Minutes: XDI TC Telecon Friday 2013-02-08

[Drummond Reed <drummond.reed@xdi.org>]

XDI TC Telecon Minutes

Following are the minutes of the unofficial telecon of the XDI TC at:
Date:  Friday, 08 February 2013 USA
Time:  09:00AM - 10:30AM Pacific Time (17:00-18:30 UTC)

ATTENDING
Markus Sabadello
Joseph Boyle
Drummond Reed
Les Chasen

REGRETS

Bill Barnhill

NEWS & UPDATES

TRANSITIONING CO-CHAIR ROLE

Drummond posted the ballot to elect Markus Sabadello to succeed Bill Barnhill last Sunday Feb 3.
The election technically ends on Sunday Feb 10, however a majority of votes are already in favor.

PRESENTATIONS/DISCUSSIONS

TYPED VARIABLES

Drummond gave a short overview about Typed Variables, which will be a key component of XDI forms:
 https://wiki.oasis-open.org/xdi/TypedVariables

DECISION POINTS FOR THIS CALL

ABNF UPDATES

After good discussion on the list, and on the ABNF discussion page, the current proposal is posted on the discussion page.
 https://wiki.oasis-open.org/xdi/XdiAbnf/Discussion
 https://wiki.oasis-open.org/xdi/InnerRoots
We discussed GCS (global context symbol) and LCS (local context symbol) characters and confirmed the following:

• Identifiers in the LCS character namespaces are always local in scope
• Identifier prefixed with ! are immutable -- assigned to an entity once and never assigned
• Identifier prefixed with ! are mutable -- they may be reassigned to different entities
• From the standpoint of immutability, the GCS characters fall into two groups:
• $ and + identifiers are immutable by default - the override to indicate mutability is $* and +*
• = and @ identifiers are mutable by default - the override to indicate immutability is =! and @!
• The reason for four GCS characters is authority
• = is for authorities who are natural persons
• @ is for authorites who are not natural persons (other legal entities)
• $ is for standards bodies (including OASIS and the XDI TC)
• + is for generic identifiers who have no single authority

We also discussed Joseph’s list of questions posted to the mailing list:

• #1: Yes, GCS and LCS characters can be a subsegment by themselves. An example is the ! used as the entire predicate to represent an XDI literal arc.
• #2: This is still to be determined. See the two ABNF variants posted by Drummond on 2013-02-09 at https://wiki.oasis-open.org/xdi/XdiAbnf/Discussion.
• #3: It was agreed we don’t need the xdi-context rule.
• #4: Joseph and Markus agreed that it would be good to do some parsing tests to profile how the grammar is working.
• #5: question answered on the mailing list.
• #6: We discussed tilde. Joseph thinks it’s there because it was a convention for UNIX usernames. No decision yet.
• #7: We should follow the convention of IRI, but there may be reasons in deployment to exclude characters in Unicode general categories. Joseph brought up homographs. We agreed we need to address that in our security considerations. Markus and Joseph also discussed that Java also has issues with characters above the Unicode BMP.
• #8: Yes, we are requiring a client to do some percent encoding of IRIs. We noted that we must include double-percent encoding rules.
• #9: We agreed we want to use a modified IRI grammar. Joseph will propose this.
• #10: By using a special IRI character grammar, we should be able to greatly simplify parsing.
• #11: Yes, we should support the full IRI scheme. We also reinforced our decision that it is not necessary to require a leading colon for an IRI. Drummond said that, for the record, the concept of doing so is completely consistent with all other XDI subsegments having leading delimiters. However, since the parentheses themselves are a delimiter, it is conceptually and visually simpler not to require another delimiter character in the form of a colon.

We agreed that all major substantive decisions about the ABNF are now decided. We also agreed that the ABNF we publish in the XDI 1.0 specs should be optimized for semantic expressiveness and not for parsing speed. We can also publish a separate informative version optimized for parsing speed.
So the remaining tasks to complete the ABNF are:

1. Decide on how we want to optimize it for semantic _expression_.
2. Specify the IRI character set allowed to provide for simplified parsing.

Our goal is to move this into Last Call next week.

ACTION ITEMS ASSIGNED

# DRUMMOND to post proposals for optimized semantic _expression_ to https://wiki.oasis-open.org/xdi/XdiAbnf/Discussion (DONE).
# JOSEPH to submit proposals on the options for IRI inclusion in the ABNF.

EQUIVALENCE LINKS (DRUMMOND & MARKUS)

Last week we agreed this proposal could go into last call, but in working on examples of $ref! links, Drummond realized that the semantics of $ref and $ref! links were different enough that it would be much better to use a different $word for $ref! links. He and Markus propose  $rep (for “replacement”). Drummond revised the proposal to use $rep and found it dramatically easier and clearer to explain.
 https://wiki.oasis-open.org/xdi/EquivalenceLinks

Drummond also pointed out that now the proposal clearly defines both the semantics and processing rules for 3 distinct kinds of equivalence links:

1. Identity links ($is)
2. Reference links ($ref)
3. Replacement links ($rep)

Markus suggested changing the proposed $noref dollar word to $deref. Drummond agreed.

# DRUMMOND to change $noref to $deref in the proposal.

# MARKUS to review this semantics and see if he has any other revisions.

Our goal is to move this into Last Call next week.

DECISION POINT QUEUE REVIEW

The decision queue stack is now shown on the following three auto-generated Category pages:
 https://wiki.oasis-open.org/xdi/CategoryLastCall
 https://wiki.oasis-open.org/xdi/CategoryCurrent
 https://wiki.oasis-open.org/xdi/CategoryHighPriority

See also this list of proposals that need to be developed:
 https://wiki.oasis-open.org/xdi/XdiPendingIssues

We discussed the need for two new proposals:

• A proposal to extend https://wiki.oasis-open.org/xdi/SimpleHTTPTransport to include a method for doing an XDI $get with an HTTP $get.
• Potentially this proposal could be extended to a completely RESTful interface where an XDI $add could be performed by an HTTP PUT, and XDI $mod by an HTTP POST, and an XDI $del by an HTTP DELETE.
• A proposal to extend the XDI Discovery spec to specify the function of HTTP/S proxies.

NEXT CALL

The next call is next week at the regular time.