search

dask / community

For general discussion and community planning. Discussion issues welcome.
20 stars 3 forks source link

Live coding events #75

Closed jsignell closed 3 years ago

jsignell commented 4 years ago

There was some discussion at the end of the dask dev meeting yesterday about the SciPy sprints. It was agreed that the turn out wasn't great and it was hard to retain people virtually. We talked about possibly having some more live coding events to offset the inperson sprints that we are missing during covid. This could look like:

  1. Maintainer sits in a room and people can drop by to ask questions and get advice (would it be ok to use the whereby room for this?). Possibly with a page of some tasks that it'd be good for relative new-comers to tackle.
  2. @jacobtomlinson style code stream where a maintainer picks an issue and streams themselves trying to fix it.

I think this kind of thing would fit well into the evangelism work that has been happening recently with the tutorials.

martindurant commented 4 years ago

It was agreed that the turn out wasn't great

I would rather say, it wasn't clear what our expectations were, nor how much work done and people introduced to the project its worth our developers' time.

jcrist commented 4 years ago

I would rather say, it wasn't clear what our expectations were, nor how much work done and people introduced to the project its worth our developers' time.

As one of two core devs that showed up (the other being Mike McCarty, AFAIK), I thought it went ok. Got a handful of people introduced to contributing, and talked a bit. A few PRs were even submitted and merged. This was on par with my experience with past SciPy sprints - I'd usually come prepared with plenty of beginner issues, but would instead end up helping people with their pet issues or getting people set up with dev environments on windows boxes. I haven't found sprints to result in a meaningful increase in long-term contributors, but they're still nice to meet and help new people start figuring out how to contribute via github.

mrocklin commented 4 years ago

In general I'm +1 on us trying new things. I expect most of them to sort of flop, but maybe some of them will actually be great.

I like the idea of option 1, which sounds like of like office hours from university.

On Wed, Jul 15, 2020 at 9:37 AM Jim Crist-Harif notifications@github.com wrote:

I would rather say, it wasn't clear what our expectations were, nor how much work done and people introduced to the project its worth our developers' time.

As one of two core devs that showed up (the other being Mike McCarty, AFAIK), I thought it went ok. Got a handful of people introduced to contributing, and talked a bit. A few PRs were even submitted and merged. This was on par with my experience with past SciPy sprints - I'd usually come prepared with plenty of beginner issues, but would instead end up helping people with their pet issues or getting people set up with dev environments on windows boxes. I haven't found sprints to result in a meaningful increase in long-term contributors, but they're still nice to meet and help new people start figuring out how to contribute via github.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/dask/community/issues/75#issuecomment-658870084, or unsubscribe https://github.com/notifications/unsubscribe-auth/AACKZTGDJLM4Q7TP25Z3GBDR3XLLLANCNFSM4O2UB6XA .

thomasjpfan commented 4 years ago

The last three three virtual scikit learn sprints, including Scipy 2020, were held on discord. For the two sprints I was involved in, we had attendees assigned to virtual “tables” in pairs, where the primary form of communicate was through voice. These tables also allowed attendees to share their screen with each other. The maintainers would sit in a “help desk” voice channel where attendees can join and ask questions. Usually this involves one of the maintainers joining to the attendees’ table to help with their specific issue.