Huge +1 to this.
I think X=6m and Y=3m makes sense, as something as simple as adding a comment
refreshes it. As such, slightly longer than a normal Flink release cycle is a
good line in the sand (as Matthias has said)
Side note : Just thinking of a typical PR review process, PRs either fall in
Hi Rui,
Good point about recording the proposal. I am not sure it fits the FLIP process
so I made a page [1] under the CHI workgroup.
That page has a summary of the proposal and links to the discussion and vote
threads.
Thanks,
Tom Cooper
@tomncooper | tomcooper.dev
[1] https://cwiki.apache
Thanks Tom for driving this discussion!
It sounds make sense to me, could you help create a FLIP or doc
to record it? It's helpful to let other developers check the stale strategy
and background in the future, thank you.
Best,
Rui
On Wed, Dec 18, 2024 at 5:16 PM Matthias Pohl wrote:
> +1 for a
+1 for adding the automation considering that other projects are doing this
as well and the high amount of open PRs doesn't help. Thanks for
summarizing the current state in such detail.
X + Y should probably be larger than our usual release cycles. X=6m and
Y=3m might be alright in this regard.
Hey!
Big +1
I think this is a great initiative and we should follow the path laid down
by other projects.
Starting with X=1 year and Y=3 months is very forgiving, I personally
wouldn't mind immediately going more aggressive to clean up the backlog.
So for me X = 6m Y = 3m is also perfectly fine.
Hi Tom,
I like the idea of starting with X=1 year and Y=3 months. That plus having
CHI review recent PRs would be helpful.
Your suggestion of having a bot which "lints" naming conventions could be
useful. Maybe there could be a CHI-bot to automate what you and others are
doing manually. :)
I wo
TL;DR
- We have over 1.2k open PRs, this is an issue as it makes new contributors
think twice about committing and looks like a problem that is too-big-to-solve
for committers.
- There have been various attempts, over the last 6 years, to enable the Stale
PR bot/action to prompt authors to refr
