home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.lang.fortran
- Path: sparky!uunet!cs.utexas.edu!uwm.edu!linac!unixhub!slacvx.slac.stanford.edu!fairfield
- From: fairfield@slacvx.slac.stanford.edu
- Subject: Re: How common are long variable names?
- Message-ID: <1992Dec11.174722.1@slacvx.slac.stanford.edu>
- Lines: 49
- Sender: news@unixhub.SLAC.Stanford.EDU
- Organization: Stanford Linear Accelerator Center
- References: <1992Dec11.221459.10841@sol.UVic.CA>
- Date: Sat, 12 Dec 1992 01:47:22 GMT
-
- In article <1992Dec11.221459.10841@sol.UVic.CA>, klassen@sol.UVic.CA (Melvin Klassen) writes:
- > Dr. Kenneth H. Fairfield <fairfield@slacvx.SLAC.Stanford.EDU> writes:
- >> Hurray for IBM's VS Fortran!!! 31-character identifiers allowed,
- >>but if you should be so bold as to use an identifier > 7 characters long
- >>for a global symbol (i.e., SUBROUTINE, FUNCTION, ENTRY, BLOCK DATA, or
- >>COMMON block name), it takes the first 4 characters and concatenates
- >>them with the last 3 characters of the (long) identifier to generate the
- >>external name. Think about the following (generic) utilities:
- >>
- >> SUBROUTINE UTL_GET_DATA
- >> SUBROUTINE UTL_PUT_DATA
- >> SUBROUTINE UTL_CLEAR_DATA
- >
- > If you CALL more than one of these routines from a single program-unit,
- > the compiler reports the error-message:
- > ILX1454I Two or more external symbols are not unique.
- > Check the "external symbol report".
- >
- > The ICA (InterCompilation Analyzer) reports the diagnostic:
- > ILX0067I Conflicting name usage -- the name, UTL_ATA has been used
- > as the name of a subroutine in UTL_ATA (compilation 2) and
- > as the name of a subroutine in UTL_ATA (compilation 3).
- >
- > The compiler is doing as best as it can!
- > It's the IBM linker/loader which is the bottle-neck,
- > due to the restriction to 8-character external names,
- > so don't blame the IBM FORTRAN-development group
- > for a problem which they have no authority to fix!
-
- Well, I _didn't_ "name names", you did ;-) I didn't say "the IBM
- FORTRAN-development group" stinks because of this kluge, I said, in
- effect, "look what happens" if you use long names with VS FORTRAN. And
- it's not just the linker/loader people, it goes all the way down to the
- file system in some cases...
-
- My posting was in sympathy to J. Giles' (?) remark about compilers
- which _accept_ long names but, in essence, don't _support_ them. Yes, at
- least IBM is flagging the possible conflict... My personal opinion is
- that it would be better for them to say, "We support long variable names,
- but externals still need to be limited to 7 characters". That would be
- clear; the other lays a trap for the unwary.
-
- -Ken
- --
- Dr. Kenneth H. Fairfield | Internet: Fairfield@Slac.Stanford.Edu
- SLAC, P.O.Box 4349, MS 98 | DECnet: 45537::FAIRFIELD (45537=SLACVX)
- Stanford, CA 94309 | BITNET Fairfield@Slacvx
- ----------------------------------------------------------------------------
- These opinions are mine, not SLAC's, Stanford's, nor the DOE's...
-