I do not have a strong feeling about whether it should be in the title or
in the description. I kind of like all titles looking "the same" versus
some PRs missing an issue that could cause the PR to get ignored like it is
less important.

I agree with Alexey's concern. I do expect that issue linking will be less
thorough now. It should still be in the checklist in the issue description.
I don't know how much people really pay attention to the checklist though.
We could potentially have the checklist enforced by a simple action that
required either an issue link or a tag indicating that there is none.

Kenn

On Fri, Jun 24, 2022 at 2:37 PM Alexey Romanenko <[email protected]>
wrote:

> My main point was to minimise the number of PRs that are not properly
> linked to their issues since it risks to keep this issues open even if they
> are already technically resolved. Imho, with an issue prefix in PR title,
> as it was before, it’s much easier and faster to verify.
>
> Though, if everybody finds it not useful then we will keep a current
> behaviour.
>
>
> > On 24 Jun 2022, at 21:40, Robert Bradshaw <[email protected]> wrote:
> >
> > #N will automatically link to the issue (or PR), which should
> > generally show up in the description. I'm not sold on the value of
> > having to have it in the title. [BEAM-nnnn] was primarily useful
> > because one had to reference an external system.
> >
> > On Fri, Jun 24, 2022 at 12:34 PM Danny McCormick
> > <[email protected]> wrote:
> >>
> >> I don't think a similar "[issue #] title" linking construct exists for
> GitHub. AFAIK the recommended method of linking is to put things in the
> description, and I don't really see a reason that putting the issue number
> in the description is harder or more likely to be forgotten; I kinda expect
> it will just take some time for people to adjust to doing this the new way.
> >>
> >> I'm probably ambivalent (or slightly opposed) to mandating an issue in
> the PR title - I mostly just see it as a small extra tax with limited value
> since I do think we should ask for issues in the description to preserve
> linking. With that said, I also haven't been around the project for as long
> and didn't previously build that feature into my regular workflow - others
> might find it more useful.
> >>
> >> Thanks,
> >> Danny
> >>
> >> On Fri, Jun 24, 2022 at 1:30 PM Alexey Romanenko <
> [email protected]> wrote:
> >>>
> >>> I really liked and used the “feature” that we had before when we asked
> developers to add a Jira issue ID as a prefix to commit messages/PR titles.
> >>>
> >>> What’s about to have a similar thing with Github Issues? As additional
> bonus we could link (is it possible?) a PR to a corresponding issue (as I
> can see now, quite often people tend to forget to add this into PR
> description).
> >>>
> >>> —
> >>> Alexey
> >>>
> >>> On 7 Jun 2022, at 16:04, Danny McCormick <[email protected]>
> wrote:
> >>>
> >>> This is definitely possible - if we have a list of the dependencies to
> ignore we can specify an ignore list, or as they come up you can comment
> "@dependabot ignore". In this case, having the explicit ignore list
> probably makes sense. I'll follow up with Tomo to make sure the GCP
> dependencies get added to the ignore list.
> >>>
> >>> Thanks,
> >>> Danny
> >>>
> >>> On Mon, Jun 6, 2022 at 8:10 PM Ahmet Altay <[email protected]> wrote:
> >>>>
> >>>>
> >>>>
> >>>> On Sat, Jun 4, 2022 at 8:30 PM Sachin Agarwal <[email protected]>
> wrote:
> >>>>>
> >>>>> This is great, thank you so much Danny! I checked my issues and all
> look correct.  Thank you!
> >>>>>
> >>>>> On Sat, Jun 4, 2022 at 6:40 PM Danny McCormick <
> [email protected]> wrote:
> >>>>>>
> >>>>>> All Jiras should now be migrated to Issues*, and the owners should
> be assigned or tagged. Hopefully this will help us be a more productive
> community and will make it easier for newcomers! If you see any issues with
> the migration, or generally with using issues, please let me know.
> >>>>
> >>>>
> >>>> Is it possible to disable dependabot for certain dependencies? GCP
> dependencies are managed in a way to happen at a lockstep, based on
> specifically validated sets
> >>>>
> >>>> /cc @Tomo Suzuki - who did lots of work in that area.
> >>>>
> >>>>
> >>>>>>
> >>>>>>
> >>>>>> Thanks,
> >>>>>> Danny
> >>>>>>
> >>>>>> * The "Beam Dependency Report Jiras" were not migrated for 2
> reasons. (1) There were >500 of them dating back to Nov 2019, many
> outdated. (2) Dependabot has been added to the repo and should take care of
> putting up PRs for outdated dependencies. If that is an issue for any
> reason let me know, it is not too late to migrate those as well.
> >>>>>>
> >>>>>> On Fri, Jun 3, 2022 at 11:14 AM Danny McCormick <
> [email protected]> wrote:
> >>>>>>>
> >>>>>>> Hey Hector, they were just enabled (thanks @Valentyn Tymofieiev
> for merging the PR).
> >>>>>>>
> >>>>>>> The existing Jiras should be migrated throughout the rest of the
> day - this will take a while due to GitHub rate limits, but it should
> definitely be done by the end of the weekend (and I expect significantly
> earlier). In the meantime, feel free to start opening issues at
> https://github.com/apache/beam/issues, and let me know if you see any
> issues with the migrated issues.
> >>>>>>>
> >>>>>>> Thanks,
> >>>>>>> Danny
> >>>>>>>
> >>>>>>> On Fri, Jun 3, 2022 at 10:42 AM Hector Miuler Malpica Gallegos <
> [email protected]> wrote:
> >>>>>>>>
> >>>>>>>> When will github issues be enabled?  I understood that it would
> be today
> >>>>>>>>
> >>>>>>>>
> >>>>>>>>
> >>>>>>>> Hector Miuler Malpica Gallegos
> >>>>>>>>
> >>>>>>>>
> >>>>>>>>
> >>>>>>>> El jue, 2 jun 2022 a la(s) 21:35, Hector Miuler Malpica Gallegos (
> [email protected]) escribió:
> >>>>>>>>>
> >>>>>>>>>
> >>>>>>>>> hahaha!!! excuse me, I'm thinking in the flink project when read
> the last email, I do not know why.
> >>>>>>>>>
> >>>>>>>>>
> >>>>>>>>> Hector Miuler Malpica Gallegos
> >>>>>>>>>
> >>>>>>>>>
> >>>>>>>>>
> >>>>>>>>> El jue, 2 jun 2022 a la(s) 18:04, Danny McCormick (
> [email protected]) escribió:
> >>>>>>>>>>
> >>>>>>>>>> Thanks for calling that out Hector - the migration will only
> migrate issues from the Beam project (that one is from the Flink project),
> so I don't think it should be an issue.
> >>>>>>>>>>
> >>>>>>>>>> Thanks,
> >>>>>>>>>> Danny
> >>>>>>>>>>
> >>>>>>>>>> On Thu, Jun 2, 2022 at 6:38 PM Hector Miuler Malpica Gallegos <
> [email protected]> wrote:
> >>>>>>>>>>>
> >>>>>>>>>>> This migration includes the issues of the component
> `Kubernetes Operator` ? like this issue FLINK-27820 ? this issues
> correspond to the repository
> https://github.com/apache/flink-kubernetes-operator, keep it in mind.
> >>>>>>>>>>>
> >>>>>>>>>>>
> >>>>>>>>>>> Hector Miuler Malpica Gallegos
> >>>>>>>>>>>
> >>>>>>>>>>>
> >>>>>>>>>>>
> >>>>>>>>>>> El jue, 2 jun 2022 a la(s) 12:33, Danny McCormick (
> [email protected]) escribió:
> >>>>>>>>>>>>
> >>>>>>>>>>>> Given the consensus here, I updated the tool to do this. This
> means that we won't update the JIRAs to be read-only until after the
> migration is complete. I'll rerun the tool if any extra jiras come in
> during the intervening period. The tool will also still write the mapping
> to the file in case there are unforeseen issues so that we can backfill if
> needed.
> >>>>>>>>>>>>
> >>>>>>>>>>>> Thanks for the suggestion and followup Brian, Ahmet, and
> Alexey.
> >>>>>>>>>>>>
> >>>>>>>>>>>> On Thu, Jun 2, 2022 at 12:16 PM Alexey Romanenko <
> [email protected]> wrote:
> >>>>>>>>>>>>>
> >>>>>>>>>>>>> +1 That would be very helpful for mapping!
> >>>>>>>>>>>>>
> >>>>>>>>>>>>> On 2 Jun 2022, at 17:48, Ahmet Altay <[email protected]>
> wrote:
> >>>>>>>>>>>>>
> >>>>>>>>>>>>> Is it possible to add comments on the JIRAs with a link to
> the new corresponding github issue?
> >>>>>>>>>>>>>
> >>>>>>>>>>>>> On Thu, Jun 2, 2022 at 8:47 AM Danny McCormick <
> [email protected]> wrote:
> >>>>>>>>>>>>>>
> >>>>>>>>>>>>>> Thanks for the feedback, I agree it would be good to keep
> that option open - I updated the tool to write those to a file when we
> create an issue. I'll share that after the migration.
> >>>>>>>>>>>>>>
> >>>>>>>>>>>>>> Thanks,
> >>>>>>>>>>>>>> Danny
> >>>>>>>>>>>>>>
> >>>>>>>>>>>>>> On Wed, Jun 1, 2022 at 7:03 PM Brian Hulette <
> [email protected]> wrote:
> >>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>> Thanks Danny. Regarding links to GitHub issues, if we
> could at least save off a record of jira <-> issue mappings we could look
> at adding the links later. I think it would be nice to have those links so
> that anyone landing in a jira through a search or an old link can quickly
> find the current ticket, but I don't think that needs to block the
> migration.
> >>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>> On Wed, Jun 1, 2022 at 7:05 AM Danny McCormick <
> [email protected]> wrote:
> >>>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>>> Hey Brian,
> >>>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>>> 1. Right now, the plan is to (1) turn on the issues tab,
> (2) make the JIRA read only, (3) run the migration tool. Since the
> migration tool won't be run until after Jiras are read only, there
> shouldn't be issues with making sure everything gets captured.
> >>>>>>>>>>>>>>>> 2. That current ordering does mean it's difficult to add
> a link to the newly created Issue, and I hadn't built in that feature. With
> that said, I will ask Infra if they're able to put up a banner redirecting
> people to GitHub for the Beam project - that should hopefully minimize some
> of the issues - and I'll also look into updating the tool to do that in
> case the banner isn't doable. I'm also planning on doing a few passes to
> update our docs and code comments from Jiras to issues once the migration
> is done.
> >>>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>>> Thanks,
> >>>>>>>>>>>>>>>> Danny
> >>>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>>> On Tue, May 31, 2022 at 8:09 PM Brian Hulette <
> [email protected]> wrote:
> >>>>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>>>> Thanks Danny, it's great to see this happening!
> >>>>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>>>> A couple of questions:
> >>>>>>>>>>>>>>>>> - Is there something we can do to remind people creating
> a jira that they should create a bug instead (e.g. a template)? If not I
> suppose we can just re-run the migration tool a few times up until jira
> creation is disabled to make sure everything is captured.
> >>>>>>>>>>>>>>>>> - Will your migration tooling comment on the original
> jira with a link to the new issue in GitHub?
> >>>>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>>>> Brian
> >>>>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>>>> On Tue, May 31, 2022 at 9:57 AM Robert Bradshaw <
> [email protected]> wrote:
> >>>>>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>>>>> Thanks for finally making this happen.
> >>>>>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>>>>> On Tue, May 31, 2022 at 7:18 AM Sachin Agarwal <
> [email protected]> wrote:
> >>>>>>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>>>>>> Thank you Danny! This will help us a lot, especially
> with new contributors. Thanks so much!
> >>>>>>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>>>>>> On Tue, May 31, 2022 at 4:10 AM Danny McCormick <
> [email protected]> wrote:
> >>>>>>>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>>>>>>> Hey folks, this is a reminder that we will be
> migrating from Jira to GitHub Issues this Friday (6/4). A few key details
> to keep in mind:
> >>>>>>>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>>>>>>> 1. All active Jiras will get automatically migrated
> and assigned over the course of the weekend.
> >>>>>>>>>>>>>>>>>>>> 2. Starting Friday (once the the Issues tab is open),
> please stop creating Jiras and start creating Issues instead. You should
> also reference issues in your PRs and commits instead of Jiras. The Jira
> creation flow will eventually be disabled.
> >>>>>>>>>>>>>>>>>>>> 3. If you encounter any issues that can't be resolved
> by looking at the doc updates, please let me know and/or follow up in this
> thread.
> >>>>>>>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>>>>>>> I'm looking forward to seeing how Issues can minimize
> friction for new contributors and I'm hopeful that this will be a smooth
> transition. If you have any last minute concerns let me know. For more
> context, see the original thread on this topic.
> >>>>>>>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>>>>>>> Thanks,
> >>>>>>>>>>>>>>>>>>>> Danny
> >>>>>>>>>>>>>
> >>>>>>>>>>>>>
> >>>
>
>

Reply via email to