#312 makes me think we should cut a release before v2.0 so resolvers can include polysemy together with first-class-families-0.7.0.0. It would include all changes up to #311, which I'll merge when I have time to write its tests. It would probably be the final release before v2.0.
polysemy has restrictive upper bounds on first-class-families (>=0.5.0.0 && < 0.7), which causes it to fail to build with first-class-families-0.7.0.0. I'm posting an issue here (...
Closes #310.
How would you go about creating tests for this? Something like "take 100 . fst . run . runLazyOutputList $ forever (output ()) must terminate within 2 seconds?"
Continuing this thread for the v1.3 release. Can someone remind me why we turned off loop-breaker? Has that problem been fixed by now? If so, we should reenable it.
Also, @TheMatten can you add polysemy-1.3.0.0 as an upper bound for polysemy-zoo on versions 0.6.0.0 and 0.6.0.1 on Hackage? Otherwise the migration of Tagged will cause conflicts.
Also, I think every version of polysemy-pluginbefore 0.2.5.0 also now needs polysemy-1.3.0.0 as an upper bound due to the change in internals with the membership proof rewrite.
Merged in #311 and the v1.3 pr. Once we can fix the bounds for the older versions of polysemy-zoo and polysemy-plugin, we'll cut new releases for all three.
Done! All left now before cutting the releases is to merge #62 of polysemy-zoo, and before I do that I'm gonna check if there are any dependencies (other than hedis) that have been made redundant with the removal of Redis support in polysemy-zoo.
Oh right, I'm not part of the uploader group yet. I've sent away a mail to the hackage trustees, but I can't cut the releases myself just yet. If anyone else wants to do the work, feel free, otherwise I'll do it when I can.
All releases cut. Documentation for polysemy-zoo failed to build on Hackage, so we need to upload our own documentation for it, and I'm not sure how that system works.
#312 makes me think we should cut a release before v2.0 so resolvers can include polysemy together with first-class-families-0.7.0.0. It would include all changes up to #311, which I'll merge when I have time to write its tests. It would probably be the final release before v2.0.
Alternatively, is it possible to change the first-class-families dependency bounds of polysemy v1.2.3.0 on Hackage?
Yeah, it is --- you can just do it in the maintenance panel without needing to reupload
@TheMatten Could you do that?
Oops, sorry, wasn't here - is it already sorted out?
Not yet
Nothing interesting happened in new version of fcf, right?
https://hackage.haskell.org/package/first-class-families-0.7.0.0/changelog
No breaking changes, it seems
Okay, should be done
Continuing this thread for the v1.3 release. Can someone remind me why we turned off loop-breaker? Has that problem been fixed by now? If so, we should reenable it.
Also, @TheMatten can you add
polysemy-1.3.0.0as an upper bound forpolysemy-zooon versions 0.6.0.0 and 0.6.0.1 on Hackage? Otherwise the migration ofTaggedwill cause conflicts.Ok, will do
Also, I think every version of
polysemy-pluginbefore0.2.5.0also now needspolysemy-1.3.0.0as an upper bound due to the change in internals with the membership proof rewrite.they need
polysemy-1.3.0.0as an upper bound, sorry, that was a egregiously bad omissionHaha, I don't have access to
polysemy-zoo:big_smile:nor
plugin@Sandy Maguire ? :slight_smile:
:upside_down: whoops
Is it possible for multiple people to have access? If so, perhaps the entire team should be given it.
I don't have any experience, but I'm sure I can figure it out.
It seems like I can add people to
polysemy- do you have hackage account?Just created one.
KingoftheHomeless
@Love Waern (King of the Homeless) Now you're officially maintainer of
polysemy! :stuck_out_tongue_wink::tada: thanks!
Merged in #311 and the v1.3 pr. Once we can fix the bounds for the older versions of
polysemy-zooandpolysemy-plugin, we'll cut new releases for all three.Unless something crops up
just made yall owners of the research org
@Sandy Maguire We still aren't maintainers of
polysemy-pluginorpolysemy-zooon Hackage, though, which is the issue.(Or at least I'm not)
fixed
sorry about that!
@Sandy Maguire Thanks :slight_smile:
I'm gonna start on updating the bounds on
polysemy-zooandpolysemy-plugin.Done! All left now before cutting the releases is to merge #62 of polysemy-zoo, and before I do that I'm gonna check if there are any dependencies (other than
hedis) that have been made redundant with the removal of Redis support inpolysemy-zoo.zoo-#62 merged. We may now cut the releases.
Oh right, I'm not part of the uploader group yet. I've sent away a mail to the hackage trustees, but I can't cut the releases myself just yet. If anyone else wants to do the work, feel free, otherwise I'll do it when I can.
no? i made you a maintainer. that's not enough?
Tried to upload a package candidate and got 403:d.
Now part of the uploader list. Too tired to cut the releases now though, I'll do it tomorrow.
All releases cut. Documentation for polysemy-zoo failed to build on Hackage, so we need to upload our own documentation for it, and I'm not sure how that system works.
Does it build locally just fine? Does Hackage show you some logs?
It builds fine locally. Hackage logs don't tell me much.
https://github.com/ekmett/lens/blob/master/scripts/hackage-docs.sh
Stack uses cabal, right? cabal's apparently not in my PATH.
Whatever, I'll just install
cabal-install.It uses
Caballibrary, notcabal-installitselfDocumentation up for
polysemy-zoo. That should be everything for this release.Thanks! :slight_smile: