Received: by 10.66.77.131 with SMTP id s3mr1218865paw.24.1352922336561;
        Wed, 14 Nov 2012 11:45:36 -0800 (PST)
Received: by 10.68.143.74 with SMTP id sc10mr6176478pbb.5.1352922336544; Wed,
 14 Nov 2012 11:45:36 -0800 (PST)
Path: s9ni10239pbb.0!nntp.google.com!kr7no15023129pbb.0!postnews.google.com!glegroupsg2000goo.googlegroups.com!not-for-mail
Newsgroups: comp.lang.prolog
Date: Wed, 14 Nov 2012 11:45:36 -0800 (PST)
In-Reply-To: <k80gau$2bd$1@news.albasani.net>
Complaints-To: groups-abuse@google.com
Injection-Info: glegroupsg2000goo.googlegroups.com; posting-host=129.120.60.155;
 posting-account=RZjJKwoAAAB5CuPf8h3-ZShZVHgfq-hI
NNTP-Posting-Host: 129.120.60.155
References: <95ac36f2-fca3-4f0d-b599-7a377413eedc@googlegroups.com>
 <2012Nov14.152438@mips.complang.tuwien.ac.at> <k80gau$2bd$1@news.albasani.net>
User-Agent: G2/1.0
MIME-Version: 1.0
Message-ID: <e356fadd-04f9-431c-9a70-e7ff60fc319e@googlegroups.com>
Subject: Re: ANN: open source Prolog compilers: BinProlog and Jinni Prolog
From: Paul Tarau <pta...@gmail.com>
Injection-Date: Wed, 14 Nov 2012 19:45:36 +0000
Content-Type: text/plain; charset=ISO-8859-1

Both BinProlog and Jinni provide essential logic programming services and a lot of experimental extensions but none of them is very strong on most of the "gourmet features" of the ISO specification. 

My hope by opensourcing them as Google projects is to provide a good starting point with 90% of the work done (WAM, networking, multi-threading, objects, agents, GUI etc.) for someone who would like to build on them. 

In particular, if there's enough interest in their other features, someone could provide the remaining 10% needed for full ISO conformance.

Paul

P.S. My  newest Prolog-like system is Scala based (see
 
http://code.google.com/p/styla/ 

) and it is the only one that I am actively developing.

Still, it might end up diverging radically from Prolog-as-we-know-it, so
in this case ISO conformance would be like requiring an airplane to
flap its wings just because birds do so :-)

 

On Wednesday, November 14, 2012 10:19:42 AM UTC-6, Jan Burse wrote:
> Ulrich Neumerkel schrieb:
> 
> > Are these the only Prolog you are maintaining currently?
> 
> >
> 
> > What are your plans for ISO/IEC conformance?
> 
> >
> 
> > http://www.complang.tuwien.ac.at/ulrich/iso-prolog/
> 
> 
> 
> I would embrace if *all* JVM Prolog systems would
> 
> try to reach ISO/IEC comformance.
> 
> 
> 
> And I would embrace if there are test suites for
> 
> syntax and predicates, that can be executed *unattended*.
> 
> 
> 
> I have already once proposed some predicates that
> 
> would support such test suites. Namely:
> 
> 
> 
>        with_out_to(+Goal,?Container)
> 
>        with_in_from(+Goal,?Container)
> 
>  
> 
> http://www.jekejeke.ch/idatab/doclet/blog/en/docs/std/07_compliance/07_diagnosis/01_runner/02_helper.p.html
> 
> 
> 
> With the above write/read side effects can be tested
> 
> resp. emulated. I already use them in a predicate
> 
> test suite.
> 
> 
> 
> I guess they would be also suitable for a syntax test
> 
> suite, although a very small fraction of the syntax
> 
> test cases would probably not work.
> 
> 
> 
> There is a caveat for this helper predicates. They make
> 
> use of setup_call_cleanup/3 (*) which is not yet standardized.
> 
> So I didn't yet check the JVM Prolog systems, but
> 
> those that don't have setup_call_cleanup/3 would need
> 
> to find a workaround, either implement the helper
> 
> natively or find a substitution for setup_call_cleanup/3.
> 
> 
> 
> But those that have a substitute for setup_call_cleanup/3
> 
> could probably provide setup_call_cleanup/3 in the
> 
> first place?
> 
> 
> 
> Bye
> 
> 
> 
> (*)
> 
> http://www.complang.tuwien.ac.at/ulrich/iso-prolog/cleanup
> 
> 
> 
> BTW.: We might also have a hard time for Jinni:
> 
> ?- setup_call_cleanup(true,true,true).
> 
> prolog.kernel.Machine prolog 0 returned:: 
> 
> exception(error(undefined_predicate_in_metacall,setup_call_cleanup(true,true,true)))
> 
> Restarting Prolog Machine: prolog.kernel.Machine prolog 0