Date
Attendees
Discussion items
Time | Item | Who | Notes |
---|---|---|---|
announcements: | Some discussion of need for "global" review of spec before release for review so that changes all cohere, more polished view, and can give good high-level Curtis and John spoke with some of repondants to poll. Comment from Pablo ? from Argentina that it is hard to find docs on ARKS (where to go - spec out of date, do I go to AITO?) He went to arks.cdlib.org – Need to clean this up quickly (part of outreach groups remit) (Pablo created Spanish-language ARK wikipedia page, is willing to translate FAQ as well) Roxana reporting on enhancements to their resolvers, etc. ; waiting on final spec for other changes | ||
(from last time) research handling of query parameters that have no explicit value (eg. ...??) by HTTP servers and reverse proxies. | Tom | No clear one place to go for clarity on behaviors of reverse proxy servers; Specs vs. was is coded; suggestion is to write a script and test against actual implementations; Suggestion that Tom will continue to monitor, he will try against top 3 proxies to see results | |
question of collapsing ? and ?? | Greg: refers to Bertrands suggestion to use query term (non-empty) - why would we not? John: question as to whether appearance of English words in standardized query an issue (internationalization) (eg info) 2nd issue: historically IETF reluctant to do anything to standardize query string - although using ? or ?? in effect standardizing anyway Tom: 'meta'; ark-info, ark-meta; maybe also consider as optional method of support is via request-header rather than in query string; would provide greater latitude in describing methods of formatting response Mark: value also in having URL-like string that functions Greg agrees use cases for both techniques - string easier for "naive" implementation- question would this be an extension header (X-)? Tom yes - but style now for non-standardized without X quite common; also agrees, even if we do headers, useful to have query string John: given we will approve inflection supported, considering adding one with word - John would like to continue; could have language like ?info preferred, reserve? and ?? since that has been in spec for so long Curtis - prefers info, about - ? ok if words, options to add more operations in future John - we are moving to, at very least ? means same as ??, and ?? preferred Should be one query that gives basic information plus persistence info Sheila prefers word, okay with both Mark: has gotten feedback of questions about ? – looks strange to many for usual URLs; also implementations low; using word, making it "look" like query parameter would encourage greater use and implementation; re "both" - Might be confusing to have 2 ways of doing it | ||
question of NAAN assignment corresponding to ISSN | |||
when to open spec review to arks-forum |
Action items
- John Kunzewrite up summary of new take on inflections