Sandro Hawke wrote:
[cc'd www-archive, in case others get interested]
Great.
[You write about perhaps using the scribe tools OWL-WG is using]
Sure, I'd be happy to have you guys as the second users. (Eric just
grabbed me in the hall to talk about it.)
Is your wiki up and running?
Yes, it's at http://www.w3.org/2009/sparql/wiki/ though it's still sparse.
Where in web space do you want minutes to appear?
>
> For OWL, the chatlogs, which start as IRC logs and are then cleaned up
> are named like:
>
> http://www.w3.org/2007/OWL/wiki/Chatlog_2009-02-23
>
> and the formatted minutes are named like this:
>
> http://www.w3.org/2007/OWL/meeting/2009-02-23
We're pretty much duplicating the Web-space structure that OWL WG is
using, so I'd imagine chat logs at:
http://www.w3.org/2009/sparql/wiki/Chatlog_%yyyy-%mm-%dd
with formatted minutes at:
http://www.w3.org/2009/sparql/meeting/%yyyy-%mm-%dd
I was a bit confused by
http://www.w3.org/2007/OWL/wiki/Minutes_Review_Conventions which
suggests that the minutes are on the wiki and editable by any WG member
when, in fact, that's not the case?
You'll also need a wiki page like:
http://www.w3.org/2007/OWL/wiki/Participants2
which doesn't exactly match the other databases of participants because:
- it needs to list all the irc names used for people which
can not be derived by the algorithm
- it needs to list everyone who has attended any meeting,
including guests and part-participants
OK, I have an initial stab at this at:
http://www.w3.org/2009/sparql/wiki/Participants2
(One of the current bugs in the system is around the handling of the
participant list. The list really needs to have a transaction-log
style, so that we can tell who the participants (really "attendees")
were at any given point in time. If that list changes later, then a
name might BECOME ambiguous after-the-fact, which would make old minutes
harder to work with. So, for now you can use the Participants2 style,
and expect at some point we'll have to upgrade this to a transaction-log
style.)
OK.
Out first meeting's logs were at
http://www.w3.org/2009/02/24-sparql-irc.txt
http://www.w3.org/2009/02/24-sparql-minutes.html
What next? :)
Lee
BTW, the script has a clear object model of MeetingEvents which could be
dumped in RDF/XML easily. I was going to dump them all into virtuoso at
one point, but never quite got there. Of course I'm not sure it would
be socially good to have it easy to datamine the meeting records.
