if ($self->is_dork) {

Again, finding it difficult to switch into writing mode due to the fact that work is going well, and part of me wants to keep doing job-stuff. I have started to pull together my first application for the MTL, you see, and I'm very impatient to get it into a do-somethingable state. I allowed myself to make one discovery, after coming home, as to why something didn't work, and why something else worked very slowly. The nice feeling of having my work cut out for me will have to tide me over until tomorrow morning.

FWIW, the cause of slowness, I discovered to my chagrin, was Mason. One page-component I'm making involves an HTML-generating loop that iterates several hundred times per page (and unfortunately, it's not a results-set that can be broken across several pages, or anything like that). My first solution is somewhat elegant and makes use of Mason's object-oriented-ish features... and it's butt-slow. I found that if I get experimentally sloppy and paste all the code into one big ugly blob, the page builds much faster. Tsk

Fortunately, I think this presents a capital opportunity to learn about Mason's cache API, which I've yet to really use. These particular pages, each of which represents a plate full of tasty soup from ICCB's chemical-tools libraries, are the perfect candidates for caching. So, that's what I'll do tomorrow.

So full of code-foo was I today that I wrote a little patch for Andy's DBIx::Abstract module, which, though it be just a few lines, makes me feel good. I followed this up with a call to Andy, at the Arcus offices. Jim answered: "Ah, Jmac! We were just disparaging your code." It seems I, mm, didn't quite remove all the silly debug crap from my last Arcus assignment before I turned it in. Andy spent 30 minutes (of vice-president time) hacking it all out, after the site went live. mumble oh well...