Skip to main content

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
Re: [che-dev] Issue tracking, triage, and sprint planning: is there a better option?

On 2/22/21 1:58 PM, Angel Misevski wrote:
Thanks for the context Brian, this should have been considered earlier in the process here.

A topic that was brought up during the call that I feel didn't get enough discussion is the distinction between "moving to Jira" and "moving to the Jira instance at issues.redhat.com".

It's one thing to move to some (currently non-existent) Jira managed by the Eclipse foundation, it's another to move to a Jira that is controlled by Red Hat. While GitHub is a Microsoft product, it's a platform, and the Eclipse foundation ultimately has control of the Eclipse organization on GitHub. Moving to a Red Hat-owned Jira blurs the lines of ownership in a way I'm not entirely comfortable with -- does Red Hat support a way for external contributors to manage Red Hat Jira issues, create sprints/manage priority/etc., or are the benefits of Jira we're discussing only available to those of us with @redhat.com emails? What admin-esque privileges would the Eclipse foundation have within the Jira system?

I think the only administrative privileges that apply to issue management is who gets to assign components/priority (currently labels), and create Kanban boards, sprints, etc. With GitHub we currently restrict those privileges to committers only, but this is a GitHub specific restriction. In other systems like Bugzilla, users cannot file an issue without a component, so choosing a component (currently label) isn't something that's restricted. I believe non-committers can also adjust the severity of their bugs in Bugzilla, so in this case we are more restrictive than other issue tracking systems. I could argue that this makes GitHub more community unfriendly than other trackers, but that's a topic for another time. :)


Put another way, is moving to Jira in the way we're discussing something that another project, e.g. Theia, could also benefit from? I suspect that most of the upsides we're discussing here apply mostly to us.

It really depends on each project's needs. IMO the features lacking from GitHub make it a poor issue tracker for any project with a large number of full time developers working on it. I'm sure it's fine for the individual developers themselves, but it's hellish for anyone who needs to manage them or the project.

I know there are other huge projects using GitHub, but we can't really infer from their mere existence that their workflows are satisfactory. Maybe they are having the same debates we are, who knows. At any rate, it seems like GitLab has a lot of missing issue management features that GitHub lacks, and it's hosted by the Eclipse Foundation. I think that's something we should explore further.


Eric




Back to the top