All TLJH issues triaged!

Since I had some other important work to do, I procrastinated instead by triaging all issues on the littlest JupyterHub repo! I closed a bunch, fixed a bunch and responded to almost everything.✓&q=is%3Aissue+is%3Aopen+no%3Alabel is empty, showing 0 unlabeled issues. Makes me feel great. Should make it easy to keep up with new issues now.

This was inspired by Steve Klabnik’s post about open source gardening - Also by @mrocklin’s Highly recommended.

Would love for more people to help triage and work on stuff in TLJH. :heart:


Big effort!! Also thanks for doing the same to the repo2docker issues!!

What is your feeling for the trade off between spending time triaging issues and dealing with them then and there?

In my mind triaging is for when there are several people on the project and one person can triage things and that saves time for the rest of the team as they look at their speciality. For small teams where it is more “everyone can/has to tackle everything” it seems not so useful to spend my time triaging because I could just reply then and there (it is highly unlikely that someone else will look). I’ve been wondering however if it is because we don’t triage things that we don’t increase the size of the team of people who items could be redirected to.

One thing I’ve pondered while reflecting on how many notifications we all get and people declaring “bankruptcy on GitHub notifications”: should we strive to only trigger a notification if the person who receives it has a clear next step when they read it? If we could get really good at that then every time there is a notification and you decide to act on it there is a clear thing for you to do. Often it feels like we generate a notification and the people who react to it are greeted with a “yeah well, what am I meant to do here” state. Pondering how/if we can use labels to help with that. Maybe we should collect a list of the next steps that are frequently used and then have a system for “the next step here is step #1234 from our list of next steps”

Stream of thoughts over :slight_smile:

I’m a big fan of triaging in general - I find that labels are helpful, and making sure there’s an actionable next step written out in the issue is great as well. The other benefit of triaging is that it helps you

  1. Get a global view of activity in the repository, assuming you aren’t combing through each new issue in real-time as it comes in
  2. Prioritize the issues that are most urgent or that you are particularly useful in helping with. Responding to issues one-by-one means you might spend a bunch of time on a less-important issue and miss a more important one

My goal is to spend an hour or so triaging each week, and the rest of my “issues time” actually responding etc.

Also thanks yuvi for the triage effort! Much appreciated :slight_smile:

This is gonna vary from person to person. In this case, if I could provide a useful, considerate response in under 5minutes, I responded to it (and labeled it). If I could make a PR in under 15minutes, I did that too.

I partially did this was to make it easier for myself. Now if I feel like working on documentation, I can filter and pick one! Same with bug squashing, etc. I’m sure this works for other people too.

It also makes me feel less prone to having to declare GitHub bankruptcy. A lot of times I don’t want to respond to an issue that’s been stale for a while simply because it’s been too long and now I feel super awkward (see Now I feel I can triage, and provide a welcoming message pretty soon after an issue lands. I think this will be good for me personally and good for the projects / communities in general too.

1 Like

I think that sounds like a cool rule to use :+1:

Thank you for the link to OS management. It was very insightful. Thanks for your triage!

I was surprised by a slew of comments in my inbox. It inspires me to try to commit more time to this project.

You have been a model of support for me. That said, I have been pretty successful at providing local services in our internal network.

Thank you for your personal support at JupyterCon 2018 and my efforts since then. You, and others in, have been very rewarding!

1 Like