1GTLEJS@CALSTATE.BITNET (Ed Skochinski) (02/10/90)
I feel compelled to argue for a lowest common denominator: Pascal. The uncommon architecture which makes CYBERs perform so well is what gives rise to the lack of an adequate C compiler for NOS. In general, C likes 8-bit, byte-addressable, VonNeumman architectures, preferably speaking ASCII. Pascal does not depend on national characters, and makes no assumptions about architecture. Pascal is far better defined for the international, multi-vendor environment. Please, could the BITNET pathalias be implemented in ISO standard Pascal? I usually argue for the lowest common denominator---it's just that the CYBERs rarely fall into this category. I don't think that this list is the place to debate the nature of C and the flukes of NOS; C under NOS is a problem, and Pascal is more widely available. If you feel compelled to flame me, send your commentary to me personally, and if I get sufficient time, I'll argue my case. Ed Skochinski Operating Systems Support The California State University
GETTES@PUCC.BITNET (Michael R. Gettes) (02/10/90)
On Fri, 9 Feb 90 09:00:47 PST Ed Skochinski said: >I feel compelled to argue for a lowest common denominator: Pascal. The >uncommon architecture which makes CYBERs perform so well is what gives >rise to the lack of an adequate C compiler for NOS. In general, C likes >8-bit, byte-addressable, VonNeumman architectures, preferably speaking >ASCII. Pascal does not depend on national characters, and makes no >assumptions about architecture. Pascal is far better defined for the >international, multi-vendor environment. Please, could the BITNET >pathalias be implemented in ISO standard Pascal? Well, this is a sticky thicket. I believe this is an issue for the Technical Committee which should set some kind of decree that sites participating in our network should have C or some such beast. Or C and Pascal and therefore all tools would have to be provided in both languages... or whatever. Anyway... my feeling is that if you really want to consider the least common denominator -- IT'S FORTRAN! 'nuf said. I am not going to develop networking software in FORTRAN. Sorry folks. I suggest the CYBER folks make a real push for a better C as the IBM folks did for IBM machines. Keep in mind, the IBM C is still quite lacking in my book. So, as for implementing pathalias in Pascal, unless it is done in a automated fashion of programatic translation, I vote no. I do not think it is an area for significant investment of time. However, if it is something that could easily be translated by some PD translation software with minimal effort following, go for it. I just don't think we should waste a lot of time and effort in porting pathalias to Pascal. /mrg
mwh@IVORY.EDUCOM.EDU (Michael Hrybyk) (02/10/90)
> I just don't think we should waste a lot of time and > effort in porting pathalias to Pascal. > > /mrg > agreed. From talking to Ed, there is a Cyber C which is serviceable but slow. Since the number of tables to be generated on any one Cyber would be small anyway, the performance hit would not be significant. Mike H.