On 16 Aug 2005, at 02:19, Mark Nottingham wrote:
On 15/08/2005, at 5:09 PM, Robert Sayre wrote:
I'm interested in getting things working,
not playing the syndication format advocacy game.
Not sure how feed-history will work without a unique id.
? Sorry, you lost me there.
Yes. If you were to take a subscription feed "feed.atom" and map it
to a graph
you might get something like this:
_feed
|----id------> <tag:myblog,2005/feed>
|---entry----> _
| |----id-----> <tag:myblog,2005/e1>
| |---title---> "first entry"
|--fh:next---> <http://my.web/feed/history1.atom>
The graph of the history1.atom archive may be:
_feed2
|----id------> <tag:myblog,2005/feed>
|---entry----> _
|----id-----> <tag:myblog,2005/e2>
|---title---> "another entry"
The role of id is to help us merge the two graphs. IE the point of
the history
tag is to help us think of the history.atom document as a
continuation of
the same document, or rather that they are both descriptions of the
feed with
id <tag:myfeed,2005/feed>. Merging the above two as simple graphs at
most gives
this
<tag:myblog,2005/feed>
|--1/id---> _feed
| |---entry----> _
| | |----id-----> <tag:myblog,2005/e1>
| | |---title---> "first entry"
| |--fh:next---> <http://my.web/feed/history1.atom>
|--1/id--->_feed2
|---entry----> _
| |----id-----> <tag:myblog,2005/e2>
| |---title---> "another entry"
|--fh:next---> <http://my.web/feed/history1.atom>
But what we really want is something like this:
<tag:myblog,2005/feed>
|---entry----> _
| |----id-----> <tag:myblog,2005/e1>
| |---title---> "first entry"
|---entry----> _
|----id-----> <tag:myblog,2005/e2>
|---title---> "another entry"
I think we require some form of inferencing beyond what is available
in OWL
to achieve this result. This could probably be achieved through some
simple
N3 rules.
Henry Story
