XForms has explicit data models, in HTML the model is implicit. XForms can have type definitions, repeating blocks, event processing without scripting. HTML5 has similar widgets, they share XML, dom, events mechanism.
HTML5 can be XForms host language, has the widgets. Has the Dom interface and APIs. Need schema validation, event processing. Dom events, no problem. XML events, no. XForms action elements, of course not. Repetitions, sorry.
The XForms.js processor. JS objects for every XForms element. No schema support, no XPath extensions, no synchronous submission, no local file access.
Firstly, it occurs to me that my notes of the sessions aren’t worth beans, so if anyone actually wants to know what they’re really about, you might want to consult the actual proceedings. They’re in preliminary form while the conference is going on, but the presenters will make their final revisions based on responses from the audience and add any supplementary content, such as the slides, shortly thereafter.
So: tonight’s dinner. Pizza, of course, since computer geeks and pizza is the cliche that keeps on giving. At first I thought I would be keeping to myself as usual, but Michael Sokolov from Safari Books sat at my table and we fell into a nice conversation. The intersection of tech and publishing is my comfort zone, so we just naturally had a lot to talk about. Then Simon St. Laurent, whose articles I’ve been reading for at least the past dozen years and who had the most delightfully subversive presentation today, joined us as well, and it finally felt like a party to me.
Learned a lot, had some laughs, finally excused myself to put in some gym time before bed. This feels like something worth getting used to.
AJAXForms developed into XSLTForms 1.0. Added native xslt engine, json, csv converters. DOM level 4 sounds like a bad idea. Proposes a sort of attributes axis and schema type definitions within a document, a lot else besides.
RSDL, Restful Services Description Language. For specifying RESTful services. Documentation: There’s a home document, resources, media types, link relations. Service metadata. “You can’t apply an HTTP method to a representation. You can’t parse a resource.” Goes into detail about how resources are described.
My god, that was the furthest thing over my head I think I’ve ever heard. He slammed ultra-complex concepts at super-speed without mercy. I should have expected this, based on the forum messages I’ve seen of his. He proposes XPath as a common language shared for arbitrary future XML processing languages.
Simon St. Laurent
Savagery, changefulness. Only programmers read Christopher Alexander! Markup as architecture. XSLT as data pachinko.
This was an amazing poetic plea for a more humanistic approach to markup. I have no idea how to apply it. After a stunned pause, there were many passionate responses.
Students had assignment of building a better XML. The markup language is nicely streamlined. The schema language seemed very sophisticated. The scripting language…god, I’m going to have to read the paper. This gives a good overview of general markup theory.
Uses DocBook5 XML for documenting LaTeX. XSLT2 for distribution files, which can accomplish look-ahead while LaTeX cannot. Remarks on his (apparently remarkable) “tag abuse”. Very funny presenter, in that gruff, pragmatic pedantic Brit sort of way.
Processing XML can be accelerated with multicore techniques, data parallelism, SIMD, SSE2, AVX2. They investigated Xerces parsing with their Parabix tools. If the different phases of processing can be performed in different cores, great gains can be made.
Just gave an example of using parallel bit logic as a basis for parsing. People are mesmerized. He does this several times.
That was amazing. While the topic itself was as dry and mathematical as anything I’ve ever seen here, the energy was intense. The presenter wasn’t flashy in any way, but he really had a way of working the room. I might have also been picking up on the enthusiasm of the audience, who were all as attentive as a roomful of cats when a can of tuna fish starts getting opened. Really remarkable.
B. Tommy Usdin
Originally thought that “semantics” meant the study of the meaning of words. (Isn’t it?) “If you understand how to manage it, it’s syntax. If there’s still magic and mystery to it, it’s semantic.” “There’s only one appropriate syntax for semantics now?!?” She submits that disputes about the syntax of semantics should be moot.