* for the record, my steampunk goggles were bought off of Etsy.

Much of an object’s strength lies in its behavior, not its data. If you only try to provide the minimum, you end up with multiple clients duplicating code for common cases. In cases like flatten you end up with a bunch of people writing their own recursive functions. It’s not hard, but why should they bother when it’s not that rare a case?

Harold thinks the class is too large and needs refactoring.

There’s simply no reason for 78 methods in a basic List [Array] class. In fact, there’s no reason for 78 public methods in any class. 78 public methods in one class is a code smell. 78 public methods make a class hard to learn, hard to use, hard to test, and hard to maintain. When a class has 78 public methods, it’s time to refactor.

I think Harold’s right when it comes to designing one’s own classes: make them small, focused and testable.

However, if a language has core classes like Array, I think there’s a reason to build in convenience methods so that users don’t have write them from scratch. Projects like Ruby have resources for writing optimized implementations of those methods, and testing them.

This summer I ended up writing about a dozen additional methods for the PHP libxml2 object. I can see other projects needing the same methods, and it would had been nice to have them baked into PHP.

Elliot Rusty Harold and Simon St Laurent have been blogging the conference in full. Here’s the presentations that grabbed me so far.

• ERH showed us a tool, written in Java, for ofuscating an instance of XML so you can send it to someone as a test case without revealing confidential information. This was a hit, with several people suggesting additions. He put a sheet down in the coffee break area where we could write suggestions.
• Examplotron, Eric van der Vlist’s tool for generating Relax NG schemas from annotated references, influenced a couple of presentations:
  • Angelo Di Iorio described design patterns for document structures. They motivated their argument with an overspecified model of an address that didn’t fit ‘real word’ instances.
  • Ken Holman showed us a second generation of an XSLT 1.0 stylesheet that produces a stylesheet from an annotated example of an instance document.
• A couple of presentations on Topic Maps and RDF that went right over my head. Both were about trying to reconcile one to the other.
• Matthijs Breebaart from the Dutch Tax Authority showed off a system his group built to describe bits of law and regulation (i.e. the chapter and paragraph) in XML. From there you transformed it into a URI. His department was able to get the commerical publishers (who produce the online versions of law and regulation) to sit down and agree on a common URI format. The benefit of the exercise: his customers have a common way to get to the online documents and commentaries without needing to dive down into each vendor’s website.
• Ann Wrightson taught us some Situational Semantics, that is, under what conditions do human or machine interpretations of observations make sense. She left a challenge: one of Shakespere’s sonnets translated into Klingon, with markup in Tengwar. Under what situation does this document make sense?
• C. M. Sperberg McQueen gave a math lecture. Brzozowski derivatives, an algebra of regular expressions, give you nice way to generate validators without building state machines.

Tomorrow, we get the presentations on overlaps. This is one of the “elephants in the living room” that Tommie Usdin, one of the co-chairs, warned us about in the introduction. The spectre of Ted Nelson will haunt Extreme 05.

My experience so far confirmed my theory that Extreme is the Potlatch of XML conferences. Small, a little clubby (but self concious of this, and people do take the time to say hello,) presentations above the ‘beginner’ level, and one main track of programming.

And the food in Montreal? C’est bon. C’est bon.

Note: 2 – 4 require Java.

From that list, mod_murka looks neat. It looks for a cached HTML version of the request URI, and if it’s not found, looks for the .xml file, transforms it using whatever stylesheet’s mentioned in the file’s processing instruction, and caches the result.

There’s a standard text format for representing games, along with commentary, and multiple branches of play. The applets in The Interactive Way To Go use them, and you can find more game files at the KGS Beginner’s Room. If you download a tool like Goban, you can load them and walk through commented games and tutorials.

While I was in the desert last weekend (okay, staying in an air conditioned condo doesn’t really count as being in the desert, but it was 110 in the shade outside) I got to sit down to a few games with an old friend from college. He told me about the Kisedo server in the hopes that I’ll play more Go.