Help with open issues on github.

[Daniel Lee]

Hi all,

I could really use your help. There are way too many open issues on our repos and I really want it to be cut down to something reasonable.

Here is our current status:

- Stan: 218 open issues. https://github.com/stan-dev/stan/issues
- Stan Math: 112 open issues. https://github.com/stan-dev/math/issues

If you have any time to spare, please help out. You don't have to be coder. You can do this with any 15 mins.

1. Go to either the Stan issues (https://github.com/stan-dev/stan/issues) or Math issues (https://github.com/stan-dev/math/issues) on GitHub. Log in.

2. Go to any page of the issues (preferably something older, but anywhere is good). Click on any issue. It'll look like this: https://github.com/stan-dev/math/issues/61

3. Take some action:

  - Read the issue.

  - If it's irrelevant now, please close the issue! That's the best thing to do.

  - If it's still relevant, add tags to the issue if there aren't any. Write a comment that you've seen it. Assign someone. Do something with it. Ask a question. In other words, don't just read it and let it go.

Thanks. Please help out if you have a few minutes. It would make me feel a lot better about the stability of our project if the number of issues were manageable.

Daniel

[Jonah Gabry]

I'll do some right now.

[Jonah Gabry]

I don't know if a lot of these can be closed, but I can at least make comments asking the people involved in the issue if it can be closed.

[Daniel Lee]

Thanks. Most of the ones you're asking about can be closed. We should really close things unless they have a reason to be open.

Daniel

On Fri, Mar 4, 2016 at 1:15 PM, Jonah Gabry <jga...@gmail.com> wrote:

I don't know if a lot of these can be closed, but I can at least make comments asking the people involved in the issue if it can be closed.

--
You received this message because you are subscribed to the Google Groups "stan development mailing list" group.
To unsubscribe from this group and stop receiving emails from it, send an email to stan-dev+u...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Jonah Gabry]

Ok, if nobody responds today saying to keep them open I'll go back through and close those ones I asked about.

[Daniel Lee]

Thank you.

On Fri, Mar 4, 2016 at 1:38 PM, Jonah Gabry <jga...@gmail.com> wrote:

Ok, if nobody responds today saying to keep them open I'll go back through and close those ones I asked about.

[Daniel Lee]

Thanks, all. Seeing all the closed issues is making me happy. 

If you need to move issues between repos, this is useful. https://github-issue-mover.appspot.com

[Bob Carpenter]

Usually this kind of flurry of activity has your name
attached to almost every post. Nice to see everyone
jumping in here.

- Bob

[Michael Betancourt]

Sorry, a little bit late to the game.  What do we want to do with long term

goals that were something of a replacement for our massive todo list?

I can understand if we want issues to be “short term” to avoid clutter,

but I would prefer to have a “future” tag so we can keep these long

term goals without cluttering up the active issues.  Thoughts? 

[Bob Carpenter]

We have the "future" milestone already.

I think we should keep the longer-term goal wiki.

I'd very much like to restrict the issue tracker to:

* concrete feature requests with a functional spec
in the form of a basic interface design

* bug reports (should these be required to be reproducible
before allowed to stand?)

That will keep the issues narrowly scoped and still allow us
to have bigger design discussions on the outside.

- Bob

[Jonah Sol Gabry]

On Saturday, March 5, 2016, Bob Carpenter <ca...@alias-i.com> wrote:

  * bug reports (should these be required to be reproducible
    before allowed to stand?)

Ideally yes, but I can imagine a user stumbling on a legit bug that they themselves can't reproduce consistently. But maybe then the right place for that is a question on Stan-users so that together we can find a way to reproduce it and then move it to the issue tracker.

[Michael Betancourt]

Wikis have proven prone to falling out of date, so I’ll add
a vote for using the “future” tag to maintain long-range
goals. But I’m guessing I’ll be outvoted and won’t make
a fight.

[Daniel Lee]

An alternative could be to create a separate repo where we use the issue tracker. We can migrate issues when they become fully formed.


Daniel

[Allen B. Riddell]

I think the solution here is to mark as outdated (or delete) any wiki
pages which are out of date. Github doesn't make this kind of
housekeeping as easy as it should be.

p.s. there are features in Discourse which would make wiki pages
unnecessary, see https://www.discourse.org/about/ ("Wiki posts").
Basically you can have a long-running discussion thread which has wiki
features. Since everything in a discussion thread is dated, it's very
easy to see whether or not something is out of date.