Index ¦ Archives ¦ Atom

TC Report 18

Intro

Feedback from last week's first attempt at a weekly overview of TC activity was positive enough to continue. Suggestions on how to make it more useful welcome. Main change this time is that I've added some information on stuff happened outside the meeting, a link to the meeting minutes, and a section on stuff we talked about last week that we said we'd pick up later but haven't yet.

Prior to the Meeting

Communications

Shortly after producing the first version of this newsletter last week I was approach by Flavio who reminded me that there has been a communications working group for the TC that had plans to provide regular updates on the state of things TC. We agreed that such a thing should still happen, that I ought to be involved, but that I would probably still to do this, so that I could editorialize freely if I wished.

Then in discussion of dropping the regular meetings2 it came up that if we do that it will be very important to have plenty of structured communication to the mailing list to replace some of the cadence marking that the meeting provides, something the TC chair might provide . As I'm typing this, this week's meeting has started and it is clear (by the immediate rush of discussion) the topic is of dropping the meetings is going to a big deal and if followed through will be a significant shakeup to how interactions happen within the TC and between the TC and everyone else. It is especially important for people who are not on the TC but want to interact with it regularly in a conversational way. If you have thoughts on this, read and respond to2.

Draft Vision for the TC1

This is mostly sitting idle, awaiting feedback from sessions in Boston. The idea is to use the vision to establish some goals for the TC and OpenStack. If you're interested or invested in that future, your feedback is important (on the review, in email, in the survey that was sent out, or in the sessions next week). You might look at what's there and think "what is all this fantasizing?". If that's your reaction you should say so, and say what you think should be talked about instead. Or you might love it. If you do, you could say why. That would be useful.

This Week's Meeting

Minutes and Log: http://eavesdrop.openstack.org/meetings/tc/2017/tc.2017-05-02-20.01.html

Killing the Meeting2

As mentioned above we leapt right into talking about killing the meeting. Wide variety of opinions on what function the meeting is providing in the first place, thus a broad selection of suggestions on what can be done instead of the meeting to serve those functions. Eventually we realized we weren't getting anywhere and there was a motion to move the discussion to email because:

  • it would be visible there to everyone
  • gerrit is a poor medium for exploratory or expansive discussion
  • email can be digested at whatever pace the reader requires

There were some ideas on how to make sure a thread moves forward to a conclusion. A regular summary and reality check of "is this where we are" every small number of days is a useful idea.

Not Making Decisions Synchronously3

This is related to killing the meetings; the idea is that making decisions synchronously excludes everyone who cannot be there at that specific moment in time or who cannot digest the language quickly enough to participate at full speed in a synchronous environment. There's some confusion over whether this should be a goal for just the TC or the entire OpenStack community. We eventually had to punt on this because we didn't really know. The conversation will move to the review.

(In my observations of the TC for the past couple of years, this is a common pattern. There's often lack of clarity on intent of a resolution or other proposal. What are the real problems it is trying to address, or the environments it is trying to create? People have very different interpretations and when it gets difficult or unclear, rather than reaching the bottom of the difference, the conversation is shifted to another time or medium. Often this is due to time constraints, but frequently the topic is never rejoined so incomplete understandings accumulate in a massive pile.)

Change the target for this goal to uWSGI not Apache mod_wsgi4

General agreement about doing this change, not a big deal, but some concern about changing a cycle goal in the middle of the cycle ("moving the goal posts"). Agreement was that changing details of implementation are not the same thing as changing the goal (especially when it is a simplification) so it is okay as long as the change is reflected in the doc, not just the git history.

More on maintenance-mode

There's a newish tag called status:maintenance-mode which means that a project is receiving limited attention for a period of time. There's a proposal5 that the TC should become core on such a project to make sure there are people to handle urgent matters. The question is whether this is necessary since:

  • the TC can get those privileges at any time on any project when there is an urgent matter
  • being in maintenance-mode is supposed to mean there is sufficient attention from project team members for urgent matters, if not the project is abandoned

This turned out more contentious and confused than expected and it too was punted to the review5.

Open Discussion

The above filled pretty much the whole hour, suggesting that perhaps we all have a lot more to say to one another than a single hour allows. That was acknowledged and suggestions were made that we really need to use email more and better, even though it can be challenging. That is, we need to level up our email skills.

To help ensure more talking to one another and do a bit of near-term planning, a TC gathering will happen in Boston late next week. Evidently I will be spit-balling, and no one will be sitting near me.

Dropped Stuff

A section of reminders of things we said we'd talk about more but haven't yet.

OpenStack moving too fast and too slow

At last week's meeting, while discussing the findings from the user survey there was discussion6 of

the complicated problem of OpenStack moving both too fast and too slow at the same time, depending on who was looking. And the difficulty with lack of centralized control over the technical direction of OpenStack and (probably most importantly) the application of resources.

that was supposed to move the mailing list7. As far as I can see it did not. dfisher have you got the cycles to pick that up again here on the list? Or if not you, maybe mordred, fungi or dhellman? If it was already discussed, my apologies for losing it, can someone point it out to me?

Colophon

This is an opinionated overview of Technical Committee activity from my perspective. As such it is subjective and potentially wrong enough to cause disagreements. That's a good thing if it leads to discussions that make things better or more correct.

© Chris Dent. Built using Pelican. Theme by Giulio Fidente on github.