A few questions on using compute credits

[ericoporto]

Hi,

Trying to understand how to use compute credits

• do I HAVE to mark it in the YAML, or there is some interface I can just say "hey you can use compute credits for this ci run specifically".
• is there some way to not mark anything in the yaml but have it using compute credits for the organization repository only?
• if the compute credits reach zero, will it eat negative credits that put me in debt or will it just stop using? I want to know if I have to do something to not have negative compute credits.

This is it. Overall I would prefer if there was a less sudden transition - I also don't look into the blog and got the knowledge that this would happen because of seeing accidentally in a GitHub issue in another repository. I also would really prefer if cirrus-ci had a better manual interface or better metrics (a time frame lesser than months), overall is really hard to access what is our consumption and if an effort to reduce consumption is effective or not...

This meeting in Conda forge has an interesting item

https://conda-forge.org/docs/orga/minutes/2023-08-09.html

Running out of credits would mean it would stop suddenly (bad UX story)

This isn't even true, running out of "free credits" makes it stop and there's no button to make it actually consume the paid credits.

To me it would be ideal if there was a paid plan that simply raised the "free" compute credits. Also the way it works now anyone in the org can pay to add compute credits but only some specific org members can change the plan because it's locked in the GitHub interface.