Why are scorecards restricted to component types?

[Jan-Hendrik Spieth]

So far, we have the option of creating scorecards per component type. We can then limit the scorecard down further to certain components, e.g. by label. But the basic restriction is by type.

After the first few experiments with scorecards (basic health metadata), we find ourselves copying scorecards just to have them available for several component types.

What's the logic behind this?

Why should scorecards by specific to component types? Why now allow us to define basic scorecards that are applicable to all component types?

To give you an example, we were thinking about checking component ownership like so:

• field - has owner
• metric - sth that rings an alarm when a component is "orphaned" due to staff leaving the company (last Jira login date, or user active/deactivated status - we're not sure yet this can be done, but that's our idea right now).

This "ownership" scorecard is applicable to all component types.

[Patrick Hill]

Hi @Jan-Hendrik Spieth

Thanks for reaching out! I'm the Product Manager for Scorecards and would love to chat more about this - mind finding some time using https://calendly.com/patrick-atlassian/let-s-talk-about-compass?

When we started out to build Compass, we found that specific types of Components had certain fields that only applied to that particular domain - an example here is on-call for Services.

As we've expanded our thinking over time and made component types more flexible with Metrics and Custom Fields, we realise that you may want to have a Scorecard across multiple Component types. I'm super curious about how else you'd want to use Scorecards - targeting on different properties or fields, or new types of criteria so if you or others have ideas please let me know!

Thanks,

Patrick

[Jan-Hendrik Spieth]

Hi @Patrick Hill thanks for your reply.

We've just started with Compass! And quite some of our experimental components are external. That's a bit special and not what Compass has been built for it seems.

But, let us do a bit more experimenting with metrics and scorecards. I'll then get back to you!