Jim, Ali, and Ed, (01)
I'm back home again, and I'll make a few more remarks. (02)
JR:
> My experience is that most programmers know FOL. They may not know it well,
> granted. They do not use it because of:
>
> 1. lack of a practical methodology for using FOL in the course of writing
> procedural programs (Prolog and its relatives deliberately excepted);
>
> 2. lack of tools for managing large corpora of FOL;
>
> 3. concerns about efficient execution of logic programs;
>
> 4. difficulty of debugging.
>
> This last item is one of my gripes about the semantic web technology, OWL in
> particular. It is very nice to know that you have an inconsistent ontology,
> but you get no help from tableaux reasoners to quickly track down the
> conflicting axioms. (03)
I strongly agree with all those points, especially point #1. (04)
As an example of a real success story for FOL, I frequently cite SQL,
which is the most widely used version of logic on the planet. I used
to call SQL the worst notation for logic ever invented, but that was
before I saw OWL. And I mean that seriously: SQL forms a bridge
between databases and all the major programming languages, and
OWL does not. In fact, programmers don't have a clue about what
to do with OWL. (05)
SQL became successful because (a) there was a need for a bridge
between databases and programming languages, (b) the alternatives
like hierarchical (IMS) and network (CODASYL DBTG) were harder to
use, and (c) Oracle established SQL as the de facto "Intergalactic
Dataspeak" (as Stonebraker called it). So programmers learned
how to use FOL, at least in the SQL notation. (06)
JFS:
>> What got us into this mess is that people said "Oooh, FOL is
>> too hard to learn." So everybody invented a different notation
>> for writing some version of it, and nobody gave any thought
>> to how their version related to anybody else's. (07)
AH:
> This is a remark that resonates and it seems like a missed opportunity
> - it would have been nice had people gone beyond trying to directly
> express FOL in new notations. (08)
I agree. And if there are widely used notations, such as SQL and UML,
it's important to build bridges to them. For years, many people
(such as Jim's former manager, Ted Codd) wanted to extend RDBs with
more semantics, such as a solid type system. If the DL people wanted
to find markets for their technology, they should have started by
using DLs to provide a type system for relational DBs. (09)
That, in fact, is another of my complaints about the Semantic Web.
When it got started, there was a large community of developers who
were familiar with both SQL and UML. They had well established
methodologies for using them with the major programming languages. (010)
There was a golden opportunity for the Semantic Web: provide
an integrated semantic foundation with a smooth transition from
legacy systems to the future. Instead, the W3C declared themselves
to be a complete break with the past. They told the world to ignore
the established *standards* in favor of their own *recommendations*.
That was the height of hubris, and they failed. (011)
AH:
> I believe there would more benefit in developing "notations"
> for the representation of models that satisfy sets of axioms
> expressed in FOL (or Common Logic or some other formal language). (012)
I agree. But I'd emphasize three points: (013)
1. Those notations have to build on what developers already know. (014)
2. The tools must be integrated with current systems and provide
a smooth migration path from legacy systems to the future. (015)
3. There must be good methodologies that show the developers
what to do with the new tools and demonstrate some solid
benefits from using the new facilities. (016)
EB:
> "Safe upon the solid rock, the ugly houses stand.
> Come and see my shining castle, built upon the sand."
> -- Edna St. Vincent Millay (017)
Great quote. The Semantic Web has built a huge castle of syntax
with no foundation in semantics. SQL may be ugly, but it is
based on FOL, and it successfully runs the world economy.
Instead of going with SQL, the DL crowd chose the sand. (018)
John (019)
_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J
To Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx (020)
|