[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: Minutes: XRI TC Telecons Thur/Friday 2006/1/12-13
Minutes of the following two XRI TC telecons: US/ASIA TELECON Date: Thursday, 12 January 2006 Time: 9:00pm - 10:30pm PT PRESENT Kunal Mohit Drummond Steve US/EUROPE TELECON Date: Friday, 13 January 2006 Time: 10:00am - 11:30am PT PRESENT Marty Les Gabe Drummond Steve Jamie Clark (OASIS) AGENDA 1) OASIS VOTE PREP (discussed on the Friday call only) Jamie Clark, OASIS Director of Standards, joined us to continue discussion of preparations for the OASIS Standard vote. Gabe clarified that in discussion with Jamie and OASIS staff following the decision to delay the vote in mid-December, it was decided that would be prudent not to attempt to have the vote in February and to delay it at least until March so that we can have the conversations necessary to ensure that the vote will be successful. In the meantime we are proceeding to the Committee Specification level on the remaining specifications in the XRI 2.0 suite as quickly as possible. Jamie explained that one step in this process he is helping with is preparation of a more formal reply to the comments submitted by the W3C TAG during the XRI public review period. It was agreed that this response should ultimately become a public document in the XRI TC repository. Marty asked to clarify the goal of this document. There was concensus that the end goal is a successful OASIS Standard vote, and that towards this end the purpose of this document will be both for communication with the TAG W3C as well as to educate the rest of the OASIS voting audience. Marty felt strongly that this communication should not be adversarial but simply clarify that XRI was developed based on a different perspective of the use of identifiers. Jamie asked for input about what we felt we should focus on as the heart of this document, i.e., the rationale for why XRIs and XRI resolution are needed vs. just using URIs (particularly HTTP URIs) and DNS/IP. Two potential responses: a) no, you can't do everything with HTTP URIs and redirects, and b) that you should never rule out future identifier schemes. TC feedback: 1) The issue of persistence is not the strongest case for XRI architecture. It is true that XRI is the only architecture that offers a common syntax and resolution protocol for both persistent and reassignable identifiers and a method for establish equivalence between the two. However the proponents of HTTP URIs may argue endlessly that you can accomplish the same thing with careful identifier management and HTTP redirects. 2) A much stronger case is that XRI is the only identifier architecture that implements cross-references (polyarchy), which among other things permits identifier metadata to be included in identifiers. A wonderful example is the NAC XRI Profile, which makes extensive use of the XRI Metadata specification, which is based entirely on XRI cross-reference syntax. 3) Another strong suit is Gabe's argument for why Visa has been pursuing XRI architecture (see http://lists.oasis-open.org/archives/xri/200601/msg00002.html and http://lists.oasis-open.org/archives/xri/200601/msg00003.html.) Gabe's "stitching" argument about "what may be closed now may later become open" is a very strong one for the XRI authority model, which is much more flexible than HTTP URIs. 4) Another key point is that XRI offers an identifier architecture that is fully compatible with the Web but may also extend to resources that are not on the Web. This applies to a closed body of data that is behind firewalls and not available on the Web (but which may one day become available on the Web.) 5) Lastly, a final key point we should be making is that with the revised XRI 2.0 resolution framework, we have achieved a high level of backwards compatability with HTTP URIs, i.e., there is a defined transformation of all XRIs into an HTTP URI format (which we call "HXRI"). This last point is important because it, together with the fact that XRI Syntax defines a transformation of an XRI into an IRI, shows that we are making sure XRI is compatible with W3C TAG architecture. Furthermore XRI ends out being a case in point of the adaptability of URI syntax that W3C has long championed -- it is one of those identifier schemes that illustrates the need to allow for the continuing evolution of identifiers. Marty pointed out that XRI is in the same category as IRI in this respect. We then discussed the potential form of the W3C TAG response. Jamie proposed that OASIS staff make the response, with support from the TC in the form of the document discussed above. Finally, we discussed other preparations for the vote. Jamie suggested that historical evidence at OASIS shows it is worth campaigning if a TC has a good pitch. Who should do it? Jamie said that the TC and OASIS staff need to work together to decide this. The final action items from this discussion: ACTION ITEM: DRUMMOND, GABE, MARTY to begin TC work to create the "XRI Rationale" document discussed above (beginning with choosing a name for this document.) ACTION ITEM: GABE, MARTY, and JAMIE as the "vote preparation coordination committee" need to decide which conversations need to happen with whom. 2) METADATA REPORT Marty and Drummond briefly reported that they are in the final stages of preparing XRI Metadata 2.0 Working Draft 08. They will direct further questions to the email list. 3) XRI RESOLUTION REF PROCESSING ISSUE At the very end of Friday's call, we discussed the proposal from Steve Churchill for how to handle recording XRD elements that result from following a Ref element into an XRDS document. See: http://lists.oasis-open.org/archives/xri/200601/msg00019.html The conclusion was that we should follow this general pattern but use nested XRDS documents instead of introducing a new element. ACTION ITEM: XRI Resolution 2.0 editor's team to incorporate this into the next editor's draft.
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]