rework workflows for consistency with other repos

[altendky]

Draft for:

• [ ] review TODO:s

[altendky]

So aside from the general tidying and such that I need to do... @arvidn and @richardkiss (or even @cmmarslender), what do you think of sticking with ABI3 wheels vs. building separately for each Python version? My main interest in considering dropping ABI3 is just for consistency across our workflow definitions. There's enough complexity and variation between projects without having the main matrix also being different. Though, I also feel a bit silly with that. But I'm not sure that ABI3 really brings us a lot of value particularly. Mostly that the wheels would be forward compatible with future Python versions, what else? But I don't think that publishing a wheel is a big cost for us and we really ought to test the new Python version first anyways.

So, how do you all feel about this?

[arvidn]

at least chia_rs depends on the buffer protocol API, which is not part of abi3, but is part of all python versions we support. I don't have a strong opinion about standardizing on one or the other across wheels. Though, the wheel built from this repository (clvm_rs) is only used by clvm_tools's brun command, nothing else as far as I know. Other parts of clvm_tools is being replaced by clvm_tools_rs as far as I know.

[richardkiss]

I strongly believe that if we can use ABI3, we should do so. Why would we want to produce more artifact outputs when we can get away with producing fewer?

[altendky]

My main interest in considering dropping ABI3 is just for consistency across our workflow definitions. There's enough complexity and variation between projects without having the main matrix also being different. Though, I also feel a bit silly with that. But I'm not sure that ABI3 really brings us a lot of value particularly. Mostly that the wheels would be forward compatible with future Python versions, what else?

[github-actions[bot]]

'This PR has been flagged as stale due to no activity for over 60 days. It will not be automatically closed, but it has been given a stale-pr label and should be manually reviewed.'