[Geopriv] Draft GEOPRIV minutes from IETF 75

[Richard Barnes <[e-mail]>]

Draft minutes for the GEOPRIV meeting at IETF 74 are below.  Please send 
comments to the list no later than Friday, 7 Aug 2009.
--Richard


----------
Minutes - GEOPRIV - IETF75

Summary (prepared by Richard Barnes):

1. Agenda Bash
Brian Rosen requested 10 minutes at the end of the meeting to discuss 
his drafts on extensions to the PIDF-LO civic address elements.  James 
Polk volunteered 10 minutes of his time for dhcp-lbyr-uri-option to 
extend discussion of geopriv-arch.

2. Geolocation URI
    draft-ietf-geopriv-geo-uri
Alex Mayrhofer presented a brief update on the WG draft describing a URI 
scheme for geolocation.  The current version adds a CRS parameter, and 
the next will address comments from the URI-Review mailing list.

3. Location Filters
    draft-ietf-geopriv-loc-filters
Brian Rosen presented a brief update on the WG draft describing a filter 
language for location updates.  The current draft is a significant 
update from prior versions, basing the filters on the general RFC 4661 
filter syntax.


4. GEOPRIV Architecture
    draft-ietf-geopriv-arch
Alissa Cooper presented an update on the WG draft describing an overall 
privacy architecture for GEOPRIV.  The developemnt of the current 
version was focused on refining terminology, in particular the meaning 
of the term "LIS"; discussion of that topic continued in the meeting, 
with no clear resolution.

5. Location URIs in DHCP
    draft-ietf-geopriv-dhcp-lbyr-uri-option
James Polk presented an update on the WG draft describing a mechanism 
for carrying location URIs in DHCP.  Hannes Tschofenig submitted an 
extensive review of the current version of the document, and James is 
still working through these comments.  James agreed to send a summary of 
the open issues in the draft to the list.  Several participants said 
that the current prohibition against the use of HTTP URIs should be 
modified to permit at least some classes of HTTP URIs.


6. Updates to DHCP Geodetic Location (RFC 3825bis)
    draft-ietf-geopriv-rfc3825bis
Bernard Aboba presented an update on the WG draft that makes a series of 
udpates to address errors and unclear points in RFC 3825.  Individual 
changes are being tracked using the issue tracker on tools.ietf.org, and 
most are awaiting text from their assigned authors.

7. IEEE Liaison
Dorothy Stanley, chair of IEEE 802.11 TGv, presented a liaison statement 
from 802.11 to GEOPRIV requesting that GEOPRIV develop a binary encoding 
for the GML shapes that are available in XML, mainly for use in interior 
location scenarios.  Some participants addressed doubts as to the 
utility of such a translation, but others supported working on this 
topic.  Discussion will continue on how to respond to this request.

8. HELD Extensions
    draft-winterbottom-geopriv-held-deref
    draft-winterbottom-geopriv-held-identity-extensions
    draft-thomson-geopriv-res-gw-lis-discovery
    draft-thomson-geopriv-held-measurements
Martin Thomson led a discussion on a series of proposed HELD extensions. 
  He gave a brief description of each document, with some group 
discussion after each description.  Privacy concerns continue to be a 
significant concern for the HELD Identity extension, and there is 
continuing debate over the need for special mechanisms for residential 
gateways.  Shows of hands indicated varying degrees of support for these 
drafts, but rough consensus to work on all four.  Discussion on how to 
sequence these drafts will continue.

9. PIDF-LO Civic Address Extensions
    draft-rosen-geopriv-pidf-interior
    draft-rosen-geopriv-prefix
Brian Rosen introduced two drafts that extend the PIDF-LO civic address 
structrure to include (1) "prefix" elements that match current "suffix" 
elements, and (2) a generic element "INT" to represent interior location 
elements.  The major issue with the INT element right now is whether to 
register values for it: Some view registration as necessary to avoid 
ambiguity, while others note that the lack of standards for building 
models could cause a lot of noise in the registry.  A show of hands 
indicated strong support for working on the -prefix draft; discussion 
will continue on the -pidf-interior draft.



Raw notes from Marc Linsner follow:

------------------------------------------------------------

Geopriv Notes

Agenda Bash: brian wants to discuss INT
     James wants lbyr longer

status update: charter updated; W3C GeoLocation last call deadline this
Friday;

Lightning Round:

Alex - GeoURI - discussion around CRS - consensus seems to be that WGS84
should be default, but don't preclude other CRS, solution in the current
draft.  Is a URI parameter registry needed? What about privacy policies?

Brian - loc-filters - 05 released this morning; now based on RFC4661;
several changes - read the slides/text!  Issue: Normal Reference to -dynamic
which is experimental

Alissa - Geo-arch: not trying to dramatically change from existing Geopriv
work or implementations. Describe what a 'LIS' means; James - the current
definition of LIS seems to fit the HELD arch but no necessarily the DHCP
architecture.  Brian - there is current usage of LIS includes dereference.
Marc - the term LIS is still muddy. Jon - What's wrong with the current
text? - Brian - like I just said, 'own location' does not cover
dereferencing.  Ray - ???

James Polk - lbyr uri - chose the auth. security model; rewrote the intro;
addressed Ted's concern; Hannes is addressing things from the 00 draft 2+
years ago.  Some of Hannes' comments were good.  Keith: this is a wg draft?
You are not the arbitrator of the text, the wg is.  Hannes is commenting on
the jabber...(read the jabber).  Jon: I'm lazy, I read the doc for the first
time today.  I am curious why only SIP, SIPS, PRES uri?  Why not HTTP uris.
James: Jon you agreed to this earlier.  Ted: you need structure around the
URIs, hence this restriction.  We need to work on this.  Jon, we need HTTP
uri support. Ted: if we had support for a HELD uri would that satisfy your
concern.  Brian: I want to support HELD uris  James: this needs to be run by
Lisa. Ted: I'll take the task to run this by Lisa.  Cullen: I agree Ted,
we'll work this out.

Bernard - rfc3825bis - started with 3825; we have an issue tracker; we will
make changes based on list discussion and consensus.  Issue 9 & 10 closed;
Issue 1 resolved; Issue 2 needs list discussion; Issue 5 has been sent to
the list; Issue .... (read slides) Keith: does the assignee have more
authority?  Bernard, all text will be discussed on the list.  Martin: I have
proposals...I'm don't have motivation.  Richard: please copy/cut paste from
your other draft.

Dorothy Stanley - IEEE liaison - chair 802.11tgv and liaison to IETF -
summarize the letter sent last week.  (read slides)  covered background of
802.11 location work.  IEEE is requesting the IETF to extend the BINARY
representation of the location objects to include shapes, etc.  Cullen -
verify the dates....wg doc. Brian questioned the usefulness. Martin
supported Brian.  Marc - IEEE is asking for xml to tlv mapping, not critque
of their application.  Ted - decide to do the work, then have the
application discussion Hannes - 3GPP has already done this work. Dorothy -
we chose to come to IETF first.  Gabor - Nobody has this defined, not in
3GPP

Martin Thomson - HELD extensions - 4 drafts - deref, identity extensions,
res-gw-lis-discovery, held measurements.  Derefernce - do people think this
is useful/necessary?  Identity - (read slides) - Marc:   Brian: I don't see
anything in the doc about a 3rd party using IP address to ask for someone's
location. Cullen: I believe we agreed to not do the 'authorized third
parties'.  (missed some)  Ted: I agree with Marc...you are changing the
rules around LCP.  The draft needs to explain why/how we are changing the
LCP rules before becoming a wg draft.  Martin: the doc talks about the need
for authorization.  Jon: We need to solve this problem and need to
prioritize this as the first problem.  Bernard: Maybe break off the third
party issue and deal with it separately.  Lis Discovery: a large group of
home gw devices don't support this. Cullen: it's too strong a statement,
some devices do support.  (Ray Bellis): this overloads option 15 and use of
it.   Cullen: if we have a solution that is supported on some of the
existing, we need to use it.  Ray Bellis: this draft will work with no work
in the home gateway.  Measurements: necessary for cooperative location
determination.  Brian: I think this is the least interesting of the 4  Marc:
please characterize measurements and identity extensions.  Ray Bellis: In
the UK, we need identity and res-gw-discovery

Richard: Any more comments on the prioritization??  Identity Extensions then
HomeGw LIS Discovery

Brian: Concerns over putting deref off for another year.

Jabber room: all 4 are equally important

HUMS:
Those in support of the group working on Deref:  (no hums)
Those no in support of the group working on Deref: (no hums)

Should the wg work on a deref for HELD: (little hums)
(never asked for the opposite)

Should geopriv address the problems of Identity?  (17 hands raised)
Against (1 hand raised)

Gateway discovery problem, in favor (9 hands raised)
Against? (1 hand raised)

Measurements, (8 hands raised)
Against? (3 against)

Brian - Additions to PIDF - prefix draft - 2 additions to handle prefix in
civic addressing.  Can we take this on as a wg item.  pidf-interior - we
can't support a lot of interior spaces.  This works in more cases.  James -
no registry leads to interoperability problems.  Ted - there is no standard
for interior spaces.  Richard: Aren't these drafts at odds with each other?

Hum:

Should we do prefix??  (14-15 hands)
Opposed?  (none)