TPLP Special Issue on Locic Programming and the Web

URL: http://www.dsi.unive.it/~tplp/

================================================================
Material:
================================================================

JW:

	* ICLP-05
	Optimising SeRQL expressions
	
	* ISWC-05
	Triple20: triple-centered SW application design

	* ISWC-03
	RDF-DB: Scalability & performance

	* SWI-Prolog HTTP documentation (Threaded HTTP & HTML-WRITE)

LM:

	* Facetted based search using Prolog as web document processing
	tool

ZH:

	* XDIG
	Overview of XDIG architecture (generic DIG I/O), PION and MORE

Optional:
	
	* E-Culture demo
	Web-service on the basis of SeRQL infrastructure

================================================================
Message
================================================================

We MUST make clear why LP is a good pardigm for Web Processing.

Why is LP good for Web Processing:
----------------------------------

	* Safe & automatic memory management (same for Java)

	* Compact

	* Good at rewrite (DCG)


Why is LP bad for Web Processing:
---------------------------------

	* Too small community

	* Mostly inexperienced programmers

	* Not enough reusable resources (libraries)


Highlights we have
------------------

	* SeRQL implementation has a nice point: compilation of
	SeRQL to Prolog, followed by subclause reordering using
	generate-and-test or programs: simple and elegant.

	* I think the PION and MORE components could make a message
	like that, but the XDIG paper is not detailed enough.

	* Lourens must make the message that LP offers a nice and
	compact rewrite system.
