Release preparations for 2.0

[sgraf812]

I mostly completed the ChangeLog from looking at this list.

If someone would give me the necessary hackage credentials (user sgraf812), I would be happy to do the hackage release.
I would just use nix run nixpkgs#hkgr -- publish to do so. Or is there some documented procedure one should follow for releases? At any rate, we should document the procedure. Perhaps in CONTRIBUTING.rst?

[Ericson2314]

happy/doc/introduction.rst

Lines 147 to 156 in 652c4ca

	Current Maintainers
	~~~~~~~~~~~~~~~~~~~
	- Andreas Abel (@andreasabel)
	- Vladislav Zavialov (@int-index)
	- John Ericson (@Ericson2314)
	- Simon Marlow (@simonmar)

do you want to be an official maintainer? :)

[sgraf812]

do you want to be an official maintainer? :)

I would be honored to, and added myself to the list :) 🎉

[Ericson2314]

If @int-index or @andreasabel signs off on this, I'll give @sgraf812 the haddock hackage perms :)

[int-index]

If @int-index or @andreasabel signs off on this, I'll give @sgraf812 the haddock perms :)

Please do!

[Ericson2314]

Hackage perms granted! :)

[sgraf812]

Thanks :) I'll wait for another 9 hours to give @andreasabel a chance to react before doing the release.

[andreasabel]

Thanks for all the good work, @sgraf812 !

I did local tests of cabal build, cabal test, cabal install, and using the new Happy for building Agda and other projects.

Could you update the release candidate on hackage (https://hackage.haskell.org/package/happy-2.0/candidate) before pressing the final button? The current one is

Uploaded by int_index at 2024-02-03T14:08:09Z

[sgraf812]

Here's a package candidate for happy-lib: https://hackage.haskell.org/package/happy-lib-2.0/candidate

And the one for happy: https://hackage.haskell.org/package/happy-2.0/candidate

Shall I do cabal gen-bounds as suggested? Also the missing documentation seems strange (I had uploaded it as well)

[andreasabel]

Reading the changelog in another context (on Hackage), I found a typo.

Here's a package candidate for happy-lib: hackage.haskell.org/package/happy-lib-2.0/candidate

Thanks!

Atm, you are the sole maintainer, this is probably not intended:

Package maintainers
sgraf812

Shall I do cabal gen-bounds as suggested?

Na, I would ignore this.
Let's wait for real breakage, then one can still place upper bounds via Hackage revisions.

Also the missing documentation seems strange (I had uploaded it as well)

The happy-lib does have the documentation.
Executables usually do not have any documentation, so this looks alright, if you ask me.
(In contrast, Agda is both an executable and a library, so it has documentation.)

[sgraf812]

Atm, you are the sole maintainer, this is probably not intended:

Thanks for pointing me to the link where I could change it.

The happy-lib does have the documentation.

That was due to an earlier experiment on my part where I simply made the main component expose all modules directly and deleted all sublibraries (apologies, should have reverted). The latter step was sadly necessary to see the full docs, which seems like a no go after the discussion in #288.

I just re-uploaded the candidate with the state where we have multiple private libraries and after fixing the ChangeLog.

I'm rather reluctant to get rid of the private libraries again (#288), so let's just release as-is and reupload docs when cabal has sorted out the situation wrt. merging documentation of multiple private sublibraries (haskell/cabal#10368). I honestly do not expect much use of happy-lib anyway, and the few users that we attract are likely experts.

A quick glance at https://hackage.haskell.org/package/happy-lib-2.0/candidate and a nod would be appreciated.

[andreasabel]

LGTM ! (except that the typo still seems to plague the candidates).

[sgraf812]

Thanks! Released as https://hackage.haskell.org/package/happy-lib-2.0 and https://hackage.haskell.org/package/happy-2.0.