Stackage: algebraic-graphs-0.1.0

[juhp]

QuickCheck-2.11.3 is out of bounds for:

• [ ] algebraic-graphs-0.1.0 (>=2.9 && < 2.11). Andrey Mokhov [e-mail] @snowleopard. @snowleopard. Used by: test-suite

[snowleopard]

@juhp Thanks! I've fixed this in https://github.com/snowleopard/alga/pull/57.

What's the protocol? Shall I reupload algebraic-graphs-0.1.0 with a patched Cabal file, or shall I upload a new version algebraic-graphs-0.1.1 instead?

[juhp]

A Hackage revision is fine or you can upload a new version: it's up to you :)

[snowleopard]

Thanks :-)

When do you expect to release the last LTS with GHC 8.2.2? I'd like this version to make it there.

[juhp]

Not exactly sure what you are asking? LTS 11 will continue for some time (months) while Nightly continues to stabilize I would say.
We can include it soon in lts-11 if you open an lts-haskell ticket - good to check it builds there with stack

[snowleopard]

@juhp Thanks, I think you answered my question :)

algebraic-graphs works with GHC versions as old as 7.8.4, so my understanding is that if I would like to include it into any prior LTS (not necessarily lts-11), I would just have to raise an issue in lts-haskell repository. Is that right?

[snowleopard]

I've uploaded https://hackage.haskell.org/package/algebraic-graphs-0.1.1, which fixes this issue.

[juhp]

Thanks

Yes, that's correct, but much older LTS releases are no longer maintained so much, due to issues with updating them: currently only LTS 11 is really being actively updated regularly.

(Curators: maybe we could using NOPLAN=1 to add packages easily to old LTS releases?)

[snowleopard]

@juhp So, now that algebraic-graphs-0.1.1 is on Hackage, how do we bring it to the latest Stackage? Do I need to send a PR similar to this commit:

https://github.com/fpco/stackage/commit/d02329d32aa23d6497396683245eadbb9e195b8e

but using algebraic-graphs >= 0.1.1 instead?

[snoyberg]

@snowleopard You can send a PR that reverts that commit, it'll do the same thing.

[snowleopard]

@snoyberg Done: #3479.

[snowleopard]

@juhp I guess this can now be closed.

[snoyberg]

Thanks!