data standardization, microformats, not just all hype
this weekend at creative camp we were talking about various professional/social networking tools, and how it’d be handy if there was a single sign-on solution for all of them.
one of the other presenters said, and i quote(paraphrase),
“If you made one place that managed my account login as well as personal info, so i didnt have to retype it every time, for:myspace livejournal facebook monster linkedin bloglines etc etc etc, not only would I use it but you’d probably make a million dollars.”
so yeah. of the bunch he rattled off, the only one with no api support at all to my knowledge is myspace. single sign on/standardizatoin of information IS a good thing. and profitable. cuz happy(and free(and open)) developers are productive developers, right?
Messina has been jabbering on about microformats forever. I recently got around to asking, “what’s the big deal”? Essentially, the content doesnt change but the packaging does. Formats like RSS, while effective, are inefficient because they require an additional packaging process: the generation of the feed. Microformats take the built in id and class properties of HTML elements and use them for their underlying purpose…no, not to style and format. That’s a secondary use. The primary function of id and class is identify an “object” within the DOM. ID’s for single objects, classes for recurring objects. Microformats exploit these identifiers in such a way that a web document itself acts as the publishing feed…a parser can go through looking for a standardized format for information such as calendar and address book info. no secondary republishing. and, every instance of support means that another developer has to do one less thing in making his data scrape work.
data standardization is good.
im weaseling a little bit of hCalendar into the app i’m building at work. my hope is that, when phase 2 rolls around, and it comes time to hook in syndication services, i can say “well, its actually mostly done already”. thats the kind of stuff that makes my bosses happy.
18. September 2006 at 15:47
For single sign-on, check out OpenID (another technology I yammer on about)!
18. September 2006 at 16:32
theoretically, it wouldnt be impossible to use the openID open source code to tie together the services which offer an API that handles authentiction. even better, something that scales to ANY service that has an API for it’s authentication scheme.
hmmm…time to get reading.
18. September 2006 at 16:44
You could even start by OpenID enabling your blog…! ;)
So the real power is in hcard + openid + XFN.
Think on that for a spell.
18. September 2006 at 18:53
the glaring problem i see with the openid project is a lack of coherant documentation.
for example, the task of installing OpenID for blog commenting on wordpress 2.0.
going to the openidenabled.com website and searching for “wordpress” i get 7 results, only 2 of which are really relevant to wordpress at all. and one of those is simply a link to the other. So that one relevant link, which is http://www.openidenabled.com/Members/http-3A-2F-2Fedulix.videntity.org-2F/openid-plugin-for-wordpress/openid-full-tar.bz2/view
is mostly a list of comments complaining about the lack of documentation! (the zip for the plugin contains NO readme. not good)
Ok, so onto google. A quick google of wordpress turns up some more useful links, which are more detailed than the directions on the openid site itself. Essentially, the “beta” plugin has dependencies as documented a few comments down on http://photomatt.net/2006/04/14/openid-for-wp/
overall, a pretty tedious process compared to how easy most other WP plugins are.
the openid site is effectively expansive, but difficult to navigate. this is a huge hinderance to the effectiveness of the mission of the project. how is it that something dedicated to make one process easy exists in a site that is possibly one of the most confusing i’ve ever experienced?
OpenID libraries exist for half a dozen major development platforms, including php , java, and .net. But with all of the other authentication schemes that are practically turn key, what incentive is there for a developer to impliment something that requires tons of digging just to get a mediocre set of documentation? The concept is gold. unfortunately, it’s a the bottom of a pretty rapid riverbed right now. I’m not sure how to even start panning for the value.
19. September 2006 at 08:22
Dude!!! are you coming this year on the ski trip?
we are planning on going to alberta canada… to ski banff, lake louise, and kicking horse… although we are hoping to focus on kicking horse primarily.
zev is already working on our accomodations, and erik is planning on coming with his girlfriend.
dates are feb. 4th to the 11th 2007
here’s to hoping you can make it!