Thursday, July 10, 2008

Invites and Elsewhere

I've recently gotten a chance to sign up at a couple of online services that are apparently in 'invite-only' beta. But they have given me invites. So I thought I'd extend, to you, beloved readers, those invites. I've got a couple of invites over at Twine, and a few at dailymile. Twine is supposedly a 'semantic web' application. As far as I can tell it's a new place to store bookmarks (though you can also post comments, notes, blog), and keep them well organized. The more info you put into Twine, the better it knows you, and the better it can recommend new items. That's my understanding, anyway. Dailymile is for people who run, bike, or swim. Some sort of social network for such people, where you can keep some information about your training as well as upcoming events. So anyway, if you'd like to try either service, leave a comment below, with your email (or, if you don't want your email showing up in public comments, you'll have to find another way to contact me - or I suppose I could delete the comment afterwards). I'll update this post, at the top, if/when I run out of invites.

So, since I've been signing up at places online and trying things out, I thought it might be a good idea to make a list of those places. And since they're all reasonably public, I went ahead and posted them on the 'contact' part of my other webpage. The list includes: this blog, twitter, reader shared items, twine, dailymile, mapmyrun, goodreads, flickr, youtube, facebook, and friendfeed. Feel free to find me there.

In the process of adding those links, I thought I should try something else out. This is the XFN (XHTML Friends Network) microformat. For the links I made to my other pages, I added rel="me" to the (x)html. Supposedly that's helpful in some sort of semantic web sense. Reading a little, though, I found that rel="me" is supposed to be symmetric. So I'm supposed to now to go all of those other places and make pointers to all of the other places from them? Some of them don't have that sort of capability setup. So am I not to claim rel="me" when I link to them? I mean, I think I see the argument for rel="me" to by symmetric - if it's not, I can go around claiming any page I want.

Of course, if I'm only linking to my pages, it's not much of a 'Friends Network', huh?

I've also wondered about linking to pages that mention me before, maybe storing them all in some sort of 'mention file'. Not because I really care that if somebody finds me online, that they also find all of my other mentions. What I was thinking was if I linked to pages that mentioned me, bots might be able to distinguish between various instances of Nicholas Hamblet online (I think there is at least one other, in case you were curious). It seems like this could be helpful somehow. Like... I find an interesting mention of a 'A. Person' and wonder if it might be the 'A. Person' I went to high school with. So I go to some web services that has crawled 'mention files', and put in the page I found them on and the name (that page could, after all, mention many other people). If the web service has found a 'mention file' for that link and person, it'll let me know. Clearly this has issues: (1) several people could claim the same mention, either intentionally or not, (2) spam, as always, (3) more popular people get mentioned all the time, and keeping their 'mention file' up-to-date would be a hastle. All the same, it's something I've wondered about. I guess one way would be to set up a web service that crawls, finds names, and stores them. If it keeps track of context, it may be able to guess intelligently to distinguish individuals with the same name, and with the help of some 'mention files' and appropriate microformats, perhaps it could do ok? Perhaps there's already such a thing? Either the 'mention files' or the web service?

No comments: