GenAI SIG: Update alternate meeting to 16:00 PST #2437
Conversation
drewby
requested review from
alolita,
austinlparker,
danielgblanco,
jpkrohling,
mtwo,
mx-psi,
svrnm,
tedsuo and
trask
as code owners
There was a problem hiding this comment.
While I'm personally ok with this, I feel it helpful to recap some points last discussion about timezones which led to this time in SIGs such as this.
- folks with family duties in UTC+8 may struggle with this time
- Having not in UTC+8 will likely lead to folks in the timezone missing due to misunderstanding US timezone impact
- This makes places like india quite late or early
| @@ -126,7 +126,7 @@ Other languages are not in the scope (yet). | |||
| We'll use GenAI Semantic Convention meetings until more suitable time for new participants is determined: | |||
|
|
|||
| - **Every other week**: Thursday at 10 AM PST | |||
| - **Every other week**: Thursday 14:00 UTC+8 | |||
| - **Every other week**: Thursday at 4 PM PST | |||
There was a problem hiding this comment.
can this stay UTC+8 so we don't need to worry about timezones?
Yes, this was much the motivation to move the meeting the first time, but we did not get the attendance. I'm OK to change to UTC time but wanted to be consistent with what was already there. Should we change both? |
One suggestion I had when this first started was to give more time to allow things to settle before canceling again (some SIG cancel then resume). I hinted to someone I forget, at least 6 months. Personally, I wouldn't propose this change or 👍 it, at best +0. We are raising awareness of a new group composure, and a new product. This has a good chance of drawing in more people, and ratcheting back the timezones will add barriers to entry. The other thing, is while I do agree we aren't there, yet, we do want a group to eventually be more diverse globally. When all meetings have to be US timezone friendly, it counters that point. These are the same reasons for the original timezone layout.
I think how it was before, one in pacific one in UTC+8 is fine as it was about the audiences involved. Mainly not anchoring UTC+8 to US timezone related things. If folks outside the US want the former normalized to UTC, that's ok with me. |
|
Discussed at the GenAI SIG call:
@patcher9 and @AmanAgarwal041 we'd like to check with you folks - it seems this change would be the most problematic for you folks. I wonder what your thoughts are and if you're planning to attend current APAC call (14:00 UTC+8). I'll also post in the GenAI instrumentation channel. |
|
I wonder if we are being too narrow with the impact of this, thinking mostly about india and openlit specifically. Doesn't this also affect israel and europe? While participation has indeed been sporadic, it seems we should consider if this closes the SIG to folks who we should be pulling in more often. Many folks I work with are in europe and I'm a little surprised no feedback on this. Thoughts? |
|
Adding my 2 cents from Elastic: we are looking to increase our attendance at the SIG meeting and many of us are in Europe. So we would very much appreciate a meeting at a EU-friendly time. Thanks! |
Given that we have plenty of active participants from Europe, I'd prefer to focus on extra slot in Europe-friendly time. Proposed slot is 1am CET (4pm PDT) is impossible to attend in Europe. Our current slots might, however, work for some contributors there:
@estolfo do you think either of those can work? If not, I wonder if we can find the slot that works for most people in Europe - other OTel projects use 4pm-7pm CET (7am-10am PDT) which mostly work for the US too, but there are a lot of conflicts for folks who participate in multiple otel projects. 4pm CET/7am PDT would probably be the least crowded. |
|
The current time is 14:00 UTC+8 which is 7am berlin time, and this still allows folks in UTC+8 to attend. I think what was suggested means an extra time slot for europe, so wouldn't necessarily cut out anyone (like me), but also allow existing folks to attend.
The extra slot was been tried by @mx-psi and @jpkrohling who would be able to comment on this. |
|
Aside on quorum especially during busy periods, and what quorum means. Personally, I mentioned that I had wasn't able to attend meetings for nearly a month (due to medical which also affected my normal job). I also mentioned that many I knew were tied up with kubecon related busy stuff which not only affected meetings but other sorts of sig participation. I saw chat negotiation with OpenLit who are in india, who were ok being unable to attend anymore due to a switch to earlier UTC+8. I'm still troubled about cutting out India, so prefer a way to keep it possible even if OpenLit are ok without attending meetings. Despite that, if we look at the logs, folks have been joining in all existing timezones. https://docs.google.com/document/d/1EKIeDgBGXQPGehUigIRLwAUpRGa7-1kXB736EaYuJ2M/edit?tab=t.0#heading=h.ylazl6464n0c However, if this is about spec approvers, it is a different topic than willing participants. Maybe we can focus on what we feel is a failure at the moment. While in python and javascript there are diverse approvers, there are less in the main semver spec SIG. Maybe we can work at getting more spec approvers who are in these timezones, and/or set a goal of it. @karthikscale3 was added recently, and that's awesome, but he's also west coast canada. Regardless of what we do, let's not cut out UTC+8 or make it hard for folks to join. I've been doing a lot of work in APAC to increase interest in otel, and sometimes I've not been able to attend meetings due to that, ironically. In fact, I'm doing yet another presentation on our SIG next week in singapore (last month it was kubeday melbourne). The old time frame was designed to allow participation by folks in UTC+8 who have family responsibilities in the morning. I don't want us to accidentally undermine community engagement work in APAC. |
|
updated above comments as I misunderstood the last suggestion. I think IIUC was to create another timeslot (so 3), which is far better than choosing which region to black hole. I also added notes about what quorum means, and personally believe we can find folks who can step up into approver role who aren't clustered in one region. |
|
In the Collector SIG, we've been having three slots for quite some time. The result is that most people are able to attend two thirds of the calls, meaning that we all know that SIG calls are more like "office hours" than a meeting where we decide things. Decisions are made via GitHub only, and what we discuss via SIG calls and Slack is summarized in GitHub as well. While I recognize that our APAC-centric slot has lower attendance, the fact that we have prominent contributors there (hi @MovieStoreGuy!) means that good discussion can happen there even if other maintainers can't join (ie, no quorum). |
|
@lmolkova 4pm CET/7am PDT would be perfect, if that's an option. |
We are moving the alternating meeting to 16:00 PST to get better attendance. The current time is not getting a quorum to make decisions.
@lmolkova @tedsuo @nirga