One of the data points in OPML tradition is the great attribute vs element debate, which extends beyond into XML itself. This is a spot where OPML catches a lot of grief, and I've sided with the griefers before. But, having played and thought a bit more, I think I get the point now. It's all about compromises, really, and considering both the producers and consumers of a format. And, since there's an XOXO and an XSLT (among other things) you always have an escape hatch from OPML if it makes you uncomfortable. There's a rough edge or two I'd like to see smoothed out on OPML to help ensure things fit through the escape hatch, but I think I've made my peace with things. ";->"

A great comment from Dean Edwards on my AJAX homepages mini-rant: "AJAX homepages are like those useless gadgets you can buy: an umbrella with a torch in the handle, a kettle that is also a radio, etc."

To be honest, I've yet to grok Instant Outlining as a collaboration tool. Something about the lack of automated change tracking to show what's new, on an outline-node level, when the buddy list pings. I tried it back in 2002 or so, started up again half-heartedly now, but ended up building a .plan. I don't think that this is quite what Dave had in mind for IO, but it's what most makes sense to me right now.

This Penny Arcade doughnut incident has made me hungry.

Why XOXO?

• Donovan says: "gotta learn more about gettin' my xoxo on. i've read your page, les but still i have no idea what xoxo is or why i'd want to convert my opml files to it."

• Good question, with a few different answers. And these are, of course, all my humble opinon and not any sort of official decree from the Gods of XOXO.

• First answer I have is this: Don't convert your OPML files, because ideally a pair of XOXO-colored glasses should help get your OPML seen as XOXO, should the need ever arise. The reason you shouldn't convert is because you've got an OPML editor that's helping you be productive, and as far as I know there's no comparable tool for XOXO. Here, I think the utility of the tool trumps the format - especially when the format can be massaged by machines down the line, and not by your valuable brain. Also, someday soon, I predict that someone who likes XOXO will write a tool for the OPML editor that can export XOXO.

• Second answer is this: XOXO is a format based on XHTML. XHTML is a format with which lots of other people have been productive, and it's directly viewable and CSS-styleable in web browsers without much fuss. This is the idea behind microformats: a Sudoku puzzle solution to the problem of publishing data that's both human and machine readable on the web at large. This is coming at the problem from the ground up - formats and data first, tools later, for the sake of imbuing the format with as much clear expressiveness as possible to enable future pathways.

• I think right now, XOXO is more useful for people consuming the data than for people producing it. And, OPML is easier to produce than consume in some cases (but not all).

• Where I think XOXO will really shine is in embedding more complex items of microformat data, something that I don't think will ever be a core feature of OPML. And that's okay - OPML will be good enough for capturing a lot of human brain output and it can be munged into XOXO whenever we need it as such. So, I'm happy to rest in the center and not become a cheerleader for either/or.

I don't know what I'm doing, posting to the OPML 2.0 spec review mailing list. I've stayed out of all spec and format discussions on general principle so far, avoiding the Atom process like the plague. Life's too short for sausage making, I believe in general. But, maybe this OPML thing will be different.

Is it just me, or does the new ActionScript 3 spec look like a bit of a trainwreck? It could just be me, because I've not yet taken the plunge from AS1 into AS2 - and I've yet to see a need to do so. From glances at AS2, I've gleaned that there are attempts to weld static typing and class definitions into an otherwise nicely capable and wildly dynamic scripting language (eg. JavaScript). And, as I think we're seeing with AJAX and Web 2.0 madness, JavaScript and its prototype-based style of programming has a lot of expressiveness yet to be fully explored. But, no, it looks like ActionScript yearns to take the Script out of JavaScript and maybe become ActionJava?

Sony Online Entertainment sends Penny Arcade 1200 Krispy Kreme donuts in response to their latest comic. Now, that's some funny shit. "Well played SOE...well played indeed."

It might be just me, but I don't get the buzz about AJAX homepages. I mean, sure, they're AJAX and sparkly and draggy and all that - but they're still portals, aren't they? I never wanted portals to be more dynamic, or even open to third-party-authored widgets - I wanted them to go away altogether. My feed reading is too expansive to fit inside little headline widgets, my weather report sits in my menu bar, and for anything else that does fit into a little widget I've got Dashboard. This just seems like another thing that's neat like digital watches, with a horrible user interface for doing anything actually useful.

Morning, everyone. I don't like this game - when I look at my inventory, it says: "You have: No coffee." Somehow we ran out, and now I'm considering stopping by a Starbucks for something extra-shot'ed and stupor-clearing.