Repository navigation
Implement "lifetime contribution" approach to credit #135
Description
This is recording for view in this issue tracker a message from @taldcroft on the mailing list a while back:
In the last coordination meeting and in previous years, we have all agreed on the importance of giving appropriate credit to the people in the trenches that make Astropy really work. Unfortunately we never got much past that point, often getting stuck on implementation of how to define and measure contributions.
I have a proposal now that involves nothing fancy and can be done very quickly with a single static HTML page and a google survey. The (hastily created) mock-up web page below should basically explain most of it.
The key new idea is to just measure "contribution" with self-reported hours working on Astropy-related tasks. I think this is reasonable and basically the only way forward. Any automated / programmatic way of measuring or recording contributions is going to fail in a wide variety of ways (which we have spent a lot of time already discussing with no useful outcome). I believe most people are basically honest, but of course there would be some sanity checks to the responses at the point of updating the web page each year.
The self-reported hours would be based on responses to the semi-annual roles survey, with two new questions:
- About how much time per week did you spend on Astropy-related tasks over the last 6 months?
- Less than 2 hrs/week
- 2 hrs/week
- 4 hrs/week
- 1 day/week
- 2 or more days/week
- Do you feel your lifetime contribution to Astropy is more than 2000 hours (roughly one year working full time)? You can think about a rough percentage of time you spend and how long you have been involved with Astropy. If so and you are not already on the list, then congratulations and thank you!! Please write one or two sentences describing the work you have done for the project so we can include it in the Lifetime Contributors section of the Credits web page.
I'm interested in feedback! With the idea of doing something, even if it is not perfect.
Activity
eteq commentedon Sep 16, 2020
@mhvk added:
Hi Tom,
I like the idea. I agree to rely on people's honesty - especially since
for larger contributions we can typically know/check.
All the best,
Marten
eteq commentedon Sep 16, 2020
@astrofrog added:
I agree - this is a great idea!
If we do this, we should probably align the roles survey with the
calendar year? (e.g. one survey at the end of June, and one at the end
of the year?)
eteq commentedon Sep 16, 2020
@hamogu added:
I like this idea, too. I agree with self-reporting and the pretty broad
catagories that Tom suggests (factor 2 or so) mean that cheating by a
bit does not matter, and we would probably notice if someone cheats by
several catagories. Of course, we never know extactly how much time goes
into a PR anyway, so such a broad list is good. Sometimes there are just
a lot of abandonned ideas that never made it into a git commit, but
still need developer time.
eteq commentedon Sep 16, 2020
Then after that, it was discussed in the governance WG, and the eventual decision was to not include it as a formal mechanism in APE0 (since an alternative for the initial voting membership was devised), but rather would be a great thing to have under the purvue of a role or committee that worked on the broader problem of credit.
eteq commentedon Sep 16, 2020
Now I'll add my opinion, which is that I like this idea too! I think the one caveat I have is making sure the mechnism for keeping this up to date is something that doesn't depend on a sole maintainer to keep an eye on it, but I'm confident that can be worked out if we agree on the principle here.
eteq commentedon Sep 16, 2020
Not also that #35 is related to this in that that might be a way to not have to roll our own page. But I haven't looked closely enough to be sure that it's not more trouble than it's worth to keep track of @taldcroft's original criteria using allcontributors
mhvk commentedon Sep 16, 2020
👍 to just going ahead with this, since we send out those e-mails anyway. A bot may well help in the end, but it seems more effort and would need a cross-check with something like what is proposed here anyway.
taldcroft commentedon Sep 22, 2020
The CoCo has talked about this and will go forward with this idea shortly, starting with the next roles survey that will go out in the next few weeks.
hamogu commentedon Feb 18, 2022
We've been doing this for a while, see https://www.astropy.org/credits.html so this issue can be closed.