2018-11-08-rsoc-minutes
slides-interim-2022-rfcedprog-08-sessa-2018-11-08-rsoc-minutes-00

RFC SERIES OVERSIGHT COMMITTEE (RSOC)
November 8, 2018 RSOC Meeting

Reported by: Cindy Morgan, IETF Secretariat

ATTENDEES
---------------------------------
- Sarah Banks (Chair) 
- Heather Flanagan (RSE, non-voting)
- Tony Hansen 
- Christian Huitema
- Cindy Morgan (Scribe, non-voting)
- Adam Roach
- Peter Saint-Andre
- Robert Sparks (Lead) 
- Portia Wenze-Danley (IETF LLC Board Liaison, non-voting)

GUEST
---------------------------------
- Sandy Ginoza

MINUTES
---------------------------------

1. Administrivia

  The minutes of the 2018-10-24 teleconference were approved.

2. Agenda Bash

  No new items were added to the agenda.

3. SLA Deep Dive

  Heather Flanagan walked the RSOC through the current RFC Production 
  Center SLA. The SLA has three tiers:

  - Tier 1: When there is a normal amount of input, the SLA is 67% 
    of documents published within the period have an RFC Editor-
    controlled time that adds up to six weeks or fewer. 
    * 'normal' is defined as less than 1950 Pages gone to EDIT (PGTE).

  - Tier 2: When there is a moderate burst in the amount of input, 
    then the SLA shifts to 50% of documents published within the period 
    have an RFC Editor-controlled time that adds up to 12 weeks or fewer 
    within the given quarter or the subsequent quarter (marked as Tier 
    2*). 
    * 'moderate' burst is defined as 1950 - 3072 (inclusive) Pages gone 
      to EDIT (PGTE).

  - Tier 3: When there is a large burst in the amount of input, then the 
    SLA must be discussed and renegotiated. 
    * 'large' burst is defined as greater than 3072 Pages gone to EDIT 
      (PGTE).

  This model has been used for the SLA since 2014. Heather Flanagan 
  noted that they have not hit Tier 3 in a while, but that it is likely 
  that they will again when Cluster 238 breaks free.

  Heather Flanagan said that she is not proposing any changes to the SLA 
  now, but that the RSOC should look at it again once there is some 
  experience with the running code of the new RFC format. There are 
  concerns that the new format will make editing documents more 
  difficult, and that more hands-on effort will be required to get the 
  XML correct. 

  Robert Sparks noted that the goal is to have tools that make 
  everything as easy as possible for the RPC, but there will be no way 
  to tell how things will be affected until they have some real 
  experience. 

  Christian Huitema asked if there are any consequences for missing the 
  SLA. Heather Flanagan replied that there are no financial 
  consequences, but if the SLA was missed consistently for no good 
  reason, it would be something to discuss when reviewing the RPC 
  contract. Robert Sparks added that if the SLA is consistently being 
  missed, then there is likely a bad model of what we are asking the SLA 
  to do, and if the SLA is never missed, then it is overly permissive. 
  The occasional miss is a good metric that what the SLA is looking at 
  is actually meaningful.

  Christian Huitema asked why the SLA measures the number of pages 
  rather than the number of documents published. Heather Flanagan 
  replied that looking at the number of documents did not capture the 
  amount of work required, since a document could be 5 pages long or 200 
  pages long and would count the same in a document-centric SLA. Robert 
  Sparks added that with the new format and XML becoming the canonical 
  format, pages may no longer be a useful reference; something else may 
  need to be synthesized  that acts as a proxy for pages coming in. 
  Heather Flanagan said that in the meantime, they would count the 
  number of pages in the PDF versions of the documents.

  Adam Roach asked if the RPC would be touching the SVG diagrams in 
  documents. Heather Flanagan replied that the RPC would make sure that 
  the SVG passes required checks, but that they will not be trying to 
  fix the SVG themselves for authors. Sandy Ginoza added that there will 
  likely be a basic level of cleanup to make sure that spelling matches 
  between the images and the text of the document itself.

4. Format Update and Transition Plans

  Heather Flanagan updated the RSOC on the current status of the RFC 
  format update and transition plans.

  - Project status and discussion
    * https://trac.tools.ietf.org/tools/ietfdb/wiki/FormatToolsPlan
    * https://www.ietf.org/mailman/listinfo/xml2rfc-dev
 
 - Released to the RPC and the community
    * id2xml: text to v2
    * id2xml: text to v3
    * xml2rfc v2 to v3 converter
    * rfc-xmldiff
    * rfclint
    * svgcheck
    * publication formatters (TXT, HTML)

  - In Progress
    * preptool
    * publication formatters (PDF A/3)

  Heather Flanagan reported this information during the WG Chairs' Forum 
  at IETF 103 and noted that people seemed engaged and eager to start 
  playing with the new format.

  Tony Hansen asked what tools should be made available in the web 
  interface from XML2RFC besides the translation to I-D format and the 
  v2-to-v3 translation. Heather Flanagan replied that idnits and 
  svgcheck would be helpful.

  Sandy Ginoza said that she is hoping there is a way to differentiate 
  between UTF-8 characters where they are wanted versus where they are 
  not wanted. She will check with AMS IT support and see if they have 
  any recommendations. Adam Roach added that idnits can flag for things 
  like smart quotes and long dashes.

  Heather Flanagan reported that works continues on the HTML format, and 
  that she is open to EPUB as a format if someone can write a reasonable 
  spec. 

  Christian Huitema noted that there is a cost to having multiple 
  formats, both for the RFC Editor and for the authors who have to 
  verify all of the formats. 

  Heather Flanagan reviewed the v3 transition goals:

  - Keep documents moving and minimize backlog during transition
  - Ensure quality of RFCs published in the new format
  - Clean switch to publication of v3 RFCs
  - Avoid having simultaneous v2 and v3 AUTH48 processes

  Adam Roach noted that there is a risk that Cluster 238 will break free 
  around the same time that the process transitions to v3; the work of 
  converting all of the Cluster 238 documents to v3 would be 
  substantial. Heather Flanagan said that her preference would be to 
  delay the transition until after all of the Cluster 238 documents are 
  in AUTH48.

  Transition components include:

  - Operations and Queue Management
    * Add a conversion step to the existing process 
      (remaining v2 files -> v3 files)
    * Expand the scope of AUTH48 
    * Use a new state for a tooling-related hold (e.g., if a document is 
      held pending the resolution of a bug in the new toolset) 
  - Education
    * Update internal procedures and train staff
    * Provide new AUTH48 instructions
    * Contribute to documentation, FAQ, tutorial
  - IT Support
    * Preparations started months ago; includes a set of updates to the 
      scripts, the public web resources (e.g., info pages, 
      rfc-index.xml, errata system), staff web interface, and internal 
      directory structure. _
    * Cutover to new system before publishing the first v3 RFC.

5. RFC Brand, Streams, and Statuses

  Heather Flanagan reported that she has met with Tenet to discuss the 
  RFC Brand, Streams, and Statuses project. The scope of the project is 
  to determine how to clearly measure the confusion issue around streams 
  and statuses, and to review options for resolving the main confusion 
  issues.

  An initial investigation will be required to determine who uses RFCs: 
  network operators, governments, developers, and the IETF community.

  Tenet proposed:

  - Interview approximately five IETF members (assuming they might be 
    able to provide a clearer picture regarding the RFC audience)
  - Alternatively: provide any additional audience demographics: 
    geographies, job functions, companies, etc.
  - $3000-5000 for the initial phase. (5 interviews x 1 hour plus 5 
    hours of prep and 5 hours of report preparation.)

  Heather Flanagan noted that the 2019 IETF budget has been closed, so 
  she is not sure where the funds would come from. Portia Wenze-Danley 
  replied that the budget was prepared in such a way so that the new 
  board would have some flexibility later on, and that she would take a 
  look and see if she could find some discretionary funds for this.

  Adam Roach questioned how useful an interview process that only talked 
  to five members of the IETF community would be; if the only input is 
  from the IETF then they may as well listed to the recordings from the 
  RFCPLUSPLUS BOF.

  Heather Flanagan asked if this seems like the right path to take.

  Robert Sparks replied that he is skeptical that this would bear fruit, 
  but he is willing to see it out. He suggested that the RSOC could ask 
  people in the community to write up what their concerns are. Heather 
  Flanagan said that getting that anecdotal concerns would be easy 
  enough, but that it would be difficult to get something more 
  measurable.

  Christian Huitema said that a first step could be to compare those 
  anecdotes; Tenet could take the anecdotes and set up a panel to see if 
  there is a way to get more measurable data.

  Sarah Banks asked if the RSOC cares strongly enough about this to take 
  further action.

  Robert Sparks replied that conversations about this confusion have 
  happened on multiple IESGs and IABs over the past 20 years; there is a 
  subset of the community that thinks that the harm being done by the 
  current confusion is great enough to do something very radical, i.e. 
  getting rid of all streams other than the IETF stream, or publishing 
  documents that are not RFCs. 

  Adam Roach said that the only thing there was consensus for during the 
  RFCPLUSPLUS BOF was that the RFC Editor should take on the next steps 
  in reviewing the RFC streams and statuses confusion, and that this 
  would be her taking it on. He added that it is not the same specific 
  people who have even considering the problem for 20 years; the issue 
  has repeatedly arisen in different IESGs and IABs.

  Heather Flanagan updated the RSOC on already-existing RFC Series 
  marketing and branding efforts:

  - Google Scholar indexing of RFCs
  - Adding stream to references (in progress)
  - DOI assignment to support academics
  - Socialize idea of using separate CSS for different streams, 
    statuses (ongoing)

6. SDO Comparison

  Heather Flanagan provided the RSOC with some information about how 
  other SDOs publish documents:

  - W3C: Project management staff act as copyeditors; multiple CSS; one 
    style guide
  - ITU-T: Editors are assigned to a WG for the duration; also act as 
    project management and coordination; one style guide
  - ISO: Secretariat includes copyeditors; multiple style guides
  - IEEE-SA: Staff copyeditors; one style guide

  Christian Huitema noted that an ITU editor is more like a Working 
  Group Chair in the IETF.

7. Misc Dates

  Heather Flanagan asked the RSOC to make not of the following dates:

  - January 1, 2016 - current RPC contract for two years, with 2 two 
    year extensions allowed
    * Re-bid and vendor selection = 2021
  - January 1, 2018 - current RSE contract for two years, with 2 two 
    year extensions allowed
    * Re-bid and vendor selection = 2023

  The RSOC is responsible for running the process when the RFC Editor-
  related contracts go out to bid.