October 2021 - ghc-steering-committee

#425: invisible binders in type declarations; rec: accept
by Richard Eisenberg 07 Apr '22

07 Apr '22

Hi all, I am the shepherd for proposal #425, https://github.com/ghc-proposals/ghc-proposals/pull/425, proposing to add invisible binders in type declarations. The main payload of the proposal is to allow definitions like > data Proxy @k (a :: k) = Proxy instead of today's > data Proxy (a :: k) = Proxy which has no explicit binding site for k. This new syntax solves a number of smallish syntax conundra, as very well outlined in the proposal. In addition, the proposal includes two small unrelated tweaks to the syntax of type family instances; these are points (6) and (7) in the proposal. Both changes will break some obscure (but still realistic, knowing Haskell) programs, but both fixes are backward compatible. --- I recommend acceptance. The proposal is motivated nicely (do check out the examples) and solves a real problem. The new syntax fits in with other similar features. The small cleanups to existing syntax will lead to better error messages. The one drawback, as I see it, is that this proposal includes point (5), which is a breaking change to the way type synonyms work. Right now, we can say > type P = (Proxy :: k -> Type) and GHC will infer P :: forall k. k -> Type. Under this proposal, you would have to write > type P @k = (Proxy :: k -> Type) bringing k into scope explicitly. The fix is not backward compatible, and so I think this proposal should come with a migration strategy, where we warn about the former version for some releases before banning it. (Continuing to support it is possible, but it's very awkward to have a variable mentioned only in a synonym's right-hand side.) Sorry for the delay in producing this recommendation! Richard

5 8

Please review #392: Clarify modifiers design principle, Shepherd: Alejandro
by Joachim Breitner 28 Jan '22

28 Jan '22

Dear Committe, Clarify modifiers design principle has been proposed by Richard https://github.com/ghc-proposals/ghc-proposals/pull/392 This is an amendmend to #370, see the PR description for links to diffs etc. I propose Alejandro as the shepherd, as he shepherded #370 before. Please guide us to a conclusion as outlined in https://github.com/ghc-proposals/ghc-proposals#committee-process Thanks, Joachim -- -- Joachim Breitner mail(a)joachim-breitner.de http://www.joachim-breitner.de/

4 14

Please review #319: NoFallibleDo proposal, Shepherd: Eric Seidel
by Joachim Breitner 08 Jan '22

08 Jan '22

Dear Committee, this is your secretary speaking: NoFallibleDo proposal has been proposed by Cale https://github.com/ghc-proposals/ghc-proposals/pull/319 https://github.com/obsidiansystems/ghc-proposals/blob/no-fallible-do/propos… I propose Eric Seidel as the shepherd. Please guide us to a conclusion as outlined in https://github.com/ghc-proposals/ghc-proposals#committee-process Thanks, Joachim -- Joachim Breitner mail(a)joachim-breitner.de http://www.joachim-breitner.de/

6 18

Proposal #281: Visible "forall" in terms; rec: accept
by Richard Eisenberg 01 Nov '21

01 Nov '21

Hi committee, Proposal #281 has been submitted for our consideration. Proposal PR: https://github.com/ghc-proposals/ghc-proposals/pull/281 <https://github.com/ghc-proposals/ghc-proposals/pull/281> Proposal text: https://github.com/int-index/ghc-proposals/blob/visible-forall/proposals/00… <https://github.com/int-index/ghc-proposals/blob/visible-forall/proposals/00…> The text of the proposal is long and detailed, but do not be daunted: it is simpler in practice than it appears. The main payload of the proposal is: Introduce a new extension -XRequiredTypeArguments. With this extension enabled, the `forall ... ->` syntax, currently in existence in kinds, is now available in types (of ordinary functions). This means that function definitions and call sites are sometimes required to write a type argument. The type argument is *not* preceded by @ or any other syntactic marker. This is useful for defining what would otherwise be ambiguous types. Example: > sizeof :: forall a -> Sizeable a => Int > sizeof = ... > > intWidth = sizeof Int There are further examples and motivation in the proposal. The rest of the proposal is simply about dealing with odd corner cases that come up with the main payload. In particular, mixing types in with terms with no syntactic signifier means that we must be careful about parsing and namespaces. If a type argument is written in the syntax that is shared between types and terms (including function application!) and uses identifiers in scope in only one of the two namespaces, nothing unusual can be observed. But, of course, there are corner cases. Here are some of the salient details: - Define type-syntax and term-syntax, where the choice of syntax is always driven by syntactic markers, such as :: or @. See the dependent types proposal <https://github.com/goldfirere/ghc-proposals/blob/dependent-types/proposals/…> for more details. Parsing and name-resolution are controlled by whether a phrase is in type-syntax or term-syntax. For name resolution, if a lookup fails in the first namespace (the term-level namespace in term-syntax or the type-level namespace in type-syntax), we try the other namespace before failing. - Because term- vs type-syntax is controlled by syntax, a required type argument is in *term*-syntax and gets name-resolved *as a term*. In the absence of punning, this works out fine, but it is possible that a punned identifier will cause confusion. The proposal includes section 4.3 allowing users to write `type` to signify a switch to type-syntax. - The proposal also includes a way to avoid punning for the built-in types with privileged syntax: lists and tuples. This method allows users to specify -XNoListTupleTypeSyntax to disable the list and tuple syntax in types (but still allows it for terms). The proposal also suggests exporting type List = [] from Data.List and other synonyms for tuples from Data.Tuple. --------------- I recommend acceptance. When doing type-level programming, the lack of this feature is strange, leading to ambiguous types and easy-to-forget arguments and easy-to-make type errors. The design space here is difficult, but this proposal is very much in keeping with the design sketch of our recently-accepted #378, in particular its section on this point <https://github.com/goldfirere/ghc-proposals/blob/dependent-types/proposals/…>. I believe the design described here is both backward compatible with what we have today (users who do not use this feature will not notice a difference) and forward compatible with a cohesive design for dependent types. There are several optional pieces: - The `type` herald <https://github.com/int-index/ghc-proposals/blob/visible-forall/proposals/00…>. I am unsure about this one, but others have felt strongly in favor, and I have no reason to object. - Types-in-terms <https://github.com/int-index/ghc-proposals/blob/visible-forall/proposals/00…>. I think this is necessary in order to avoid annoying definitions of type synonyms for one-off usage sites. It is a straightforward extension of the term-level parser to allow previously type-level-only constructs. It is necessary in order for us to achieve the vision of dependent types in #378. The only challenge here is that this requires us to make `forall` an unconditional keyword in terms. This does pose a backward-compatibility problem. I see, for example, that the sbv package exports a function named `forall`, so we may need to think more carefully about how to proceed here -- possibly by guarding the keyword-ness of `forall` behind the extension for some number of transitionary releases. - Lists and Tuples <https://github.com/int-index/ghc-proposals/blob/visible-forall/proposals/00…>. This section describes the -XNoListTupleTypeSyntax extension. I am not convinced that this change needs to be part of this proposal (thinking it belongs more in #270), but I do think we'll need it in the end. Is it OK to export new type synonyms from Data.List and Data.Tuple? Not sure, though I'd like these exported from some central place. What do others think? Thanks, Richard

7 31

Please review #413: GHC Maintainer Preview, Shepherd: Vlad
by Joachim Breitner 31 Oct '21

31 Oct '21

Dear Committee, this is your secretary speaking: GHC Maintainer Preview, a change to GHC's release process, has been proposed by Hécate Moonlight. https://github.com/ghc-proposals/ghc-proposals/pull/413 https://github.com/Kleidukos/ghc-proposals/blob/patch-1/proposals/0000-ghc-… I propose Vladislav as the shepherd. Please guide us to a conclusion as outlined in https://github.com/ghc-proposals/ghc-proposals#committee-process Thanks, Joachim -- Joachim Breitner mail(a)joachim-breitner.de http://www.joachim-breitner.de/

2 1

#433: the Unsatisfiable constraint; recommendation: accept
by Tom Harding 27 Oct '21

27 Oct '21

Hi all, Firstly, apologies for the radio silence - I’ve apparently been having some undiscovered email trouble (perhaps even since joining the committee), and so I’m now going to be using my usual, and hopefully far more reliable, email address. With that said, I’d like to recommend proposal #433 <https://github.com/ghc-proposals/ghc-proposals/pull/433> for acceptance. The proposal does a good job of outlining the difference between this and the current TypeError mechanism, and I think the case is well justified. Certainly, the ability to bypass fundep checks in error case “instances” immediately brought to mind a handful of places in my own libraries that could benefit from this check. I think the thread already contains a lot of committee activity, but I welcome any further comments! Thanks, Tom

4 3

#283: Local modules (again), recommendation: accept
by Spiwack, Arnaud 18 Oct '21

18 Oct '21

Dear all, It is time to review the local modules proposal, from our own Richard Eisenberg, again. In fact it’s been time for a while, and I must apologise to Richard for not having understood that. Sorry, Richard. Link: https://github.com/ghc-proposals/ghc-proposals/pull/283 Since the last time, the proposal has been both simplified and made more precise. The main issues solved by this proposal are - The ability to export qualified names from modules - The ability to define functions with qualified names within a module ( *e.g.* you could have both T.f and U.f). - The ability to locally refer to qualified data as unqualified (suppose Set.map is in scope, then the proposal lets you write let import module Set in map f (map g some_set). I think these are three important issues, and the proposal makes a good case that they ought to be solved together. I also think that the specification is clear and doesn’t leave any dark corner. Caveat: I have contributed to writing the current specification, so I may be missing some flaws. The proposal does introduce a somewhat byzantine semantic difference between module X and module qualified X in export lists. This is to keep the current meaning of module X which is already not very intuitive (cf https://ro-che.info/articles/2012-12-25-haskell-module-system-p1 ). I think this is unavoidable. There are also 3 optional pieces of specification that we are asked to rule on (§4). I’m personally in favour of option 1. I’m weakly in favour of option 3 (with a preference for the variant described in Alternative 7). On the other hand option 2 seems to add a bit too much syntactic subtlety to be worth it (though if we choose option 2, I see no reason not to accept Alternative 3 as well). /Arnaud

8 23

Please review #433: the Unsatisfiable constraint, Shepherd: Tom
by Joachim Breitner 16 Oct '21

16 Oct '21

Dear Committee, this is your secretary speaking: Unsatisfiable has been proposed by Adam Gundry https://github.com/ghc-proposals/ghc-proposals/pull/433 https://github.com/adamgundry/ghc-proposals/blob/unsatisfiable/proposals/00… I propose Tom Harding as the shepherd. Please guide us to a conclusion as outlined in https://github.com/ghc-proposals/ghc-proposals#committee-process Thanks, Joachim -- Joachim Breitner mail(a)joachim-breitner.de http://www.joachim-breitner.de/

1 0

Visible forall: 3-Year Anniversary
by Vladislav Zavialov (int-index) 13 Oct '21

13 Oct '21

Dear GHC Steering Committee, Today (10 Oct, 2021) marks three years since the pull request for the #281 “Visible forall” proposal was opened. https://github.com/ghc-proposals/ghc-proposals/pull/281 There has been a lot of discussion, which led to numerous refinements and improvements to the proposal text and its content. When the initial version of the proposal was written, it wasn’t clear if this feature would be a good addition to Haskell. But now that #378 “Design for dependent types” has been accepted, the job of #281 “Visible forall” is merely to elaborate the details of a design that we already agreed on. Because of its reduced scope, I hope there won’t be a 4th anniversary. So I invite everyone to take a look at the current iteration of the proposal and vote on it. The proposal text is at: https://github.com/int-index/ghc-proposals/blob/visible-forall/proposals/00… - Vlad

3 3

GHC20xx review; GHC2022 yes or no?
by Joachim Breitner 11 Oct '21

11 Oct '21

Hi Committee, when we defined the process for GHC20xx, as written in https://github.com/ghc-proposals/ghc-proposals/blob/master/proposals/0372-g… we wrote Likely, the first iteration of this process will be vastly different from the following ones: The first one is expected to add a large number of uncontroversial extensions; so the next iteration will likely only make a smaller, but more controversial change. Therefore, this proposal does not commit to a fixed cadence. Instead, 6 months after the first release of a version of GHC that supports a GHC20xx set, we evaluate the outcome, the process, and the perceived need of a next release. At that time we will refine the processes, if needed, and set a cadence. The first version of GHC that supported GHC20xx is 9.2, released in March. So we should do this evaluation now. My impression is that 9.2 hasn’t reached the masses yet: NixOS stable doesn’t even have it, and the default is at 8.10. Stackage LTS is at 8.10, and nightly at 9.0 Debian is at 8.8. So it seems premature to try to evaluate its impact, and what, if anything, we should do differently for a hypothetical GHC2022. So I suggest we postpone this review for another half year or so. Also, I don’t see why GHC2022 would be different than GHC2021. Not that much has changed about GHC and its set of “should-be-default” extensions since last year, has it? So I suggest we don’t work on defining GHC2022, and the next update will be GHC2023 (or later). What do you think, Joachim -- Joachim Breitner mail(a)joachim-breitner.de http://www.joachim-breitner.de/

5 5