Simon
Thanks for this. The COI jobs site is certainly helpful. It would also
helpful to take a page or two of TWFY Local pages and mark them up.
The Common Tag debate I think is part of why I've hesitated to use RDFa
up to now, and been drawn towards microformats. Basically, at times it
can tend (to a outsider) to like like the CSS3 issue -- heavyweight,
high on apparently arcane issues, and with no clear consensus.
Microformats on the other hand seemed lightweight, easy to understand
and implement, and formed by consensus. (Incidentally, when you look at
the source code for many local authority sites, you'll see the heads are
littered with DC tags and the like, but the bodied are badly-written,
ill-formed HTML with not even basic semantic markup and plenty of font
tags and the like).
However, there are clearly limitations for microformats, and one of the
core goals for TheyWorkForYouLocal is to present Local Authority in a
form that can be reused. Semantically marking it up is therefore key --
they only question is by what method(s).
Thanks again for your help. Much appreciated
C
Simon Gibbs wrote:
CountCulture wrote:
Simon
Sorry -- was rushing out and couldn't think of the name. Yahoo's
Common Tag is what I was talking about:
http://www.ysearchblog.com/2009/06/11/new-common-tag-format/
Ah, OK. Well that depends on what you want to do, and I realise I made
some assumptions about that which suddenly don't seem like great
assumptions, but I'll share them to be clear.
To state the obvious, tags describe a particular act - subjectively
assessing the topic of something and making it more - but not totally
- objective. Meeting minutes are the obvious use case for this
application and certainly CommonTag would be richer* than relTag. You
could create an AutoTag for every page and it wouldn't hurt, but you
have loads of objective data hanging there like succulent fruit so
tags don't strike me as sexy.
There is an example you might want to look at, but I'll explain the
relevant features anyway. Its the COI central jobs site, which was
intended to be a big store of all public sector jobs. (Exactly the
same thing is/was done for consultations by the way).
http://webbackplane.com/mark-birbeck/blog/2009/04/23/more-rdfa-goodness-from-uk-government-web-sites
he gave a presentation on this as well:
http://webbackplane.com/mark-birbeck/blog/2009/06/slides-for-semtech2009-talk-on-rdfa
In this case, the page about the job became one with the API. Doing a
GET on the job page returned a representation that was both machine
and human readable. It must have been very cheap because it was just a
template tweak to complete the integration. Getting the jobs on the
COI site was a case of doing GETs on ASPX, PHP and other kinds of page
that already existed, all the same back end technology and human
processes were retained.
This was actually the kind of thing that I originally volunteered to
help with, as it makes sense to do it at the Government side and it
made sense to me that MySociety would be getting calls about doing it.
Regardless, your pages are clearly data driven, come from a template,
and are full of useful facts so when I saw them, I just assumed you
wanted to expose those facts. You know, for people to experiment with.
This is very straightforward given that people, organisations,
documents, pages about people, authorship and memberships are in FOAF
and this is the data you have.
If this is old news, or otherwise makes sense, then say so, but if you
like I can take a copy of a set of pages and mark them up in the way I
had assumed so you can see what I'm on about.
Simon
* it's worth seeing Tantek Celik's objections to CommonTag:
http://www.ysearchblog.com/2009/06/11/new-common-tag-format/#comment-20244
_______________________________________________
Mailing list [email protected]
Archive, settings, or unsubscribe:
https://secure.mysociety.org/admin/lists/mailman/listinfo/developers-public
