Celebs at work.

Carefully choosing the Balmoral track this morning... where I get splendid entertainment from first (mr. SAX1) David Megginson, and (mr half a parser) Simon St Laurent. The main memories from David's presentation:

• The (only useful) possible use for XML Dave is seeing is in publishing DATA online (for all the rest, there is already quite some stuff out there to everyones satisfaction)
• Folowing up on his talk there is http://www.megginson.com/pubxs/ to keep track of the status

And the funny ones:

• XML is a solution in search for a problem.
• Standardisation organizations are (in terms of biological reproduction strategy) Frogs.

[the comparison was between humans having little offspring which they nurture and protect, while frogs just throw around their eggs (specs) and don't really worry that a lot of them do not survive]

During QA: ReSTy Paul Prescod in the room relates to HTTP publishing data and having xpointer-like references to retrieve specific data.

I make the mental note of sending them both a reference to xreporter.

Validation stinks!

Efficient presentation by the people of DecisionSoft on a proposal for possibly expressing arbitrary business rules inside an XML Schema (or did you think it was already complex enough?)

I find myself feeling quite alone in the room with some remarks on

• the fact that validation is already handled in systems out there, and reinventing the wheel will not make anyones lives easier, it will just add another copy of the same business logic to the scene
• programming/scripting languages with an angle bracket syntax are to be abolished (hmm, ant too?)
• if the business analyst is really the person we want to help, then formal modelling techniques like adding OCL to the UML might be making more sense then XML-implementation specific schema-extension

Of course... to a lot of the people at the conference XML is to be seen as conceptual esoteric manna for your brain, and is surely not the handy implmentation stuff I take it for...

Sorry, for thinking it's just XML.

Jacek again.

Next to Gregor this young man has a beautifull full-text-zone-aware search engine lingering around on the laptop's HD.Mmm, maybe he's not all that lazy :-)

And we get to talk in metafors now. He wants to keep Gregor (15 months of personal coding investment) as his own (closed source) product to build a business upon. After the build-up respect in two mindsharing days I cannot make the OSS arguments about doing it different, and can only whish him all the luck. He is after all right: people that are satisfied with the free/open family car can do so and drive wherever they like, he'll be glad to sell the F1 version to those people that need it and can afford it.However: the analogy (as the good ones do) goes quite a long way: his F1 is to be used on the racing track only.

Maybe I can plant some thought in his brain as well: From as little of what I understand from what Gregor is doing, part of the pudding-proof is in how the parser-API is presenting the XML content to the translets... And basically a whole wealth of SAX based applications (why am I thinking Apache Cocoon here) could benefit from the power-driven ideas behind that.

Jacek is pretty concerned with his intelectual property here (understandable with his intelect), but I honestly think his business could only benefit from the fact that all roads out there would be getting upgraded to racing-standards... it would allow his car to seemlessly drive home after the spinning rounds on he current racing track he built for himself...

Mental note to me: I left him with the idea of making public the (non-implementation) API of some, let's call it, sax-ng. I need to follow up on that, and actively help him doing so (if and when he's ready and willing)

While at it, Simon St. Laurent's effort with Ripper might fit in as well.