Conference turris::languages

Ahh, Jovial.  (For the Youth in the audience, the acronym stands
for "Jules' Own Version of the International Algorythmic Language".

"Jules" is Jules Schwartz, of, I think, SDC or Rand.  He also developed
the first multi-user timesharing system, called JOSS -- the language
it used is the direct ancestor of both Focal and Mumps.

The International Algorithmic Language is a reference to Algol-58 and/or
Algol-60.  It's a pretty straight-forward block-structured language
that was used on a number of U.S. military computers.  It's certainly
no worse a language than a lot of the junk we use today.

You could probably write a Jovial->C or Bliss translator pretty easily.

Martin.

The compiler has options for code-generators for military computers that
we don't have specs for.  A source level translation to another language
won't help unless we can duplicate the code generator which may border on
the impossible, or even if we could it would be unacceptable to the
Air Force without substantial validation.

Just maybe - somebody's working on Jovial to ADA translation. 

Needed Jovial on VMS or UNIX.

Pat Sweeney

Over the long term you may see Jovial=>Ada translators.  Until there are
an overabundance of Ada compilers, you may see some Ada=>Jovial translators
for projects that are required to write Ada source files.

I assume the Air Force isn't pulling a fast one by using Jovial instead of
Ada.  The project must not be "mission critical" or something.

On a more constructive note, the Jun-83 10/20 LCG Software Referral Catalog
lists "Proprietary Software Systems, Inc." as the vendor of a Jovial/J73
compiler, no doubt for the PDP-10.  Perhaps they have one for the Vax?
Their address is:

	9911 West Pico Boulevard
	Penthouse K
	Los Angeles, CA 90035
	(213) 553-2997

Oh, they say explicitly that that compiler is for Tops-10/Tops-20.  But they
still might have a Vax offering.  It says "PSS has developed numerous
targets and has also rehosted the compiler.  PSS specializes in generating
highly optimized code generators for this compiler".
				/AHM

Have you try the Jovial User's Group?  Also the Airforce or some such MUST be
able to provide you with a list of many of the validated Jovial compilers,
may be filtered to just those which  host/target VAX.

Apparently one of the big problems with Jovial is for many years it was a
moving target, with many projects NEVER able to use a validated compiler 
because the compiler they would start with would soon fall behind, but they
couldn't do their complex integration testing fast enough to move to the next
compiler before it too fell behind.  

There are still waivers being issued for DoD projects that 'require' Ada, 
given the current availability issues...


/Bevin

Sanders Associates, here in Nashua, NH has (had?) a Jovial compiler for 
VMS.  However, my wife (who works there) doesn't know if they still have it,
and can only say that they got it from a group out of Wright-Patterson
Air Force Base, in Dayton, Ohio (the armpit of the universe).

    She doesn't know which group, their name or anything else.
    (Either that, or she can't tell me... I'm not cleared for top secret)
    (but I could blackmail her... :-) )

Hope this helps.

sar

There is a validation centre at Wright Paterson for Jovial, Ada, and goodness
knows what else.  

Hope this avoids some matrimonial strife... :-)

/Bevin

My best guess is you should contact this place...

	JOVIAL Language Control Facility,
	    ASD/ADOL,
	    Computer Operations Division,
	    ASD Computer Centre,
	    Wright Patterson AFB
	    OH 45433    (513) 255-4472/4473

	ARPANET: LCF@WPAFB-JALCF

for a list.




The brochure I have lists a variety of VAX and 20 hosted compilers targetted
to VAX.  Included and looking the most likely are...

Try
    J73/1750/1750A tools                (validated)
    Sharma Simmons
    ASD-AFALC/AXTS
    WPAFB OH 45433
    (513) 255-5945

or
    Martin Marietta & TI compilers      (completed)
    Contact:
	    Gil Weisbord
	    Software Engineering Assoc.
	    23864 Hawthorne Blvd
	    Suite 200
	    Torrance CA 90505

	    (213) 373-8901


/Bevin

There are a number of existing incompatible flavors of Jovial running around
the Air Force as well. If this project has to do with the B-1, you might be
stuck with the older J3/B instead of the (somewhat) cleaner J73.

	Jerry

	I have read the preceding replies, but I'm still adding this
	one, for the following reasons:

	(1) I have always wanted to write a reply after a lapse of
	    six years.

	(2) It's always possible something may have changed in the
	    six years. (Yes, I know, everyone has forgotten about
	    JOVIAL - that's a change, right?).

	(3) I already said I was cross-posting this note, so it's
	    easier to do so than go back and edit it out.

	Seriously, there are some big, important systems out there that
	don't realise that JOVIAL is obsolete. They're still running
	it, and occasionally we get a stab at updating one. So...

					-0-

 	Does anyone know of any facilities for converting JOVIAL code
	into any modern structured language suitable for critical
	real-time applications? (Ada would be preferable, but others
	such as C might be considered).

	This question has been raised occasionally in the past in the
	TURRIS::ADA conference, without any joy.

	The point is that a lot of the existing system, which is very
	large, and has been working satisfactorily for many years, is
	written in JOVIAL. It might be useful to have the option of
	reusing this code to whatever extent is possible. Unfortunately
	there is, to my knowledge, no JOVIAL compiler for VMS.

	This note is cross-posted in TYFYS::CONVERSION, TURRIS::ADA,
	and TURRIS::LANGUAGES.

	/Tom

    Re .9:
    
    A few (five?) years ago, when I was following the Ada literature much
    more closely than I do today, I could swear that I heard something
    about one or more companies or other groups (ie, funded research
    projects) attempting to produce a JOVIAL ==> Ada translator(s).  In
    fact, back then (perhaps still) ACM SIGAda and JUG were holding joint
    conferences, and I'm sure that there was *lots* of interest in such a
    translator.
    
    Unfortunately, I don't *know* of any translator.
    
    I'd suggest that you contact one or more of:
    
    	ACM SIGAda (at a minimum, scan back copies of *Ada Letters*)
    	Software Engineering Institute at CMU
    	the Validation folks at Wright-Patterson.
    
    You could also contact the AJPO (Ada Joint Program Office) or whatever
    exists today to provide the same information/coordination function.
    
    There may very well be a public-domain tool developed as part of a
    government contract, and even though it may not be advertised as well
    as one would like (advertise a free tool?), I'd wager that much of the
    outside-Digital Ada community may have heard of it, even if we have
    not.
    
    Lloyd