This has been going on for long before bytestring-0.12. I can try building with ghc alpha but last I checked we still had other problems with it in xmonad-contrib (https://github.com/xmonad/xmonad-contrib/issues/826). On Sat, Sep 16, 2023 at 2:12 PM Tomas Janousek wrote:

Hi,

On Sat, Sep 16, 2023 at 11:36:21AM -0400, Brandon Allbery wrote:

Anyone know what this is about? I get it weekly, presumably because I forked the xmonad-contrib repo. […] Packdeps: Some jobs were not successful

Yeah, I keep getting it as well: https://github.com/xmonad/xmonad-contrib/actions/runs/6204986546/job/1684766...

I think it's trying to tell us that our bounds exclude bytestring 0.12 which is now available in Hackage. What we need to do is make sure xmonad builds and works with this version, and update the bounds.

I'm not keeping up with the Haskell ecosystem these days so I don't know how to actually do that. Last time I checked, bytestring was a "boot" package shipped with GHC, so perhaps we'd need to test with GHC HEAD? Or can you ask cabal to just pull it and build it? --

Tomáš "liskin" ("Pivník") Janoušek, https://lisk.in/

-- brandon s allbery kf8nh allbery.b@gmail.com

On Sat, Sep 16 2023 19:12, Tomas Janousek wrote:

[… 10 lines elided …]

I think it's trying to tell us that our bounds exclude bytestring 0.12 which is now available in Hackage. What we need to do is make sure xmonad builds and works with this version, and update the bounds.

I'm not keeping up with the Haskell ecosystem these days so I don't know how to actually do that. Last time I checked, bytestring was a "boot" package shipped with GHC, so perhaps we'd need to test with GHC HEAD? Or can you ask cabal to just pull it and build it?

It is a boot library, but even GHC HEAD is only shipping with 0.11.4.0 currently[1], so I think it's fine to ignore this particular warning for now. I also wish one could disable it somehow, though. [1]: https://gitlab.haskell.org/ghc/ghc/-/wikis/commentary/libraries/version-hist... -- Tony Zorman | https://tony-zorman.com/

Actually there's four: `base`, `ghc`, `template-haskell`, `ghc-prim`. Everything else is fair game, although you may need to be explicit about it because cabal will prefer the installed version. On Sun, Sep 17, 2023 at 2:19 PM Joachim Breitner wrote:

Hi,

Am Samstag, dem 16.09.2023 um 19:12 +0100 schrieb Tomas Janousek:

...

Or can you ask cabal to just pull it and build it?

you can! As long as you don’t depend on the “ghc” library (which is the only library cabal cannot just rebuild, AFAIK).

Cheers, Joachim

-- Joachim Breitner mail@joachim-breitner.de http://www.joachim-breitner.de/

_______________________________________________ xmonad mailing list xmonad@haskell.org http://mail.haskell.org/cgi-bin/mailman/listinfo/xmonad

-- brandon s allbery kf8nh allbery.b@gmail.com

Hi, On Mon, Sep 18, 2023 at 07:07:44AM +0200, Tony Zorman wrote:

Oh, that's good to know; thanks! I was successful with

cabal build --constraint="bytestring ==0.12.0.2"

so I guess we can also just bump the `bytestring' dependency.

Let's do this then. Ideally this should be tested by the CI but it's not worth the effort—it will eventually be tested when bytestring 0.12 ships with a version of GHC that we add to the Matrix. And it's unlikely we'll break compatibility in the meantime. -- Tomáš "liskin" ("Pivník") Janoušek, https://lisk.in/

Hi Joachim, On Thu, Sep 21, 2023 at 12:21:04PM +0200, Joachim Breitner wrote:

See https://github.com/haskell-CI/haskell-ci/issues/667 if you care about this. I built https://github.com/nomeata/cabal-force-upper-bound so that on can make a CI matrix that exercises the upper bounds, but it hasn’t been integrated into haskell-ci so far.

Nice! Subscribed to the issue(s) so we can adopt it once it becomes easier to do so. Thank you :-) -- Tomáš "liskin" ("Pivník") Janoušek, https://lisk.in/