OpenJDK bug database: DRAFT Developer Workflow
Richard Bair
richard.bair at oracle.com
Wed Jan 4 19:58:16 UTC 2012
> But they are not sufficient.
Can you elaborate? Why are they not sufficient? You can perform any query you like based on tags, you can setup dashboards with built in filters, .... I assume by "they" you mean "tags" are not sufficient? Or do you mean the existing states are not sufficient?
> There's a world of difference between
> "open" : which could mean anything from "this bug just got assigned" to the states
> "Cause Known" : I've root caused the problem
> and
> "Fix understood" : I know exactly what to do here.
It doesn't bother me if you find value in these. I don't. But it doesn't matter, does it? JIRA tags are searchable, you can create filters, you can add these filters to your dashboard. Every time you log into JIRA you can look at pre-built searches showing what bugs yo have where cause is known vs. understood, vs. triaged, vs. any criteria makes you more productive. I don't see any reason to build these into JIRA as "blessed" sub states. They are useful tools for some, not used by most (I would venture).
Here's the problem. If the argument for making them official states is that the added data will make cross-engineer or cross-team data mining more rich, then I have a big problem with them. Because that means that, for those reports to be useful, everybody must be forced to use these sub states. But I would hate using such a system, and further, the accuracy of the reports would always be suspect. I don't want to be forced to use them.
So what is the point of having "blessed" sub states if they aren't used cross org? Just use tags, it is much more powerful and flexible than sub states and doesn't lead anybody to think that they can start producing org wide reports about bugs in "Fix Understood" or "Cause Known" and derive any meaningful information from it.
> People who don't want to use these states don't have to, but I hope they'll come to
> appreciate them. Don't regress useful functionality.
I don't see how there can be a regression when you can still perform these searches (and more!) with JIRA with tags than you could ever do by having baked-in sub states. I literally see no value in having them baked in.
> BTW there's a bunch of things that contribute to the suckiness of our jira installation.
> Its not just this one thing. Its the combination.
The only thing that irritates me on a daily basis is the fact that I have to login after 15 minutes of inactivity or whatever it is. If you have other concerns you'd better bring 'em out.
Richard
More information about the discuss
mailing list