Minutes: XRI TC Telecon 8-9AM PT Thursday 2008-10-16

["Drummond Reed" <drummond.reed@cordance.net>]

[Note: apologies for the lateness of these minutes - I had a major deadline
this week for which I had to work all last weekend. I am now attending to
the action items in these minutes - look for an update to the
XRI-as-relative-URI wiki page and the first W3C TAG email later tonight.
=Drummond]

Following are the minutes of the unofficial telecon of the XRI TC at:

Date:  Thursday, 16 October 2008 USA
Time:  8:00AM - 9:00AM Pacific Time

ATTENDING

Nika Jones
Nick Nicholas
Markus Sabadello
John Bradley
Drummond Reed
Les Chasen
Peter Davis

REGRETS

Marty Schleiff 


AGENDA

1) REVIEW OF XRI-AS-RELATIVE-URI PROPOSAL

We had excellent feedback on the list of the first draft XRI-as-relative-URI
proposal:

	http://wiki.oasis-open.org/xri/XriAsRelativeUri

Our main topic for the call was to review the following list of feedback
items and agree on any updates to the wiki page before we begin posting
emails to the W3C TAG mailing list. (Notes and action items on each feedback
item follow the bullet point.)

* Should we list examples of bindings to other protocols? (Nika)

Consensus that we should do this - list at least http: and https:

We should probably also mention ftp. We may be able to add security even if
we don't use a secure protocol.

# NIKA to check into ftp, sftp, and ftps schemes.

* Should we mention the requirement for applications to be able to
recognize/parse different bindings? (John)

Yes. 

* Should we provide more detail about how the bindings work and how they
will scale, e.g., how the algorithm for example.com.xri.net will resolve?
(Les, John)

Yes. We should explain how the proxy IP address resolution would work for
xri.net.

* Should we talk about .xri as a possible proxy root?

Nick feels that it would only open us to other criticisms. It was agreed we
should remove this from the examples for now.

* Are there any issues with running authority servers or proxy servers at
URIs which do NOT conform to the binding? (Les)

John felt there is no issue except that XRI-aware clients will not parse
them as XRIs. The issue of a handful of addresses under xri.net that are not
currently proxy resolvers (equals.xri.net, at.xri.net) should not be too
hard to work around.

* Should we define any parsing/recognition rules for pure XRIs that do not
use any binding? (Les, Drummond)

Most likely, but not an issue we need to tackle yet.

* Terminology:

John suggested the terms "protocol-dependent" and "protocol-independent"
XRIs for describing those that are not bound to a protocol and those that
are bound.

* Should we still define an "xri:" binding if no other binding is
appropriate? (Les)

There was consensus that this is a question we can defer.

* Is version compatability necessary in the bindings? (Nika)

Drummond hopes that it will not be, but it remains to be seen.

* Should version compatability use a version parameter? (Bill)

Probably. We need to see how the resolution protocol evolves in the new
specs.

* Should we be clear that XRI resolution is not limited to the http(s)
bindings? (Bill)

We have always stressed that this is only one resolution option. We haven't
yet discussed mixing resolution protocols. The XDI TC has also discussed the
option of using an XDI document as the document descriptor, but that's only
as a media type alternative, not an entirely new resolution protocol.

* Is identifier delegation adequately described? (Drummond)

Not yet. Questions from Stuart Williams to Drummond and John made this
clear, so this needs to be added to the current wiki page.

********
We agreed the next steps are:

# DRUMMOND AND JOHN to update the wiki page, then send the first messages to
the TAG list.

# ALL TC MEMBERS to join in the discussion and help answer questions in the
"socialization" of the proposal.