Best Practice For Associating Custom Fields to Specific Projects and Issue Types

J_virpzx

We have a JIRA instance with several projects and maybe 150 or so Custom Fields. What is the best way to associate these Custom Fields to specific projects and issue types?

From what I can tell, there are two ways to do it:

1. Edit the custom field context, and limit it to just the projects and issue types that use the custom field. Then create a single screen for each project that contains all the possible custom fields that are used by the issue types in that project.
2. Leave the custom field context to be global, and instead create screens and screen schemes for each project and issue type. Only place on the screen the fields that are associated with that project and issue type combination.

I've seen suggestions for both solutions. This link (step 7) recommends method 1, whereas this link and this link (step 4) recommend method 2.

We currently use method 1, but were thinking about trying to migrate to method 2. However, as I started to look into it, I ended up needed to create a screen and screen scheme for pretty much every project / issue type combination. There isn't much consistency as far as the grouping for project / issue type combinations that our custom fields need to be associated with. And all those screens and screen scheme started to get a little unwieldy.

Are there any recommendations or best practices on which of the two methods above should be used?

Holger Schimanski_9q47k4

For 1. approach creating a screen for each project is not necessary. "limit it to just the projects and issue types that use the custom field" should be enough. Simply create a screen, which contains all custom fields. The custom fields, for which no custom field context exist for the project and issue type are automatically hidden on the screen e.g. when you create or edit an issue.

I personally use 1. and because Atlassian itself advertices 1. in http://blogs.atlassian.com/2013/12/8-steps-to-jira-field-greatness/ I would go this 1. Your second link, which recommends method 2 is mainly pointing to a bug in JIRA 5.1 with problems in using method 1. But it looks like this is solved (see last comment in https://jira.atlassian.com/browse/JRA-29310).