How are issues picked?

[gundalow]

Hi, Just found out about this project. Looks very interesting. I hope this is the right place to have a discussion.

I'm responsible for Contributor Experience at Ansible and interested in there are ways for us to use codetriage better for our Issues & PR backlog

1) How are issues current allocated?

• Are the same issues sent to multiple people?
• Is their any preference on new vs old issues?

any way to add per-repo logic to only allocate from a subset of issues, say based on filters?

2) How active are subscribers

I see on https://www.codetriage.com/ansible/ansible there are 203 subscribers. Any details on how active they are?

3) Success stories

Any details on how other projects have used this or made improvements to their process?

[gundalow]

In particular for Ansible as we have a lot of PRs and issues I think it would be useful to direct people to newer PRs that are passing CI (we have GitHub labels for this). Would be interested in your thoughts @schneems

[gundalow]

What if love to do to is tweak which issues/PRs are sent out to people so we can get more contributions back that we can merge sooner.

[schneems]

All issue emails are random. And this is a distributed computing case where it's really hard to beat "random".

One suggestion that i'm open to is to start ingesting github labels and "bless" some of them as "send these before anything else". So if an issue is labeled something like "good for beginners" or "help wanted" etc. then we'll send them that first. Or maybe if they have multiple issues we pick one from that list and then another at random.

Even with "just random" selection criteria, that section of the code is already fairly complicated. We need to be careful that we don't back our users into a corner, for example, if a project has a lot of followers but only a few "help wanted" tags, then we want to be mindful that we don't send all thousand+ followers to just a handful of issues every time. If every issue you get already has 20 responses by helpers then it would seem like your efforts are redundant and not needed when in fact there are tons of issues that need just as much help, but they're not prioritized.

[LouiseBonnard]

Hi @schneems, thanks for the explanation, that's interesting! I'd like to ask you a few questions to know how we can work together, could you please reach me at [email protected]? Many thanks in advance!