[tor-bugs] #29211 [Core Tor/Tor]: Distribute config.c functionality across more modules

[Tor Bug Tracker & Wiki]

#29211: Distribute config.c functionality across more modules
-----------------------------------------+---------------------------------
 Reporter:  nickm                        |          Owner:  nickm
     Type:  task                         |         Status:  assigned
 Priority:  Medium                       |      Milestone:  Tor:
                                         |  0.4.2.x-final
Component:  Core Tor/Tor                 |        Version:
 Severity:  Normal                       |     Resolution:
 Keywords:  network-team-roadmap-august  |  Actual Points:
Parent ID:                               |         Points:  23
 Reviewer:                               |        Sponsor:  Sponsor31-can
-----------------------------------------+---------------------------------

Comment (by teor):

 Hi Nick,

 Here are my initial thoughts on the branches I reviewed.
 I want to think about them a bit more, and review and revise them before
 the sponsor 31 meeting on Thursday.
 I'd like a response at that meeting, rather than on this ticket.
 But I wanted to give you time to think through them before the meeting.

 The goal of sponsor 31 is to improve code quality.
 But I worry that this pull request series has not improved our code
 quality (yet!), team processes, or coding habits.

 I wanted more:
 * collaboration on the design,
 * summary and detailed design documentation,
 * opportunities to write code for large refactors myself,
 * function documentation and callback documentation,
 * code comments around tricky code (or less tricky code),
 * automated code transform scripts (sed or coccinelle),
 * unit tests,
 * tests for typical options used by Tor Browser, and possibly other
 significant uses of tor, (unit tests, chutney, or stem),
 * working stem tests in CI (rather than stem CI having allow failures due
 to #29437).

 I wanted less:
 * CI failures,
 * regressions (#31495, #31527),
 * technical debt,
 * missing changes files,
 * pull request merge conflicts / target branch mistakes (which make
 reviews difficult),
 * pull requests in the review backlog,
 * dependencies between commits and pull requests (which make fixes
 difficult),
 * exceptions to the normal review rules and review processes,
 * pressure to complete reviews by a deadline.

 Overall it seems that this process has been difficult for all of us.
 I would like us to pause for a while, and work through the tensions that
 this refactor has brought up.

 I have some ideas about the root causes here.

 But I would like to hear your ideas at the sponsor 31 meeting on Thursday.

--
Ticket URL: <https://trac.torproject.org/projects/tor/ticket/29211#comment:8>
Tor Bug Tracker & Wiki <https://trac.torproject.org/>
The Tor Project: anonymity online