strange behaviour of f$license

[joukj]

Hi All,

I'm running the following on a IA64 OpenVMS 8.4

bolero-jj) write sys$output f$license("C")
FALSE
bolero-jj) sh lic c

Active licenses on node BOLERO:

------- Product ID -------- ---- Rating ----- -- Version --
Product Producer Units PCL Activ Version Release
Termination
C DEC 0 1 100 0.0 (none)
1-OCT-2012


Can someone explain why f$license returns "FALSE" in this case?

Jouk

[Heuser]

Not "strange". The default producer is "HP".

write sys$output f$license("C","DEC")

give you the answer you want.

Eberhard

[joukj]

Thanks, I overlooked this. Seems to be different for Alpha and IA64.

Strange that the "default" is HP, while almost all my "educational
licenses have DEC as producer

Jouk

[VAXman-]

LEXICALS

F$LICENSE

Arguments


license-name

Specifies the name of the license for which you want to check the
status.


producer-name

Specifies the name of the company that produced the license. By
default, DEC is assumed to be the producer on Alpha systems and
HP is assumed to be the producer on Integrity server systems. To
find an exception, specify a different producer name.

--
VAXman- A Bored Certified VMS Kernel Mode Hacker VAXman(at)TMESIS(dot)ORG

Well I speak to machines with the voice of humanity.

[Phillip Helbig---undress to reply]

In article <b618a$4f314454$82a13ce5$29...@news1.tudelft.nl>, joukj
<jo...@hrem.nano.tudelft.nl> writes:

> Strange that the "default" is HP, while almost all my "educational
> licenses have DEC as producer

Yes, quality lingers on in some corners. Steve Lionel told the
interesting story of the "DEC" used as part of an identifier for
preprocessor directives for high-performance Fortran. (To a non-HPF
compiler, they are just comments and ignored, but an HPF compiler will
do something with them.) After Compaq bought DEC, someone suggested
they should get a new name. Steve replied that it stood for
Directive-Enhanced Compilation (actually a pretty good description of
the functionality) and they stayed in. :-)

[joukj]

Do you also know where MS is to stand for? because at present the
Fortran compiler on OpenVMS seems to accept as directive-identifier.

Jouk

[George Cornelius]

Phillip Helbig---undress to reply wrote:

> Yes, quality lingers on in some corners. Steve Lionel told the
> interesting story of the "DEC" used as part of an identifier for
> preprocessor directives for high-performance Fortran. (To a non-HPF
> compiler, they are just comments and ignored, but an HPF compiler will
> do something with them.) After Compaq bought DEC, someone suggested
> they should get a new name. Steve replied that it stood for
> Directive-Enhanced Compilation (actually a pretty good description of
> the functionality) and they stayed in. :-)

I don't know what a Fortran preprocessor directive is, but
historically we have had the /D_LINES option - see below.

[A fixed format line is no doubt one that honors the historical
card column alignments: 1-5: statement number if numeric, comment
if col 1 = 'C', 6: continuation indicator, 7-72:language statement].

Probably unrelated, but potentially a source of confusion
to those of us who have not used HPF (High Performance
Fortran?).

George

--------------------------------------------------------


EISNER> help/nopage/noprompt fortran /d_lines

FORTRAN

/D_LINES D=/NOD_LINES

/[NO]D_LINES

Specifies that the compiler should treat fixed-format lines that
contain a D in column 1 as source code rather than comment lines.

The default is /NOD_LINES, which means that fixed-format lines with
a D in column 1 are treated as comments.

This qualifier is ignored for free-format source code.

--------------------------------------------------------

[glen herrmannsfeldt]

George Cornelius <corn...@eisner.decus.org> wrote:

(snip)

> [A fixed format line is no doubt one that honors the historical
> card column alignments: 1-5: statement number if numeric, comment
> if col 1 = 'C', 6: continuation indicator, 7-72:language statement].

Still standard as of Fortran 2008, along with the alternative
free-form version.

-- glen

[Ken Fairfield]

On Thursday, February 9, 2012 1:54:23 PM UTC-8, George Cornelius wrote:
> Phillip Helbig---undress to reply wrote:
> > Yes, quality lingers on in some corners. Steve Lionel told the
> > interesting story of the "DEC" used as part of an identifier for
> > preprocessor directives for high-performance Fortran. (To a non-HPF
> > compiler, they are just comments and ignored, but an HPF compiler will
> > do something with them.) After Compaq bought DEC, someone suggested
> > they should get a new name. Steve replied that it stood for
> > Directive-Enhanced Compilation (actually a pretty good description of
> > the functionality) and they stayed in. :-)
>
> I don't know what a Fortran preprocessor directive is, but
> historically we have had the /D_LINES option - see below.

[...]

Preprocessor directives are comment lines in a particular
format that the compiler recognizes and processes.

Unlike D_lines, which are treated either (a) as normal
comment lines without the qualifier, or (b) as source
code to be compiled if the qualifier is present, the
directives give instructions to the compiler that
change it's behaviour. Similar but not identical to
C preprocessor directives.

The Directive-Enhanced Compilation lines (brilliant
of Steve Lionel to come up with that post-hoc
meaning :-) is:

!DEC$ ...stuff...

Since VMS Fortran long supported the extension of
using the exclamation point as a valid comment character,
and indeed Fortran 90 standardized that usage, a compiler
that doesn't recognize the directives will simply skip
over them as normal comment lines. The VMS and Intel
Fortran compilers (as well as the earlier Digital & Compaq
Visual Fortran on Windows) do recognize these, although
there may be others as well.

I can't comment on how they're actually used since
I've never needed them.

-Ken

[Phillip Helbig---undress to reply]

In article
<12227592.170.1328828300653.JavaMail.geo-discussion-forums@pbbr5>, Ken

Fairfield <ken.fa...@gmail.com> writes:

> I can't comment on how they're actually used since
> I've never needed them.

Neither have I. Basically, you can tell the compiler how to distribute
things across a multi-processor machine and so on.

[Bob Koehler]

In article <4f34408f$0$63199$815e...@news.qwest.net>, George Cornelius <corn...@eisner.decus.org> writes:
>
> I don't know what a Fortran preprocessor directive is, but
> historically we have had the /D_LINES option - see below.

D lines were a vendor specific extension for adding "debug" lines,
where D would be treated as a comment unless /D_LINES was specified.
I've seen a couple vendors support D lines, and at least one use
X instead.

And, of course, if you use free-format source introduced circa
Fortran-90, all comments must start with ! anyhow.

The "preprocessor" statements look like comments to any other compiler,
but do thinkgs above and beyond the Fortran standard if/when you
have a compiler that recognises them. Using CDEC$ at the begining
of the line was assumed not to be recognized as anything other than
a comment by any non-DEC compiler.

The first I recall them being introduced was to support vector
processing on VAXen. Sometimes you had to tell the compiler that
it could make certain assumptions in order to allow code to compile
to vector instructions.

Other uses include control over things that would otherwize require
FORTRAN command qualifiers.