritchie and ANSI compatibility of compilers. Jeff Hull decvax,hplabs \ 13817 Yukon Avenue ihnp4,sdcrdcf -> !trwrb!trwspp!spp2!jeff Hawthorne, CA 90250 ucbvax,vortex / It was great when it all begaaaaaaaaan, I was a regular <USENET> faaaaaaaaaan, ...