Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP Path: utzoo!mnetor!uunet!seismo!columbia!rutgers!ames!sdcsvax!ucsdhub!hp-sdd!hplabs!hp-pcd!uoregon!dboyes From: dboyes@uoregon.UUCP (David Boyes) Newsgroups: comp.os.vms Subject: Re: Yet another bug in VMS C Message-ID: <347@uoregon.UUCP> Date: Fri, 31-Jul-87 03:19:32 EDT Article-I.D.: uoregon.347 Posted: Fri Jul 31 03:19:32 1987 Date-Received: Sun, 2-Aug-87 02:26:49 EDT References: <8707291008.AA17919@ucbvax.Berkeley.EDU> Reply-To: dboyes@drizzle.UUCP (David Boyes) Distribution: world Organization: University of Oregon, Computer Science, Eugene OR Lines: 84 In article <8707291008.AA17919@ucbvax.Berkeley.EDU> writes: > > I have discovered truly bizarre behavior in C for VAX/VMS. The . . . > [... a perfectly valid question about a really DUMB feature of > VMS...] > > How does one tell it not to do it? I could find no references to this > in the index of the VMS C manual. > >You didn't look in the index very hard, did you? It took me about 30 seconds Thanks for the grace and charm, bucko. Remind me to return the favor sometime when YOU need a question answered. >to find the index entry "Main function, syntax of". A quick look at the >single section that points to makes it clear that this is the place to find >the details. Reading on, we find (page 11-31 in Programming in VAX C - the >V2.0-V2.2 version): > > DCL converts most arguments on the command line to uppercase > letters. However, VAX C internally parses and modifies the > altered command line to make VAX C argument access compatible > with C programs on other systems. > Seems like a pretty dumb thing to do -- it's one of the most non-intuitive things I've ever seen. Jeez, even IBM doesn't muck with the case of command line arguments(unless you ask it to). > >This is NOT a bug in C; in fact, it has as little to do with C as the fact >that the Unix shells will interpret various metacharacters unless you quote >them. > >There is no way to prevent DCL from forcing all the unquoted letters to upper >case, Why not? >and there is nothing VAX C can do to restore the original case - the >information is simply not available to it. Then why do they do it in the first place? > >Now, repeat 10 times: > > I will not assume something isn't in the manual until I've looked. Sigh. Like IBM, DEC manuals are badly written, poorly indexed, and really require you to know what you're looking for before you start looking for it. Also, since a full set of VMS manuals now costs upwards of $5K, did you consider the possibility that the manual might not have been available to him? It's not an uncommon thing at universities to have the manuals locked away somewhere where only the gurus can get at them. Cut him some slack, ok? > I will not assume something is a bug because I don't understand it. C'mon, guy. Even YOU have to admit that this sort of behavior for argument processing makes very little sense, especially since VMS provides no way to get at the unmodified command line. Besides, isn't this newsgroup intended to provide information about working and programming under VMS? He knew enough to ask about behavior he didn't understand -- let's give the fellow credit for being willing to stand up and admit confusion. You remember confusion -- that stuff you felt when first faced with an unexplainable weirdness in a program or OS? > I will not assume something is a bug because it's not exactly like > Unix. This can be expanded to ANY OS. I could say the same about VMS, or Unix, or MVS for that matter. The behavior he has observed is certifiably silly -- how tough is it to provide one more syscall to the 1x10e9 that VMS already has to get the unmunged command line? Give us a break, ok? In summary, lighten up. At least this fellow is trying. -- David Boyes ARPA: 556%OREGON1.BITNET@WISCVM.WISC.EDU Systems Division BITNET: 556@OREGON1 University of Oregon Computing Center UUCP: dboyes@uoregon.UUCP