New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Revert "haskellPackages.{SC,sc}alendar: nullify to fix ofborg eval" #33089
Conversation
Even broken packages shouldn't be breaking evaluation, I believe. |
I agree with @vcunat. This is the only
This isn't just about the package being broken, this is about the package being impossible, which I think deserves special attention. Update that said, if you think we shouldn't be checking even broken packages to properly eval, we could talk about it? |
@peti as mentioned I have revised that change in #33077 (comment). Would you consider that an acceptable solution? I believe that would address your concerns with regards to the usage of standard API's, and keeping our current PR checks working. |
@vcunat, the package is broken because it doesn't evaluate properly. That's the whole point of marking it broken. If it would evaluate, then I would not set |
@grahamc, yes, this is a much better; I'd be happy with that solution.
I do think it's odd to check the status of a package that's been explicitly marked as broken and I cannot easily think of a use-case for that. What kind of issue would the bot detect due to this check that would otherwise go unnoticed? |
e8a6f64
to
ec5e1d8
Compare
ec5e1d8
to
39eb15b
Compare
I don't have a strong opinion on this. Use case: I think it's not that unreasonable to put |
There are other possible motivations, e.g. avoid rebuilds for Hydra (or users) that are very likely to fail anyway (think Zero Hydra Failures), and for documentation of the state. |
Reverts #33077. Don't null out packages!
meta.broken = true
is the API we've been using this for a long time and I didn't get any memo that said that this had been changed.Cc: @grahamc