Meeting Date: 2022-10-11
Minutes taken by: Matt
Attendees:
    Dmitry Belyavskiy
    Tomas Mraz
    Richard Levitte
    Nicola Tuveri
    Matt Caswell
    Matthias St Pierre
    Paul Dale
    Shane Lontis

External attendees:
    Daniel Fiala

Agenda:

    * Nominate a minute taker and confirm agenda
    * Agree minutes from previous meeting
    * Review outstanding actions
    * Review outstanding security issues
    * RIPEMD160
    * Review issues with OTC hold (30 minutes)
    * Review draft technical policies
    * Agree agenda for next meeting
    * AOB

** Nominate a minute taker and confirm agenda

Matt volunteered to take the minutes.
Proposed: Matt, seconded: Dmitry

Agenda for the meeting agreed. Proposed: Matt, seconded: Tomas

** Agree minutes from previous meeting

Minutes from previous meeting were agreed. Proposed: Tomas, seconded: Matt

** Review outstanding actions

AI: Richard to create instructions on how to add an OTC member.
Not done

AI: Dmitry to raise a public GH issue about the infinite session prolongation
in TLS-1.3.
Done

AI: Martin to propose a plan for location and timing of the OTC F2F meeting.
Not done

AI: Matt to create a proposal on return values and return value types for new
code.
Not done

AI: Martin to propose some metrics on outstanding GH PRs and issues.
Not done

AI: Martin to identify a resource to come up with proposal for PRs to backport
to 3.1 for consideration by OTC for recommendation to OMC.
Done

AI: Martin to identify a resource to write a proposal for an improved design
process
Not done

** Review outstanding security issues

We discussed outstanding security issues.

** RIPEMD160

There was a discussion whether it was a mistake to remove RIPEMD160 in 3.0. The
question now is what to do about it. We have a proposed fix which additionally
makes RIPEMD160 available in the default provider.

OTC refer the issue to OMC for a decision. An OMC hold has been placed on the
PR (#19375). Also see issues #17722 and #16994.

** Review issues with OTC hold (30 minutes)

#17984 disable 5x interleave on buffers shorter than 512 bytes

Discussion about whether this should be backported to 3.0 or not.

[Matthias left]

OTC Vote:
OTC considers PR#17984 as a bug fix and therefore can be backported to the 3.0 branch

  Dmitry     [+1]
  Matt       [ 0]
  Pauli      [ 0]
  Tim        [  ]
  Hugo       [  ]
  Richard    [+1]
  Shane      [+1]
  Tomas      [+0]
  Kurt       [  ]
  Matthias   [  ]
  Nicola     [-0]

AI: Matt to create issue for the OTC vote on PR#17984

#18864 Should libctx be added to the OSSL_set_max_threads() and
OSSL_get_max_threads() functions?

The consensus is that, yes, we should add libctx to those functions

#19248 Loading a config file with SSL/TLS config overwrites config from another OSSL_LIB_CTX

Fixing this could be breaking. A partial fix might be that loading the config into
the NULL libctx sets the global default values. Otherwise loading config into a
specific libctx will apply onto to that libctx.

The partial fix is still not perfect.

The partial fix is the preferred approach and should be applied to 3.0. We can
consider changing this for 4.0.

** Review draft technical policies

Nothing to review

** Agree agenda for next meeting

    * Nominate a minute taker and confirm agenda
    * Agree minutes from previous meeting
    * Review outstanding actions
    * Review outstanding security issues
    * RIPEMD160
    * Nominations for committer
    * EC compression format compatibility discussion
    * Review issues with OTC hold (30 minutes)
    * Review draft technical policies
    * Agree agenda for next meeting
    * AOB

** AOB

None

** New and carried over action items [transported from above]:

AI: Richard to create instructions on how to add an OTC member.

AI: Martin to propose a plan for location and timing of the OTC F2F meeting.

AI: Matt to create a proposal on return values and return value types for new
code.

AI: Martin to propose some metrics on outstanding GH PRs and issues.

AI: Martin to identify a resource to write a proposal for an improved design
process

AI: Matt to create issue for the OTC vote on PR#17984