09:02:07 <ttx> #startmeeting tc
09:02:12 <openstack> Meeting started Tue Jun 12 09:02:07 2018 UTC and is due to finish in 60 minutes.  The chair is ttx. Information about MeetBot at http://wiki.debian.org/MeetBot.
09:02:13 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
09:02:16 <openstack> The meeting name has been set to 'tc'
09:02:17 <ttx> #chair cmurphy cdent
09:02:18 <openstack> Current chairs: cdent cmurphy ttx
09:03:12 <ttx> I think I've finally come to terms with Castellan vs. Barbican on that review
09:03:46 <cdent> smcginnis and fungi had a pretty extensive coversation here about it yesterday and it felt like it opened more questions
09:04:54 <ttx> I just commented on the review... I think one issue is that people see that document for more than it is
09:05:02 <ttx> It's not deployment guidance, it's developer guidance
09:05:17 <ttx> "You can develop features that rely on Castellan, it's ok"
09:05:55 <ttx> "that will inflict pain on operators, but it's worth it"
09:07:06 <ttx> It's not about dictating to deploy one thing or the other
09:07:22 <ttx> That comes after, through constellations, interop...
09:07:39 <ttx> which are tools targeted to deployers, not developers
09:08:11 <ttx> (At least that's how I solved the tension in my mind and ended up agreeing)
09:08:20 <cdent> in the eschaton the twain shall meet
09:09:05 <cdent> I rather think that as developers, at least ones that are listening well, we shoud create burdens on deployers, but make sure they are useful and visible. Otherwise we're stuck.
09:09:51 <ttx> right. I see the base services as guidance on which burdens are blessed
09:10:11 <ttx> (or deemed useful)
09:12:08 <cdent> I have this long running interal-to-my-brain discussion of collaboration and information sharing as somewhat religious: "blessed", "eschaton" (even "burdens")
09:12:43 <cdent> and effective leadership is basically making chosen truths visible
09:15:40 <ttx> It's still an interesting balance. On one side you want people to scratch their itches and avoid top-down mandates. On the other you need a body of shared understandings and common direction
09:16:17 <ttx> Striking the right balance there has been the story of the last 8 years of my life
09:16:33 <persia> Aside from needing context for reviews, what is the benefit to common direction?
09:17:07 <cdent> persia: I can think of a variety of things, but one big one is a sense of membership
09:17:29 <ttx> persia: I'd say that for any major change, it's hard to get started if you doubt that it's even a "wanted direction"
09:17:34 <cdent> (which of course demands "what is the benefit of a sense of membership?"
09:17:53 <ttx> We have a mechanism at project-team level to early check for that agreement
09:18:07 <persia> Identity is important for lots of reasons: I just hadn't thought that common direction drove identity (but I can see why).
09:18:24 <ttx> But for changes that span "OpenStack"... it's harder to get that sense of working on a desirable direction
09:18:57 <ttx> and as individual components mature, we get more and more of those "openstack" spanning things
09:19:06 <persia> Validation is a cultural element, but widely shared.  Hard to fix without strong access to all the educational ministries in the world.
09:22:36 <ttx> In other news... the discussion on diversity tags went a bit in every direction, trying to summarize it now
09:22:45 <cdent> Another thing I like about common direction is that it acts as counter to the unfettered (some would even say cancerous) growth that OpenStack experienced a few years ago (2014-15?). Then direction was pushing in many ways, creating a large balloon, but not going anywhere in particular.
09:23:18 <cdent> To go somewhere, the force vectors need some alignment.
09:23:34 <cdent> Yeah, the diversity thing did go many ways.
09:23:39 <cdent> Metaphor
09:23:51 <ttx> zing
09:25:51 <cdent> persia: since you've expressed opinions in this area before, I have an issue you might be able to help with: Now that these office hours are being explicitly logged and dhellmann has changed the tone of the official tc update to be a bit more verbose, I'm struggling to find my voice for my tc reports. Any desires or suggestions?
09:26:24 <cdent> Part of the problem is that because of travel I haven't done one in a long time, and inertia. But there's more than that.
09:28:59 <persia> cdent: The value of amanuensis vs. logs is the process by which the amanuesis considers the input.  Be less objective: your further speculations on the discussion and/or sharing your opinions will help others in the community who are not as closely involved feel emotional connection to the logs, if they peruse them.  For those who do not intend to peruse logs, it provides a counterpoint to the verbose update, so that those not involved can more
09:28:59 <persia> easily see consensus vs, lack of consensus.
09:29:30 <persia> Mind you, in the event that you and dhellmann agree firmly and everyone else is uncertain, the burden falls on others to ensure the breadth of dialog is preserved in the discussion.
09:30:22 <cdent> Oh, good, a) you used the word amanuensis, which I love, b) what you describe is what I was thinking too: subjectively observe where consensus is lackign
09:30:53 <persia> Also where consensus is present, to help observers appreciate the nature of the consensus (removing suprise when the decision is reached).
09:33:47 * ttx reaches out for the Merriam-Webster
09:34:11 <persia> ttx: amanuensis: a human who records things (vs. a machine).
09:34:55 <cdent> the simple definitions tend to miss out on some of the context and the implications about creating meaning out of chaos
09:35:01 <persia> Used to be a synonym for secretary, but as a result of careful presentation by a couple authors in the past decade or so, has come to specifically mean inaccurate/human processed recording.
09:35:13 * ttx returns to summarizing threads like an amanuensis
09:36:18 <persia> ttx: Your thread summaries are some of the best examples of amanuensis I could cite.  Thanks for all those in the past, and any you may choose to offer in the future.
09:38:05 <cdent> back in the blue oxen days the amanuensis was a key role in (/me takes a breath) our model for high performance asynchronous collaboration
09:38:28 <cdent> as a sort of bridge between instances of engagement
09:38:39 <persia> cdent: As a means to reduce the time required to follow the many separate chains of thought that represent a collection of IRC channels?
09:39:31 <cdent> it was more about identifying/sustaining a chain through multiple media
09:40:54 <cdent> notably things like "at the end of last week we never resolved thread A, here are the things still hanging"
09:41:04 <cdent> s/media/media over time/
09:41:44 <cdent> persia: http://eekim.com/2007/02/the-blue-oxen-way/
09:42:13 <cdent> (sadly a lot of dead links in there)
09:42:31 * persia may not have been reading the right books, as that makes the claim of "decade" above clearly not a sufficient span
09:46:40 <ttx> mnaser: I just published a "diversity tag" thread summary on the list. I still think a per-cycle report is the right way to convey that complex topic -- could be crowdsourced through the TC team liaisons, each contributing a paragraph to describe current status in that team
09:48:50 <cdent> persia: the backing "ethic" of much of blue oxen is in http://peermore.com/astool.html
09:50:17 <persia> computation as augmentation vs. computation as automation ends up being more interesting than I could have imagined :)
09:58:14 <cdent> we seem to have reached the end of the office hour
10:01:29 * ttx closes hour
10:01:33 <ttx> #endmeeting