Looking for contributors for finding closable issues

[bimlas]

As @TiddlyTweeter pointed out, there are a lot of open bugs on GitHub: https://github.com/Jermolene/TiddlyWiki5/issues/3497

Technical people like to see the GitHub repository to find out "how much the project is living". If they find that there are a lot of open issues, they may not be familiar with Tiddly because they think it's full of bugs.

Anyone who likes to help reduce the number of opened issues, please come here: https://github.com/Jermolene/TiddlyWiki5/issues/3528

[Mohammad]

Good idea bimlas!

[@TiddlyTweeter]

bimlas, Good idea.

I'm limited in what I can do in that I'm not good on understanding most of the issues so would be an unreliable commentator. On the few things I understand I will.

I noted PMario flagged a LOT of issues that might be closed. Can't he just close them? I trust his judgement. I don't think it matters if he got a few things wrong. Especially for older Issues.

Best wishes

Josiah

[@TiddlyTweeter]

I been thinking more about this.

I have also been looking at what PMario has been doing flagging things for possible "closure". Good work, but how does it get actioned?

I'm guessing there are reasons that GitHub issues opened in 2013 are still open--i.e. they have some potential.

Maybe part of the barrier to closure is that its either "closed" OR "open" ? There is no in-built in-between state.

How would it look with a third state? Like: "still interesting, but way old" ... So it would be "closed" but be flagged "potential", or something like that.

I think bimlas' main idea that having several hundred old issues IS off putting to potential new developers is very true. And that reducing the "live" issue numbers is a good idea--plus some procedure to keep flagged closed issues of some possible potential.

Hope this is clear!

Best wishes

Josiah

[PMario]

Hi,

On Monday, November 12, 2018 at 10:50:36 AM UTC+1, @TiddlyTweeter wrote:

I have also been looking at what PMario has been doing flagging things for possible "closure". Good work, but how does it get actioned?

At the moment, the creator of an issue, or the owner of the repo are able to close issues. The same is true for pull-requests (PRs).

 

I'm guessing there are reasons that GitHub issues opened in 2013 are still open--i.e. they have some potential.

That's right. Some of the open issues still have a lot of potential, even if they are old. The main reason, why they are still open is: We don't want them to be forgotten!

eg - issues labelled:

 - improvement

 - new feature

 - ...

 

Maybe part of the barrier to closure is that its either "closed" OR "open" ? There is no in-built in-between state.

That's right. IMO it doesn't make sense to have more states. If an issue is "stale" in open state, it would be also stale in an "in-between" state. So the only possibility, I see is: Assigning labels see above.

If we have them labelled, it's relatively clear, what's going on.

I think bimlas' main idea that having several hundred old issues IS off putting to potential new developers is very true.

IMO most long running repos have many open issues. ... May be the same reasons.

 

And that reducing the "live" issue numbers is a good idea--plus some procedure to keep flagged closed issues of some possible potential.

As I wrote. There are many issues, that still have value. A different way, to remember them, would be to create 3 meta-issues. 

eg:

 1) new-features - closed but still valid

 2) improvements - closed but still valid

 3) discussions

Those 3 meta-issues will be always Open

Possible and simple workflow

 a) Meta-issues 1,2,3 contain some description, AND a link like: improvement

 b) If an issue with label 1,2,3 are open longer than 1 year, they will be closed with a reference to the meta-issue

 c) Some issues may be closed, because they are outdated, not actionable ...

     They don't get the link to meta issues.

If an issue is closed with b), github will automatically create a link to the meta-issue and a backlink to the issue, that will be closed. With this mechanism it is easy to follow the trails.

just some thoughts.

have fun!

mario

[@TiddlyTweeter]

Ciao PMario

Thanks for the detail!

You very much address my cruder points in a more elegant way! And you, basically, say somewhat the same: with "closed but valid". That is pretty much what I meant.

BUT, still there is WHO would do the flagging?

I don't like the idea its just you. Its a burden. On the other hand WHO has the competence to know? I know I don't for other than a few things.

Many thanks

Josiah

PMario wrote ...