XDI TC Minutes
Following are the minutes of the unofficial telecon of the XDI TC at:
Date: Friday, 01 March 2013 USA
Time: 09:00AM - 10:30AM Pacific Time (17:00-18:30 UTC)
ATTENDING
Les Chasen
Animesh Chowdhury
Markus Sabadello
Drummond Reed
Joseph Boyle
REGRETS
Phil Windley (who tried but could not get into the call)
NEWS & UPDATES
WELCOME TO ANIMESH CHOWDHURY, NEUSTAR
Animesh has formally joined the TC and will be working in particular on policy _expression_.
PRESENTATIONS/DISCUSSIONS
FILE SYNCHRONIZATION
We discussed how XDI data synchronization could be combined with conventional file synchronization (e.g., Dropbox, Owncloud). The following points were raised:
- XDI is not designed directly for file-level synchronization.
- But it could be used in a hybrid with a file-sync protocol.
- This would be similar to the connector architecture that’s already been implemented by XDI2. The file store that doing the sync would be treated as an external data source to which the XDI graph would be mapped.
DECISION POINTS FOR THIS CALL
This week's decision queue is the following set of proposals:
ABNF UPDATES (DRUMMOND, MARKUS, JOSEPH)
We continued discussion on the key items raised on the list over the past week.
We discussed the issue of the why the Interactive APG parser is treating =markus$!(+email) as three subgraphs:
Drummond asked why =markus$!(+email) was not treated as four subgraphs:
Markus explained that the source of the problem was the special rules around $, because $ is valid as a subgraph/subsegment all by itself. Drummond noted the same was true for +, =, and @, yet the parser does not seem to have problem with those global context symbol delimiters.
Markus asked if we could change
specific = "$" [ xdi-literal ]
to
specific = "$" xdi-literal
This eliminates ambiguity and appears to make ABNF-based parsers work correctly.
Drummond asked about this string: =markus+!(+email)$!(+tel)/!/(data:,...)
Markus tested it with the APG ABNF parser and it worked - the +!(+email) string was seen as a single subgraph.
Drummond asked whether the current form of the Full Graph ABNF is causing the $ problem because it breaks down all the rules to starting with a single delimiter. He suggested trying a version of the ABNF that goes to the other extreme and includes all the delimiter patterns that can be matched directly.
# DRUMMOND to create this alternative version so we can test it to see if it changes the behavior of the APG parser.
We next discussed the literal-context rule problem. As we currently have the ABNF,
=markus$!(+email)/!/(data:,...) does not match the xdi-statement rule because
=markus$!(+email) does not match the attr-singleton-context rule.
attr-singleton-context = [ peer ] *( [ attribute-singleton ] 1*non-attr-singleton ) 1*attribute-singleton
Markus noted that that rule can be simplified to:
attr-singleton-context = [ peer ] *( attribute-singleton / non-attr-singleton ) attribute-singleton
It is not yet clear how to solve this issue.
Lastly, we discussed the proposed new terminal-context rule.
terminal-context = literal-context "()"
Markus asked if the terminal () was considered a subgraph. Drummond’s said no - in the XDI graph model, it is not a literal node.
A sample operation in a message would be:
=sender$($msg)$(!1)$do/$get/=markus$!(+email)()
Markus was not in favor of this approach because the () appears too much like a new context.
# DRUMMOND will develop a new proposal that will not be confused with a new context.
XDI MESSAGE PATTERNS (MARKUS & DRUMMOND)
Last week Drummond updated the XDI Message Patterns proposal page and formally added the Parameters pattern section.
It was agreed that the revised page made the message template easier for newcomers to understand, and to reference from the sections of the page.
# MARKUS - to add a variable called <result-type> to the OPERATION line and a corresponding section to the page explaining how to control result types.
# ALL - Review this page to see if there are any open issues, otherwise we’ll put it into Last Call after next week.
PROXY DISCOVERY (DRUMMOND & MARKUS)
Last week, Drummond updated the Discovery proposal to include the very simple rules for proxy discovery:
# DRUMMOND AND MARKUS - Add any other details about proxy discovery that may be needed.
DECISION POINT QUEUE REVIEW
The decision queue stack is now shown on the following three auto-generated Category pages:
See also this list of proposals that need to be developed:
NEXT CALL
The next call is next week at the regular time.