I also enjoyed presenting in English, knowing that there was an interpreter sitting up in a booth, translating what I said on the fly. I made an effort to speak more slowly and clearly, both for the interpreter and also for the Spaniards in the crowd who could speak English.
I must admit I had to giggle when I witnessed some of the browser heavyweights on a panel together to talk about interoperability. Charles from Opera, Doug from Microsoft, Mike from Mozilla, Allan from Konqueror and Arthur from Nokia. They each tumbled over each other to present a 2-5 minute thingy from each of their laptops, delivered in a variety of languages, from a variety of presentation programs. Time was wasted while we watched them get up and down, change plugs, open programs, and talk to their slides. Interestingly, Doug didn't even present.
Now I'm not saying that some standards aren't important, particular for web browsers, through which more and more data will be delivered on more and more hardware. It simply struck me as telling that these chaps couldn't get a blended bunch of slides together for their panel about interoperability.
What was quickly evident to me is that all of these panelists were coming from a perspective of business, where it isn't practical or beneficial to the bottom line to share and offer the free exchange of both tactics and ideas for the future. It just seemed to reinforce the pervasive difficulty of standards adoption: agreement. How can you have an edge on your competitor when you have to agree on everything, and in fact share implementation.
That's what I like to much about the emerging approach to "standards". Where, with things like Flickr, the system and how to interact with it is described in such an abstract way that freedom to push and pull the data is left to the developer coding on it. The developer has the freedom to choose her own language to write in, knowing that the 'spine' of the system is exposed by semantics-free methods.
That's a well-met design challenge: to describe a system without enforcing a language of interaction.
Sounds a little like the situationists...
Update, October 17, 2007
I had a brief email conversation about this post with my friend, Dallas. The chat made it abundantly clear that I have no idea what I'm talking about when it comes to the Semantic Web. The point I was trying to make is that Cal's (and Flickr's) approach to injecting a semantic pattern into the world, where people can make of it what they will and no correspondence will be entered into seems to contrast quite starkly with the standards movement. Even the open standards movement. The presence of some of the chaps in decision-making positions at these browser corps seemed to me to be indulging in largely academic discussions about the best way to come to agreement on things, rather than the approach of 'build it and they will come'. One chap told me grass roots only gets you so far before you need to form a committee to decide how to 'prune' the emergent standards to order. I guess that's true, but really, you're literally fighting about semantics to get agreement, and that's really hard.
As I read in the Skeptical Reactions section of the (current) Semantic Web page on Wikipedia, I am not the first to question the proposed amazingness of this idea. I can certainly see the benefit to intra-net data transfer, where the language can be invented and indeed enforced amongst a closed set of developers.
aaron falls in love with his new computer potato by elinar
That said, most of the guys on the Flickr team are total nerds, and have worked really hard to provide feeds, publishing stuff in RDF format, and XML to make it even easier to propagate Flickr data through the web. I've had lots of chats with Aaron in particular, who is a total RDF maniac who can blessedly understand my skepticism. I guess I'm really blowing a bunch of ignorant hot air, because I've never written an RDF document and had that data accessed and understood by an external 'agent'. But, you know. A girl can
I'm curious about what we might learn from the study of linguistics in this regard. Will it ever be the case where a self-published API becomes de rigueur? We've seen imitation and reflection in some of our co-conspirators' work, where their API design is in some cases identical to ours, so data transmission is smoother. I suppose this may evolve into a de facto standard at some point. Is that where the committee steps in? To manage change? Perhaps. But, perhaps language doesn't require change management. (Hence the interest in, and questions of linguistics.)
Computers love standards! People? Not so sure.