[ghc-steering-committee] Re: Qualified *

Yes on qualified strings from me as well. On Mon, Oct 6, 2025 at 3:47 AM Arnaud Spiwack wrote:

I vote yes for qualified strings (apologies, I thought I already had).

On Sat, 4 Oct 2025 at 05:28, Adam Gundry wrote:

...

Apologies that I've been struggling to find time to page all this in. I think:

* QualifiedStrings is an unqualified (ha ha) "yes" and we should accept that part as-is.

* QualifiedNumerics seems almost to have converged on a consensus design, with the sole remaining question being whether the desugaring uses `negate` even when NegativeLiterals is disabled. So I think we could probably accept conditionally on that change?

* QualifiedLists does seem to need a little more thought.

I like the general rule Sebastian proposes ( https://github.com/ghc-proposals/ghc-proposals/pull/724#issuecomment-3359455... ):

...

If M.<stx> is syntax enabled by a -XQualified<ext> extension, then any overloaded function f in the expansion of <stx> (under XOverloaded<ext>, if that exists) will be rebound to M.f instead.

Ideally we'd extend this rule to RebindableSyntax as well, something like:

...

When RebindableSyntax is in effect, then any overloaded function f in the expansion of <stx> will be rebound to whatever f is in scope.

This means everything behaves simply and predictably. The downside is that we are constrained by the existing design (of overloaded numerics/lists in particular), so we may not want to enforce this as a hard rule immediately, but I think we can treat it as a goal to aim for, and think about suitable migration strategies for getting to that point without an unreasonable level of breakage.

Cheers,

Adam

On 03/10/2025 07:23, Erik de Castro Lopo wrote:

...

Erik de Castro Lopo wrote:

...

Hi all,

Trying to drive this forward.

The general consensus seems to be to accept QualifiedStrings. I would suggest that we accept this on October 7th unless someone yells before then.

There seems to be quite a few of reservations on QualifiedNumerics, with the main sticking point being the treatment of negative literals and natural numbers. I will consult with the proposer on this.

For QualifiedLists it seems we are even further from consensus. Again, I will consult with the proposer on this.

Hi,

For the QualifiedStrings proposal I have "accept" votes from Simon Peyton Jones, Malte Ott, Matthías Páll Gissurarson, Sebastian Graf, Erik de Castro Lopo and Moritz Angermann.

I have not seen a response from Simon Marlow, Eric Seidel, Arnaud Spiwack, Arnaud Spiwack or Jakob Brünker. I have CCed the ones I do have email addresses for, but have no address for Eric Seidel or Jakob Brünker.

I am hoping to get the voting for QualifiedStrings done by October 7th.

The other two related proposals well be sent back to the proposer for review.

Thanks, Erik

-- Adam Gundry, Haskell Consultant Well-Typed LLP, https://www.well-typed.com/

Registered in England & Wales, OC335890 27 Old Gloucester Street, London WC1N 3AX, England

_______________________________________________ ghc-steering-committee mailing list -- ghc-steering-committee@haskell.org To unsubscribe send an email to ghc-steering-committee-leave@haskell.org

-- Arnaud Spiwack Director, Research at https://moduscreate.com and https://tweag.io. _______________________________________________ ghc-steering-committee mailing list -- ghc-steering-committee@haskell.org To unsubscribe send an email to ghc-steering-committee-leave@haskell.org