Path: ...!eternal-september.org!feeder3.eternal-september.org!weretis.net!feeder8.news.weretis.net!reader5.news.weretis.net!news.solani.org!.POSTED!not-for-mail
From: Mild Shock <janburse@fastmail.fm>
Newsgroups: comp.lang.prolog
Subject: =?UTF-8?Q?Re:_Can_we_trust_the_Scryer_Prolog_Gurus=3f_=28Was:_A_har?=
 =?UTF-8?Q?sh_wind_is_blowing_into_the_face_of_Prolog_now=e2=80=a6_=29?=
Date: Tue, 23 Jul 2024 00:42:44 +0200
Message-ID: <v7mn93$72fa$1@solani.org>
References: <v32hjd$s3qh$1@solani.org> <v7mmbj$71vr$1@solani.org>
 <v7mmi0$725n$1@solani.org>
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 8bit
Injection-Date: Mon, 22 Jul 2024 22:42:43 -0000 (UTC)
Injection-Info: solani.org;
	logging-data="231914"; mail-complaints-to="abuse@news.solani.org"
User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:91.0) Gecko/20100101
 Firefox/91.0 SeaMonkey/2.53.18.2
Cancel-Lock: sha1:DqzoL3yNUM9CbkKvJ2fSbrEarbs=
In-Reply-To: <v7mmi0$725n$1@solani.org>
X-User-ID: eJwNyMEBwCAIA8CVikCC49gI+4/Q3vPSYRADicjJKXKv227jDey6/ahSaS26l8VZGuXa8wfCR3jPcZHCZfIDT8wVtA==
Bytes: 4981
Lines: 102

Well I am mistaken, Prolog 0 must have had some
concept of error. For example I find:

3. OPERATION SUR LES NOMBRES
=============================
LES PREDICATS SONT EVALUES A ERROR

And this is also used in a code_type variant, i.e.
CHIFFRE and LETTRE, follow the instantiation error
idea. But we have this also in Prolog systems today:

/* SWI-Prolog */
?- code_type(0'a, X).
X = alnum .

?- code_type(X, X).
ERROR: Arguments are not sufficiently instantiated

It is not the case that modern Prolog systems always
silently fail. They silently fail in atom/1, integer/1, etc..
which makes sense, especially from a WAM implementation

viewpoint, since WAM has usually type tag branching or
switching instructions. So these atom/1, integer/1, etc.. can
be implemented quite efficiently and one should view them

as belonging to the same category as var/1, (==)/2, etc..
i.e. meta predicates that deal syntactically with Prolog terms.
Some Prolog system can even perform indexing on these guards.

Mild Shock schrieb:
> That the DEC 10 Prolog 1975 is close to Prolog 0,
> can be verified by reading the Prolog 0 manual:
> 
> MANUEL DE REFE RE NeE ET D'UTILISATION - PROLOG
> ROUSSEL Ph. (1975)
> http://alain.colmerauer.free.fr/alcol/ArchivesPublications/ManuelProlog/Pr.pdf 
> 
> 
> So at that same year there was already an
> English rip-off. If I read the french, I also
> don't find some atom/1, integer/1 equivalent
> 
> that would throw an instantiation error. Problem
> is again, what would have been an exception in Prolog 0?
> 
> Mild Shock schrieb:
>> For example one Guru claimed?
>>
>>  > Prolog were invented today, I think there would
>>  > be at least two significant differences:
>>  >
>>  > First, the type-testing predicates like atom/1,
>>  > integer/1 and compound/1 would (and should) throw
>>  > instantiation errors if their arguments are not
>>  > sufficiently instantiated.
>>  >
>>  > This is also what the original versions of Prolog
>>  > did. However, DEC 10 Prolog chose to replace instantiation
>>  > errors by silent failures, and this has been
>>  > perpetuated in the Edinburgh tradition for type tests
>>  > including the ISO standard.
>> https://www.quora.com/If-prolog-were-being-invented-today-with-no-concern-for-backward-compatibility-or-the-existing-standardization-how-would-it-differ-from-standard-prolog 
>>
>>
>> I cannot verify any of the above nonsense.
>>
>> First of all the term "DEC-10 Prolog" is ambigious:
>>
>> DEC 10 Prolog 1975
>> https://www.softwarepreservation.org/projects/prolog/prolog/edinburgh/doc/Warren-Epilog_400_400-1975.pdf 
>>
>>
>> DEC 10 Prolog 1982
>> https://userweb.fct.unl.pt/~lmp/publications/online-papers/DECsystem-10%20PROLOG%20USER%27S%20MANUAL.pdf 
>>
>>
>> The DEC 10 Prolog 1975 looks very close to Prolog 0
>> with its french predicate names. There is not a simgle
>> atom/1, integer/1 equivalent that would throw an
>>
>> instantiation error. Actually Prolog 0 didn't even
>> have some sort of exceptions, right?
>>
>> Mild Shock schrieb:
>>> Especially since good old FORTRAN has
>>> made a new appearance:
>>>
>>> TIOBE Index for May 2024
>>> I have received a lot of questions why Fortran entered the top 10
>>> again after more than 20 years. The TIOBE index just publishes
>>> what has been measured.
>>> https://www.tiobe.com/tiobe-index/
>>>
>>> Why Fortran is back in TIOBE’s top 10
>>> First, Fortran is especially good at numerical analysis and
>>> computational mathematics. Numerical and mathematical
>>> computing is growing because interest in artificial intelligence
>>> is growing, Jansen told TechRepublic in an email.
>>> https://www.techrepublic.com/article/tiobe-index-may-2024/
>>
>