Using JIRA labels

Question

How do you use JIRA labels? We're kind of starting out with JIRA with a somewhat rusty team in terms of best practices, and it would mean alot to start off the right foot here.

Between components, projects and issue types, I'm somewhat at a loss as to how to best use labels without introducing redundancy; I had labels called "think-over" and "implement" to try it out, but it didnt feel natural. Complexity is something that could fit in, the kind of work to be done ("routine", "new stuff" etc) perhaps might be able to make it too; not much else I can think of.

How do you use them?

Answer 1

To make things as clear as possible, I would only use labels to describe aspects of an issue that can't be described by other attributes such as issue type, component, version or environment. Complexity might be a good example, but the kind of work to be done sounds more like an issue type.

Depending on how many kinds of data you want to record in labels, it might be better to use custom fields; these have the advantage that you can give them a fixed set of values, whereas labels are arbitrary strings, so two people might use different labels to describe the same thing.

Answer 2

Labels make a quick and dirty method of clustering related issues, almost like a quick filter. While components can and should be restricted to bare bones actual components or scenerios (ie installation, startup) -- labels can provide meta-groupings across components. I say use them liberally in the quick-filter sense and also as a low-level prioritization scheme. Keep an eye on 'all labels' and let them evolve naturally, do some pruning and you will come up with a right fit.

In summary, think freestyle and don't sweat it -- and always remember, use the force.