[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: Minutes: XDI TC Telecon Friday 2015-03-13
Following are the minutes of the unofficial telecon of the XDI TC at:
Date: Friday, 13 March 2015 USA
Time: 09:00AM - 10:30AM Pacific Time (16:00-17:30 UTC)
Ning Zhang
Drummond Reed
Markus Sabadello
Joseph Boyle
Phil Windley
Peter Davis
Drummond uploaded a draft of The XDI Graph Model section of XDI Core. It is ready for final proofing.
You can see where this section fits in the full outline of XDI Core 1.0:
https://wiki.oasis-open.org/xdi/CoreSpecOutline
Discussion of this section took the first hour of the call. Key points of the discussion follow.
First we discussed the star symbol and its relationship to the term “authority”. Prior to the Star Shift, star had been classified as an XDI authority because we needed a way to model the identity of an entity that was not a person or a legal entity but still an independent “thing” (device, sensor, software agent, etc.) that could have its own independent XDI graph. Reclassifying the star symbol as an instance seemed to take away that meaning.
However Drummond explained that an instance can still have an authoritative graph. The only change is the application of the term “authority” to an instance node since as a “thing”, it is not actually the entity ultimately responsible for control of that graph. Some person or some legal entity is ultimately responsible—they are the real “authority”. Phil agreed with this semantics, and that the star space of instances would still correspond directly to providing XDI graphs for the Internet of Things.
Markus caught and Drummond fixed an error where the draft said that only context nodes could be the subject and object of relational statements; in fact literal nodes can also be the object of a relational statement.
Drummond showed the new Visual Graph Diagramming Notation section and confirmed that TC members were comfortable with the terms “Radial format” and “Tree format” for the two standard orientations.
We observed that we have only 3 more calls before IIW and that we want the committee draft to be complete and thought through.
#ACTION ITEM: MARKUS to send his other editorial suggestions to Drummond or the list.
#ACTION ITEM: ALL to further review Drummond’s draft and send comments/suggestions.
We reviewed a new draft of the ABNF building off Joseph’s latest draft for the Core spec. This new draft incorporates the StarShift, Immutability Symbol, and Xref proposals from last week’s TC call.
https://wiki.oasis-open.org/xdi/XdiAbnf
There were two open issues to discuss. The first is whether peer root nodes should allow “naked” URIs directly in parentheses, such as:
...or whether they should have to be expressed as an external reference, e.g.
(*(https://example.com/))
(=(https://example.com/))
(+(https://example.com/))
(#(https://example.com/))
Joseph asked what this is used for. Markus mentioned XDI connectors as an example.
The conclusion was that we should allow naked URI xrefs as it does not force the author to choose an XDI context type. However we noted that the latter examples remain valid and that it should always be safe to choose a star symbol as an XDI context because it only asserts that the target URI is an instance of some kind.
The second open issue was the question of what SHA algorithms should have XDI schemes. We reviewed the Wikipedia SHA page:
http://en.wikipedia.org/wiki/Secure_Hash_Algorithm
After some discussion, we arrived at two items of consensus:
#CONSENSUS: The ABNF for XDI scheme IDs will include hyphen characters.
#CONSENSUS: We will provide XDI schemes for all SHA-2 and SHA-3 algorithms except SHAKE128 and SHAKE256. The schemes names will be the XDI normalized version of the algorithm variants listed as SHA-* on the Wikipedia page, with the exception that SHA-512/224 will be sha-512-224 and SHA-512/256 will be sha-512-256.
Lastly we discussed the ABNF for URI characters. Peter initially suggested that it should just reference the ABNF for RFC 3986. Joseph explained that full parsing of URIs was generally going to be too expensive and overkill for XDI processors, so he had written a streamlined ABNF just for valid character checking. Peter agreed that approach made sense.
#ACTION ITEM: JOSEPH to update the URI ABNF and other rules that still have TODO comments.
There is a consensus that the second spec that needs to move to Committee Specification Draft status as soon as possible after Core is Messaging.
https://wiki.oasis-open.org/xdi/XdiMessagePatterns
Markus reported that the Messaging spec is still at an early stage - Working Draft 01 is over a year old and Working Draft 02 is still in progress. He agreed that ideally we will have a Working Draft of Messaging by IIW so that we can talk about both the graph model and the protocol from a spec standpoint.
#ACTION ITEM: MARKUS to do his best to post a full Working Draft 02 of XDI Messaging 1.0 before the special IIW meeting on April 8.
We confirmed that the Computer History Museum and IIW staff can provide us with a conference room with a phone line so we can hold an official meeting at IIW. The meeting will be held from 3:30-5:30PM on Wednesday April 8.
#ACTION ITEM: LES to send out a special Webex invite to the meeting.
The next call is next week at the regular time.
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]