W3C home > Mailing lists > Public > www-archive@w3.org > December 2001

PRIMER: Conference IRC Log

From: Aaron Swartz <me@aaronsw.com>
Date: Thu, 13 Dec 2001 12:55:20 -0600
To: Frank Manola <fmanola@mitre.org>, "Sean B. Palmer" <sean@mysterylights.com>, eric miller <em@w3.org>
CC: <www-archive@w3.org>
Message-ID: <B83E55B8.11E40%me@aaronsw.com>
Available on the Web from about:
http://blogspace.com/swhack/chatlogs/2001-12-13#T16-22-16
Until http://blogspace.com/swhack/chatlogs/2001-12-13#T18-47-49 .

Also, Sean and I hope to work up an outline today on the Wiki:

http://c2.com/cgi/wiki?RdfPrimerOutine

--
<AaronSw> +SeanPalmer
<AaronSw> +AaronSw
<AaronSw> we chould do: history, introduction
<AaronSw> serialization, namespaces, etc
<AaronSw> daml, ontology stuff
<AaronSw> ... m12n is necessary
<AaronSw> ... want something that's not too big to start and work thru
<AaronSw> and then move on
<AaronSw> sbp clarifies: three separate documents
<AaronSw> frank: what i imagined was to say what rdf _is_
<AaronSw> ... how do you model things? what are the general idea?
<AaronSw> ... then have some applications
<AaronSw> ... once you have that core, you can have modules on tools, who's
using it
<AaronSw> ... introduction to the model theory
<AaronSw> ... need a core explanation, sorta like what M&S tries to do
<AaronSw> ... don't want to mix some stuff up
*** xena has quit IRC (Excess Flood)
<AaronSw> em: intro to the model theory wouldn't be helpful here
<AaronSw> ... but would it be?
<AaronSw> ... some really get it, others say you're taking a complex thing
and making it completely inaccessible
<AaronSw> frank: primer doesn't stand alone
<AaronSw> eric: dependencies... what needs to be included?
<AaronSw> ... primer might say what they're there for
<AaronSw> frank: put into context
<AaronSw> em: wasn't initially going after that, but it does make sense
<AaronSw> ... sorta going after web architecture
<AaronSw> ... identifying things with uris, xml introduction, context on why
we structure things in this way and benefits
<AaronSw> sbp: yeah, that's aaron's stuff
<AaronSw> ... shows how rdf came up -- with xml and namespaces and uris
<AaronSw> ... and how data fits together
* em appologies for mushing your guys together
<AaronSw> no prob
<AaronSw> sbp: there are other things that come with it
<Morbus> # mushing?
<AaronSw> sbp: lots of philo discussions that get missed
<AaronSw> eric: intended audience?
<AaronSw> sbp: i think me a year ago would be good model
<AaronSw> ... i knew about xhtml, someone mentioned the semantic web on
xml-dev
<AaronSw> ... the only document was timbl's semweb roadmap, and some stuff
on xml.com
<AaronSw> .. i read it, but it wasn't enough
<AaronSw> ... i had to talk to people like aaron and so forth
<AaronSw> ... the breaking point was timbl's notation3 primer
<AaronSw> ... from there to m&s, rdfs, daml
<AaronSw> ... want a central point to link to all of these things
<AaronSw> frank: i'm targetting it towards a collection of people...
<AaronSw> ... i want to speak similar to sbp, but i'm not sean
<AaronSw> ... i come with a different background
<AaronSw> ... i found it was interesting when you get to talking about tech
like rdf
<AaronSw> ... it's the intersection of DB, Web, KR
<AaronSw> ... certain things are interesting based on your background
<AaronSw> ... i took the tact from the data model section
<AaronSw> ... people used to databases
<AaronSw> ... neat stuff is exciting issues about identifying things on the
web
<AaronSw> ... that's when things get interesting
<AaronSw> ... want to go to folks used to dealing with large data
<AaronSw> ... then you got to point out the URI business
<AaronSw> ... i'm trying to get the data folks on board
<AaronSw> aaron: same explanation doesn't work for all groups
<AaronSw> [missed stuff]
<AaronSw> frank: object dbs allow behavior in the db
<AaronSw> (for db people)
<AaronSw> for oo people it's persistence
<AaronSw> ... it's the same problem -- different ramifications for each
<AaronSw> aaron: i want to aim at app developers and get them to store data
<AaronSw> eric: me too -- i have an audience that aren't sold on this stuff
<AaronSw> ... we (this group) are sold
<AaronSw> ... but i want this aimed at the folks saying: why is this
important? what are the recipies?
<AaronSw> ... just tell me how do I do it?
<AaronSw> ... some things we need to clarify. serious issues with going to
do it
<AaronSw> sbp: use cases are good for that
<AaronSw> ... they show it works, and show how someone else did it
*** filsa: No such nick/channel
*** [filsa]: No such nick/channel
*** Morbus has left the channel
<AaronSw> aaron: have a cookbook like perl and python
<AaronSw> eric: for me it's a cookbook -- there are patterns that i see
<AaronSw> ... writing some of those recipies down
<AaronSw> ... would benefit so many people
<AaronSw> ... DC never had recipies
<AaronSw> ... and made some mistakes
<AaronSw> ... same mistakes being repeated
<AaronSw> aaron: it's a pattern language and we need to write it down
<AaronSw> frank: there are some data modelling idioms covered in M&S we need
to cover
<AaronSw> ... when do you use a container or repeated attributes?
<AaronSw> eric: is it is-a or has-a?
<AaronSw> ... a sorta follow the trail kind of thing
<AaronSw> ... controlled vocabularies
<AaronSw> ... i have seven things about this topic
<AaronSw> Python Cookbook: http://aspn.activestate.com/ASPN/Cookbook/Python
<AaronSw> ... need to tackle the 80/20 stuff
<AaronSw> ... help groups confusing author string and persons
<AaronSw> eric: 30 minutes left, how do we fill it in?
<AaronSw> frank: cookbook is arbitrarily extensible
<AaronSw> ... we get the basic stuff down in the primer, but we could have
another document
<AaronSw> ... don't want to get in the way
<AaronSw> eric: i view the rdf and rdf schema separation as a managerial
one, not a technical one
<AaronSw> ... we need to group logically, not by specs
* AaronSw agrees
<AaronSw> sbp: one of the best rdf skills is playing mix and match and
choosing the right terms
<AaronSw> aaron: have primer being core background and point towards
cookbook as evolving resouce
<AaronSw> ... perhaps ask aspn for help with this
<AaronSw> ... but not a w3c deliverable
<AaronSw> eric: good point, we keep learning more
<AaronSw> frank: recipies include not just idioms, but in fact whole schemas
<AaronSw> sbp: i looked at the daml ontology list and many have errors and
mistakes
<AaronSw> ... we can't throw them to the cookbook without teaching them
first
<AaronSw> ... serialization is the most annoying part
<AaronSw> eric: the people who get rdf the most, imo, are those who've never
seen xml
<AaronSw> frank: i'm looking at the rdf core wg page
<AaronSw> ... what the primer at least has to do is all of the explanation
that the m&S does
<AaronSw> because no one else will do that
<AaronSw> frank: does the primer need to cover rdf/xml?
<AaronSw> aaron: i want to make xml optional
<AaronSw> sbp: perhaps have a flowchart
<AaronSw> eric: in M&S we did some things for historical reasons we don't
need to do anymore
<AaronSw> ... no such thing as abbreviated syntax
<AaronSw> ... no id
*** GabeW (~gwachob@12.236.92.153) has joined the channel
*** AaronSw has changed the topic to "Taking notes from the RDF Primer
telecon,"
*** AaronSw has changed the topic to "Taking notes from the RDF Primer
telecon"
<AaronSw> eric: some things we can do to make it easier to swallow
<AaronSw> (it = rdf/xml
<AaronSw> ... we need some kind of this is what it looks like kind of thing
<AaronSw> eric: n-triples, then xml
<AaronSw> ... ntriples in the sense of: this is a simple sentence
<AaronSw> ... show the comparison between n-triples, rdf/xml, and arcs and
nodes
<AaronSw> ... we now have the tools to spit out all those formats plus svg
and png
<AaronSw> aaron: very cool to have a "other representations" web service
<AaronSw> aaron: i got the sense that we're looking at giving people the
background they need to dig in
<AaronSw> sbp: getting into the rdf mess
<AaronSw> my doc:
<AaronSw> rdf's place in webarch
<AaronSw> frank: i have a problem starting off with soemthing other than rdf
<AaronSw> eric: absolutely
<AaronSw> ... introduction should be all about rdf
<AaronSw> ... start off with an example
<AaronSw> ... but the next section should go into the context of the web
<AaronSw> ... we're building off of the web here
<AaronSw> ... different from my days in sgml
<AaronSw> ... building off of the web helps
<AaronSw> ... showing at goals
<AaronSw> frank: once you have a bit to talk about the data model
<AaronSw> ... natural hook to attach the explanation of uris
<AaronSw> ... and explain a resource
<AaronSw> aaron: have a narrative to hold things together
<AaronSw> sbp: URIs identifying things are a huge leap for people
<AaronSw> frank: you should re-read my data model thhing
<AaronSw> ... starts with URLs but shows useful relationships
<AaronSw> ... but then uses uris
<AaronSw> ... show predicates are uris
<AaronSw> [rum analogy]
<AaronSw> eric: frank's thing could be a whole document by itself
<AaronSw> frank: is this the right material? is it going to slow?
<AaronSw> frank's doc:
http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2001Oct/att-0236/01-prime
rmodel7.html
<AaronSw> eric: need some outline for tomorrow's telecon
<AaronSw> ... seems aaron and frank are saying similar things on a different
outline
<AaronSw> eric: more text than we know what to do wiht
<AaronSw> ... need a storyline/outline
<AaronSw> eric: can people commit time up until the 20th
<AaronSw> yes all around
*** GabeW has quit IRC (Ping timeout: 181 seconds)
<AaronSw> telecons seem helpful
<AaronSw> eric: take discussion to email until we get an outline
<AaronSw> ... see how far we get on that.
<AaronSw> ... once we feel comfortable, put some meat on the bone
<AaronSw> ... cc www-archive
<em> AaronSw, can you start the ball off by sending these excellent notes?
<AaronSw> sure
<AaronSw> will do
<AaronSw> and cc www-arhchive
<AaronSw> for primer: pointer to community

-- 
      "Aaron Swartz"      |              The Semantic Web
 <mailto:me@aaronsw.com>  |  <http://logicerror.com/semanticWeb-long>
<http://www.aaronsw.com/> |        i'm working to make it happen
Received on Thursday, 13 December 2001 13:57:23 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 7 November 2012 14:17:15 GMT