[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: XDI TC Notes Unofficial Telecon Friday 2015-08-31
Following are the notes of the unofficial telecon of the XDI TC held on:
Date: Monday, 31 August 2015 USA
Time: 10:00AM - 11:30AM Pacific Time (17:00-18:30 UTC)
The TC operates under a standing rule approved 17 July 2008 under which the TC does not hold regular official meetings and conducts all business by electronic ballot only. Unofficial weekly meetings are held to enable discussion among members but no business is conducted nor actions taken.
Peter Davis
Markus Sabadello
Drummond Reed
Phil Windley
Les Chasen
Joseph Boyle
Christopher Allen
Lionel Wolberger
After our decision during the XDI Core Immersion calls last week that order numbers (e.g., @~1, @~2, @~3) needed to be relative identifiers, Drummond realized they have moved further from being directly comparable to an array, and closer to an ordered list.
Which raises an issue: while the order count for an array most often begins with zero, for an ordered list it begins with one. The question for the TC: which is going to be more intuitive for developers and users of XDI graphs?
Drummond points out this quote from the Wikipedia article about natural numbers:
There is no universal agreement about whether to include zero in the set of natural numbers. Some authors begin the natural numbers with 0, corresponding to the non-negative integers 0, 1, 2, 3, ..., whereas others start with 1, corresponding to the positive integers 1, 2, 3, ....[7][8][9][10]
Peter questioned whether we need ordered sets in XDI. Drummond explained that we need them for the same reason RDF needs them, and even that HTML needs them (ordered lists).
Joseph felt that if it is computer-based, it should be zero-based. An implementer can always choose a higher number.
Markus said there are three arguments for using one as the starting number:
In RDF containers (Bag, Seq, Alt), instances start with one.
The semantics line up with human semantics.
The mathematical concept of natural numbers (counting numbers).
We talked about using other order identifier schemes. Here is an example of alphabetic ordering:
{
"=example#favorite[#author]@~:alpha:poe": {
"$ref": [
"=example[#author]=poe-edqar-allen”
]
},
"=example#favorite[#author]@~:alpha:vonnegut": {
"$ref": [
"=example[#author]=vonnegut-kurt"
]
},
"=example#favorite[#author]@~:alpha:zelda": {
"$ref": [
"=example[#author]=zelda"
]
}
}
And here is an example of default numeric ordering starting with one:
{
"=example#favorite[#car]@~1": {
"$ref": [
"=example[#car]*!:uuid:f81d4fae-7dec-11d0-a765-00a0c91ea751uuid:33ad7beb-1abc-4a26-b892-466df4379a51"
]
},
"=example#favorite[#car]@~2": {
"$ref": [
"=example[#car]*!:uuid:9ce739f0-7665-11e2-bcfd-0800200c4bd2uuid:f336a645-f5a9-41b7-ab80-ace41a8f69c2"
]
},
"=example#favorite[#car]@~3": {
"$ref": [
"=example[#car]*!:uuid:3a96e460-7be9-f7e4-b92a-83d3c45e76a3uuid:1c958708-d5aa-4213-a6a9-73dd423502b3"
]
}
}
Peter argued that in JSON, order is not deterministic, and he noted that in the JSON serialization format XDI is using, arrays are 0-based.
Markus found the following Wikipedia article:
https://en.wikipedia.org/wiki/Zero-based_numbering
Markus felt that starting with 1 makes more semantic sense.
Joseph pointed out that human readability does not really factor into this. He provided a link to where RDF processing used zero-based numbering:
http://www.w3.org/2001/sw/Europe/events/20031113-storage/positions/asemantics.html
Markus commented that the above paper only uses 0-based indices for display purposes, but not for ordering resources in the RDF graph model.
Peter and Phil said they believe the core issue is whether XDI list ordering is “programmer-friendly” or “human-friendly”. Peter believes the former. Joseph and Les also agrees. Phil agree because it will actually make it more “program friendly”.
Markus can live with it, and so can Drummond.
#CONSENSUS: We will remain with zero-based ordering as the default for order numbers.
Drummond asked whether we should allow other ordering schemes such as alphanumeric. Markus felt it should be supported but not emphasized much.
See this table from the latest draft of Drummond’s sections of XDI Core:
XDI Term | Includes Person | Includes Group | Includes Thing |
XDI person | Yes | No | No |
XDI authority | Yes | Yes | No |
XDI actor | Yes | Yes | Yes |
Drummond explained the terms XDI person, XDI authority, and XDI actor. Joseph said the choice of terms sounds okay. Peter suggested they should be covered in the glossary, and everyone agreed.
Markus asked about the relationship between these terms and the terms “XDI endpoint” and “XDI agent”. Drummond said the latter terms should be defined in the XDI Messaging spec, but expressed that the core difference is that “XDI endpoints” and “XDI agents” are not themselves XDI actors. His view is they are INVOLVED with XDI messages but do not INITIATE them.
Joseph suggested that changing the term “Order” (as one of the four XDI entity instances along with Person, Group, and Thing) to “Ordinal”.
#DRUMMOND will look into making this change in the Core draft.
We reviewed Drummond’s latest update to the XDI Core 1.0 draft coming out of “XDI Core 1.0 Immersion Week”:
https://www.oasis-open.org/committees/document.php?document_id=56383&wg_abbrev=xdi
This version of Drummond's sections of XDI Core is now complete through the Core Relations section. The two remaining normative sections Drummond needs to complete are: 1) XDI Addressing, and 2) Versioning.
Markus noted that there is another good example of nominalization, which is the parameters on an XDI protocol operation.
#MARKUS to send Drummond an example to add to this section.
The goal is to take Core to a Committee Specification Draft vote by the end of September. The schedule Drummond proposes:
Week of Aug 31: Complete XDI Addressing & Versioning.
Week of Sept 7: Complete DocBook migration and ABNF
Week of Sept 14: Complete Serialization
Week of Sept 21: Complete Introduction and Example Graph
Week of Sept 28: Complete Appendicies
We did not further discuss this topic.
The next call is next week at the usual time (Monday 10AM PT). The link to where agenda items can be posted for the next meeting is: https://docs.google.com/document/d/19oDl0lbb56Grehx2a5flZnhrgnua5l8cVvC_dJ8fTXk/edit?usp=sharing
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]