RE: Minutes of the tele-conference CIQ meeting held on 15 January 2001

[Vincent Buller <vincent@and.nl>]

Ram, all,

>
> Vincent, I have enclosed the updated NAML version.

Thanks! I will have a look at it soon. WOuld you care to also send the DTD
itself?


> we agreed during the meeting that for some of the terms used
> in the Global
> address
> spec., we are not sure what it means as the document was not
> complete and we
> agreed
> that we will have more examples from your people who created the dtd.

I have sent you guys updated documentation of our GlobalAddress
specification, and I welcome any comments. My worries though are related to
timelines and creating unnecessary interdependencies. Inserting the
GlobalAddress spec into NAML does NOT depend on getting more sample data -
we work with the spec all the time, and the level of documentation is
similar to what I've seen for NAML. If we agree to that we're fine.

What I said in December is that I acknowledge the need for people unknown to
the Address field of work, unlike you and me, to have real-world sample data
to increase the level of trust they have in the structure's ability to
express real-world data. All I'm saying now is that getting these samples
may take some time, and we should not let our progress depend on that.


> > What we talked about at our kickoff meeting is to make an extensible
> > multilevel definition. For example, for Address this would
> > mean that one
> > should be able to markup a street as a whole
> (street&premise:"prinses
> > julianastraat 26") or in parts (street:"prinses julianastraat",
> > premise:"26"). This is a next step that we can start discussing now.

Sorry guys, I may have misled you. Looking at my notes from the kickoff
meeting we did not even plan to do this for V0.5! Quote:
v0.5
- First integrate MSI and AND work, start 1st week of Jan 2001
- Release first draft xNAL and xCIL (DTD and documentation), 31st March 2001
- Public comments by 30th April
- Review and integration 15th May, release 0.5

v1.0
- Multilayer address bit
<================================================================
- Extensible customer information
- Preliminary date: Public review August 15th 2001
- Comments by Sept 30th
- Release Oct 31st

v1.1
- Internationalization, de-westernization




> > I am currently converting the GlobalAddress DTD to the latest
> > W3C schema
> > definition; this allows us to be more expressive as we discussed in
> > December. Ram, did you look at XML Schema and can you have a go at
> > converting your specification?
> >
>
> I think we said that we will have an XML Schema for XNAL and
> XCIL in version
> 2.0
> and not in this version. We have to be very careful when
> defining the Schema
> for
> names and addresses and in fact other customer data as the size of the
> fields
> varies from country to country. For example, street names and
> person names
> are extremely large in Asian countries like India, Srilanka, etc. For
> example,
> my full name is: PARUVACHI VENKATACHALAM RAMKUMAR

Do you mind if we just keep calling you Ram? ;-)

I could not find any reference to using schema in my notes, but I remember
is that you would have a look at Schema to see if you could easily switch.
If so, we'd use it, otherwise not.

By the way, using XML Schema does not enforce you to use all the expressive
power; in other words, your samples of having to define field lengths is not
a requirement for using schemas. It is possible to use XML Schema as the
language, yet define nothing more or less than with a DTD.

I would still like to use schema for this version; it's not that hard.
However, if you still feel it takes too much time or it increases
uncertainty too much we could save it for the next version (1.0 in our
December numbering).

An introduction to Schema which helped me to get started:
http://www.xml.com/pub/a/2000/11/29/schemas/part1.html




> > Ram, did you hear anything from the DocBook people on the
> > e-mail you sent
> > them?
>
> I sent a couple of e-mails to Norman Walsh of the DocBook and
> have not heard anything so far.

Well, it's in their interest too. I don't think we have a lot of time to
spare :-)


> BTW, Bad Husick of Vignette, who was leading the CPEXchange
> group has left the group as he resigned from Vignette.

That doesn't simplify things either.... I would have loved to hear from them
what has happened to CPExchange's privacy policy being questioned.... Does
Karl know?

Cheers,
Vincent