search

2i2c-org / team-compass

Organizational strategy, structure, policy, and practices across 2i2c.
https://compass.2i2c.org
4 stars 13 forks source link

Add community (formerly "upstream") projects to the Initiatives planning process #829

Closed aprilmj closed 4 months ago

aprilmj commented 7 months ago

Context

This issue tracks the implementation of #820 (proposal for including upstream work in initiatives and delivery planning).

Updates and actions

We want to have an initiative on the Initiatives Board for all upstream projects/contributions that 2i2c people are working on collectively. For now, that means making sure everything on this list is tracked as an initiative:

Grants

Pending grants

Potential grants?

Major community partnerships

Once added, we'll also need to

choldgraf commented 7 months ago

Just looking at the list that's here and wanted to share a general thought (non-blocking on this issue, but if it's important to discuss now I'm happy to do so).

I suspect that we'll have to grapple with the question of "how to introduce a WIP limit given that we have 9 upstream/external initiatives". I don't think we're there yet, but eventually we'll want to limit the total number of in-flight initiatives so that we're not doing too much at the same time.

We might also want to define our own initiatives that feed into multiple upstream projects (e.g. "identify a thing that 4 projects want, and create an initiative that will check "done" boxes in multiple projects when it is complete)

aprilmj commented 7 months ago

@choldgraf Good questions! I suggest capturing the list first, then asking ourselves what's possible or reasonable when we're planning. 9 is a lot of initiatives to have in progress, if people are actively contributing to all of them!

How do y'all identify and track work that serves multiple communities' needs today?

consideRatio commented 7 months ago

When I think about "upstream", I believe what I'm thinking of is open source projects used by several separate parties that 2i2c or our communities depend on. In this, I'd list for example jupyterlab, jupyter_server, jupyterhub, dask, ipywidgets, nbgitpuller, grafana, etc.

Is perhaps "community projects" a better fit of how to speak about these than "upstream projects"?

aprilmj commented 7 months ago

Nice suggestion @consideRatio! It started as "upstream leadership", but I like "community" better as a category.

choldgraf commented 7 months ago

+1 to that terminology as well, thanks for thinking it through @consideRatio

Also is a more inclusive terminology, since something like Jupyter For Health is more like a "community (we are serving) project" rather than a "community (we use technology from) project". I'd classify the latter as "upstream", but not the former. So "Community projects" feels like it captures both.

9 is a lot of initiatives to have in progress, if people are actively contributing to all of them!

Ah I just realized you said this, I added a similar comment to the issue you opened :-) I think it's safe to just start trying and see where it goes.

How do y'all identify and track work that serves multiple communities' needs today?

I'd say we're in the middle of building out these systems now - definitely related to the work @Gman0909 is doing, and I think this initiatives process ties into it as well

jmunroe commented 6 months ago

9 is a lot of initiatives to have in progress, if people are actively contributing to all of them!

This tells me we need a way to distinguish between an initiative which is work-in-process for the 2i2c team as distinct from a project that is longer lived and may have significant periods of time that do not require active participation from 2i2c. An objective of the the initiatives board to me is to capture the initiatives we already have as work-in-process and use that to decide if we have capacity to start work on a new initiatives.

There is a longish list of project and communities that we provide support for, we have received money for, we need to be able to assist with when called upon, but are not 'initiatives' in the sense of something we are intentional about allocating our resources towards. There are also upstream projects that we are also highly involved with (jupyterlab, jupyter_server, jupyterhub, dask, ipywidgets, nbgitpuller, ...) but none of those 'projects' can ever be considered "done" so I am not sure how to represent them as initiatives.

I agree that this is best figured out by adding some 'initiatives' to this board and using them as specific examples of what we are taking about.

For example, I am at this moment looking 'Project Pythia'. That 'project' is long-lived and hopefully will continue far into the future with or without 2i2c direct involvement. So 'Project Pythia' is not a 2i2c initiative. We are co-investigators on an NSF OSE GEO grant to fund and support 'Project Pythia'. The grant lasts three years and not all of the deliverables under the grant dependent only on 2i2c. The grant is not an 'initiative' of 2i2c either. (If it was, it will sit on this board for three years as WIP -- I don't think that is useful). There are some particular deliverables (some explicit and some implicit) for 2i2c under the Project Pythia grant:

This is not a comprehensive list but some of things rattling around in my brain under the concept 'Project Pythia' that I could see being adopted as initiatives for 2i2c. There is a much longer list of deliverables for other project partners on this grant.

Some of those 'initiatives' are more refined and ready for action. Others are require refinement. Are these the kind of the 'initiatives' we should be putting on this board?

aprilmj commented 6 months ago

Some of those 'initiatives' are more refined and ready for action. Others are require refinement. Are these the kind of the 'initiatives' we should be putting on this board?

I think we should try it and see. Seems like the first thing to try would be the "ready for action" initiatives, since those could result in action in the next sprint, if prioritized. One thing we might realize is that Pythia, for instance, is actually a group of things that need to be broken down (maybe unified by tagging and/or titling the initiatives to make it obvious they're part of the umbrella of Pythia?).

jmunroe commented 6 months ago

I've updated the top comment to group different types of community projects by 'type'. The list of types is not complete or finalized.

colliand commented 5 months ago

I think this can be closed. The Projects have been transferred from meta into AirTable-backed issues in leads. I will remove other assignees and leave @aprilmj as the shepherd with the power to close the issue if deemed ready.

aprilmj commented 4 months ago

Closing this issue with an assumption that we can revisit the entire initiatives planning process post-July and will be working differently until then.