Re: Charter question on form of API, was Re: DOM based API

Matt Womer <mdw@w3.org>, 2008-06-06 15:49 -0400:

[...]
> 3. We may explore exposing location information via markup or HTTP  
> headers

As far as exposing location info via markup, I believe that it
could, at some point, be worthwhile for this group to take a look
together at the existing mechanisms that are supported in
currently deployed browsers on GPS-enabled handsets for use in the
carrier networks here in Japan -- specifically, the DoComo @lcs
attribute mechanism, and the KDDI/Au device:location protocol
mechanism. They are essentially just different syntaxes for
achieving the same effect in the browser.

I recognize the glaring deficiencies inherent in those markup
mechanisms, and I personally have some misgivings about whether it
would be a good idea for anybody to see them as good precedents
and to propagate them elsewhere.

But that said, those mechanisms have been in use in Japan for 2 or
3 years or maybe more, in hundreds (if not thousands) of sites --
including those of some key, major content providers (e.g,,
yahoo.co.jp, goo.ne.jp). So I think they should eventually merit
some attention from this group, just due to the fact that it seems
likely that it's going to bring together a range of people and
organizations with experience and insight into the use cases and
requirements around location-aware browsing.

Finally, all *that* said, and as Matt has said too, I don't think
this group should be chartered or expected to produce a spec for
any standard markup mechanism for this (it should instead just
produce the DOM/scripting/browser API).

  --Mike

-- 
Michael(tm) Smith
http://people.w3.org/mike/
http://sideshowbarker.net/