IETF78 Maastricht PRECIS Working Group Meeting Minutes

Date:   Thursday, July 29, 2010
Time:   09:00-10:30
Chairs: Marc Blanchet and Yoshiro Yoneya

1) administrativia (chairs)

Chairs: Problem statement and framework are done in parallel. Other deliverables are after problem statement.

2) problem statement (Blanchet)
   draft-blanchet-precis-problem-statement-00.txt

 In the charter and in initial work for problem statement, only the profiles registered in the IANA stringprep profile registry were looked at. However, it was discovered during the meeting that there are at least 39 RFCs referring stringprep. see:
     <http://www.arkko.com/tools/allstats/citations-rfc3454.html>
not counting drafts
also: http://www.fenron.net/~fenner/ietf/deps/index.cgi?dep=rfc3454

Transition guidelines from current stringprep profiles to the new solution are currently not in the problem statement. It would likely be in another document, but more work has to be done on the solution before identifying the transition guidelines.

  Andrew Sullivan has kindly volunteered to be co-author of the problem-statement document.
  Consensus in the wg to adopt the document as WG document.

3) framework (Blanchet)
   draft-blanchet-precis-framework-00.txt

Some comments:
- SASL trace may be considered as bad. don't take care of it for now.
- "Commonly mapping to nothing" characters need extreme care for example used in passwords.
- Updating stringprep profiles and internationalizing them are different thing.
- Stringprep customers might not know internationalization well.
- Choosing NFKC at this moment is premature.  Shuold survey difference between NFKC and NFC.
- most likely more string classes than the two listed

 Concensus in the room that we need to survey the current specifications that are referring to stringprep (normative or not) to get a better view of the usage of stringprep, not only in the profile itself, but also outside of the profile (pre or post processing), and if the actual profile corresponds to the intent of the specification authors. This would help see if the proposed strawman approach in the framework described is the right or needs more string classes or another approach is needed.

  Looking for co-author(s) of the document.
  Concensus to keep the document as individual document for now. revisit in Beijing.

4) next steps (chairs)

  Action item: setup survey team for all existing stringprep profiles that are found. Chairs will browse each and create a specific ticket for each document referring to stringprep. The tickets will be created in the working group tracker: http://tools.ietf.org/wg/precis/trac/report/1

  Each ticket will be assigned to a working group member and will be requested to review the referring document and to provide a summary in the ticket. 

 When the tickets will be created, the chairs will send to the wg mailing list a request for volunteers to take on some tickets.

 As the first reviews are done, the review team will prepare a list of questions/topics to be filled out as a template for the reviews. Starting point is the list of issues in the problem-statement document.