Unmaintained

[sjshuck]

Some facts:

• No releases for almost 3 years (Oct. 2022)
• The last issue was closed almost 3 years ago (Nov. 2022)
• Of the current open issues (21), more than half of them (11) were created after that date
• Some pull requests since then have been subsequently closed by their authors due to lack of response, e.g. add overwrite method to MonadWriter #146

On social media, I've seen a lot of buzz about alternate effect systems, with many arguing that mtl has been superceded. However, there can be no doubt that mtl remains a fairly foundational library for many, many other libraries. "There is a total of 581 [reverse] dependencies in [Stackage snapshot] lts-24.5". Just to take for example a popular library that is very much actively maintained, lens presents a bunch of core functionality in terms of mtl typeclasses:

view :: MonadReader s m => Getting a s a -> m a

You can't just drop an mtl replacement in there.

If nobody steps up, I would like to volunteer to co-maintain, although I can't in good conscience bear the burden of full maintainership.

[hasufell]

Since mtl is a core library, I'm bringing this to the attention of the CLC.

@angerman @velveteer @TeofilC @Daniel-Diaz @ChickenProp @noughtmare @mpscholten @doyougnu @parsonsmatt

[hasufell]

@chessai @emilypi @kozross how do you feel about adding @sjshuck as maintainer?

[chessai]

@chessai @emilypi @kozross how do you feel about adding @sjshuck as maintainer?

I stepped down from maintaining mtl. I have no opinion here as a maintainer. My opinion as a user is that new maintainers would be welcome.

[emilypi]

This is largely a CLC issue, since I technically only had this as a CLC member, and then I stopped being a CLC member, and stayed on for Koz when they became a CLC member, but now no one's a CLC member so the CLC should decide what to do with it.

I would like to see someone else take this on, but I'd expect to see a merit-driven selection so we can stop the cycle of people being chewed up by the volunteer machine here. I have no idea who @sjshuck is, so I can't say whether he's a good choice or not.

[hasufell]

I can't quite follow how CLC membership relates to mtl maintainership.

But do I assume correctly that none of the maintainers listed in the mtl.cabal file are currently active maintainers?

mtl/mtl.cabal

Lines 7 to 9 in f69c6d3

	maintainer: chessai <chessai1996@gmail.com>,
	Emily Pillmore <emilypi@cohomolo.gy>,
	Koz Ross <koz.ross@retro-freedom.nz>

[emilypi]

I can't quite follow how CLC membership relates to mtl maintainership

This is a CLC library and they allocate the maintainers. It's up to them to keep productive maintainers staffed on the project and not the job of the maintainers, who may not be able to maintain it anymore for whatever reason. Many of the maintainers on core libraries are only given maintainership for the duration of their tenure on the CLC, and the expectation is (or was) that they'd stop maintaining the project once their tenure was up, which is what my deal was. When Koz replaced me on the CLC, i was not a maintainer anymore for the project. I was brought on later to help with it, but not as a lead.

But do I assume correctly that none of the maintainers listed in the mtl.cabal file are currently active maintainers?

Correct. The CLC needs to select new maintainers. I comment on here from time to time bumping things for Koz, but I have not been actively engaged with Haskell for a year or two now.

[TeofilC]

It sounds like we (the CLC) should do a call for maintainers on the Discourse and other platforms, and then pick a group of people to maintain mtl.

[hasufell]

Many of the maintainers on core libraries are only given maintainership for the duration of their tenure on the CLC

I'm not sure this is accurate as per the current CLC practices. There is no documentation of such in our README and it's the first time I hear this.

I do not know if such historical practices existed though or if such agreements were made ad-hoc.

[hasufell]

It sounds like we (the CLC) should do a call for maintainers on the Discourse and other platforms, and then pick a group of people to maintain mtl.

I would like to give @kozross the opportunity to comment on the maintainership status first.

If there's even one active maintainer left, we should give them the opportunity to resolve such issues on their own.

[emilypi]

That seems like a massive continuation issue then because it was clearly delineated in the Haskell wiki up until the latest CLC changed it. Sorry generally if that process was dropped. The CLC should probably go through and audit and figure out what their ownership burden looks like then.

[andreasabel]

@ekmett passed maintainership to the CLC in 2021 here:

• Cut a new release of mtl #69 (comment)

I also remember there was a lot of discussion about the maintainership, and I am a bit surprised that the current maintainers went passive so quickly after they won.

[hasufell]

Ok. I've added @sjshuck as maintainer on github.

Can you confirm whether this is your hackage account https://hackage.haskell.org/user/SShuck so that @Bodigrim or @andreasabel can give you hackage upload access to mtl?

[andreasabel]

Only existing maintainers (see https://hackage.haskell.org/package/mtl/maintainers/) or hackage admins (in case of a takeover, see https://hackage.haskell.org/users/admins/) can add new maintainers on hackage.

[emilypi]

I am a bit surprised that the current maintainers went passive so quickly after they won.

I'm sure you can think a little harder and use a little more imagination and that surprise will go away pretty quickly.