Re: Hi. I'm new to this mailing list and have a few questions.

Your idea is a good one. To add to Iavor's mention about Cryptol, Agda uses both positional and named parameters, and they seem to have a good time with it. To be honest, if I could go back in time, I would think much harder about the decision to do ordered parameters in GHC. I'm very sympathetic to this direction of travel, but unsure about the costs (code breakage, etc.) of doing so. Do comment on the proposal thread, Anselm! :) We need motivated people like you to keep those conversations going. Richard PS: And, yes, I believe everyone here is on a casual, first-name basis.

On Aug 4, 2020, at 2:37 PM, Brandon Allbery wrote:

It's represented the same way but handled differently; my point, though, was more about it being consistent with the rest of the language, which is less about how the compiler handles it than about how it looks to the programmer. In this case, _ is consistently used to mean a wildcard of some kind, even though it has to be handled by the compiler in multiple places including TypeApplications and PartialTypeSignatures (others include patterns, typed holes, and let bindings as an extension of simple patterns).

On Tue, Aug 4, 2020 at 2:27 PM Anselm Schüler (conversations subemail) mailto:conversation@anselmschueler.com> wrote: That’s true, but type wildcards without PartialTypeSignatures throw an error, while this doesn’t need that (which suggests that they’re represented differently internally — I don’t know, of course). Also, there are other drawbacks discussed in the ghc-proposals Proposal 0042-named-type-args https://github.com/ghc-proposals/ghc-proposals/pull/196 that was mentioned by Vladislav Zavialov mailto:vladislav@serokell.io. Which is, I think, it seems, where the discussion should take place (?).

Anselm Schüler

www.anselmschueler.com http://www.anselmschueler.com/ mail@anselmschueler.com mailto:mail@anselmschueler.com

From: Brandon Allbery mailto:allbery.b@gmail.com Sent: Tuesday, August 4, 2020 20:16 To: Anselm Schüler (conversations subemail) mailto:conversation@anselmschueler.com Cc: Rowan Goemans mailto:goemansrowan@gmail.com; ghc-devs@haskell.org mailto:ghc-devs@haskell.org Subject: Re: AW: RE: Hi. I'm new to this mailing list and have a few questions.

_ already has the meaning of a type wildcard (generalized from its use as a wildcard in patterns), so this

is consistent with its use in other type signatures. The problem I see with your proposal is that it assumes

you know the names of the type variables in the original declaration.

On Tue, Aug 4, 2020 at 2:12 PM Anselm Schüler (conversations subemail) mailto:conversation@anselmschueler.com> wrote:

And also, it’s really weird and seems like a feature that should be removed IMO. It kind of looks like the type variable _ is always in-scope, and it’s unclear if the type of the resulting function is just type variables without an explicit forall or if _ has a special meaning.

Anselm Schüler

www.anselmschueler.com http://www.anselmschueler.com/

From: Anselm Schüler (conversations subemail) mailto:conversation@anselmschueler.com Sent: Tuesday, August 4, 2020 20:09 To: Rowan Goemans mailto:goemansrowan@gmail.com; ghc-devs@haskell.org mailto:ghc-devs@haskell.org Subject: RE: AW: RE: Hi. I'm new to this mailing list and have a few questions.

True, but this makes the type less readable since id @_ :: _ -> _.

Anselm Schüler

www.anselmschueler.com http://www.anselmschueler.com/

From: Rowan Goemans mailto:goemansrowan@gmail.com Sent: Tuesday, August 4, 2020 19:59 To: ghc-devs@haskell.org mailto:ghc-devs@haskellorg Subject: Re: AW: RE: Hi. I'm new to this mailing list and have a few questions.

Just as a heads up. You can achieve what you want already by doing:

f @_ @([String]) :: forall a. (a, [String]) -> ([String], a)

Best regards,

Rowan Goemans

On 8/4/20 7:45 PM, Anselm Schüler (conversations subemail) wrote:

Thank you for the nice introduction :) !

I will check out the GHC proposals site.

And following Simon’s (I hope addressing with first name is OK) suggestion, I’m going to give an outline of the idea.

The idea is to extend type application syntax to enable explicit assignment of types to specific type variables.

For instance, say I have f :: forall a b. (a, b) -> (b, a), and I want to apply the type [String] to it. My only option is to do

f @([String]) :: forall b. ([String], b) -> (b, [String])

—but what if, instead, I want a function of type forall a. (a, [String]) -> ([String], a)?

I propose the following syntax:

f @{b = [String]} :: forall a. ([String], b) -> (b, [String])

This wouldn’t break any existing programs since using record syntax here is already disallowed and met with an error message.

A question is of course the symbol used for assignment (~, =, ::, or ->?).

I hope the code shows up as a monospace font on your end. I used the IBM Plex Mono font, which is open-source https://github.com/IBM/plex.

Anselm Schüler

wwwanselmschueler.com http://www.anselmschueler.com/ mail@anselmschueler.com mailto:mail@anselmschueler.com

From: Simon Peyton Jones mailto:simonpj@microsoft.com Sent: Tuesday, August 4, 2020 18:44 To: Richard Eisenberg mailto:rae@richarde.dev; "Anselm Schüler (conversations subemail)" mailto:conversation@anselmschueler.com Cc: ghc-devs@haskell.org mailto:ghc-devs@haskell.org Subject: RE: Hi. I'm new to this mailing list and have a few questions.

Welcome Anselm ghc-devs is a very informal mailing list, and we welcome newcomers.

For example, I have a feature idea in the back of my mind, which I imagine would be easy to implement

What Richard says is right, but you should feel free to fly the kite on this list if you want – or on Haskell Café – to get some idea of whether others seem warm about the idea, before writing a full proposal.

Simon

From: ghc-devs mailto:ghc-devs-bounces@haskell.org On Behalf Of Richard Eisenberg Sent: 04 August 2020 16:05 To: "Anselm Schüler (conversations subemail)" mailto:conversation@anselmschueler.com Cc: ghc-devs@haskell.org mailto:ghc-devs@haskell.org Subject: Re: Hi. I'm new to this mailing list and have a few questions.

Hi Anselm,

Welcome!

A good way of getting used to a list like this one is to wait a little while and observe what kind of messages others send; this will give you a feel for how the list is used. If you're impatient, you can also check out the archives at https://mail.haskell.org/pipermail/ghc-devs/ https://nam06.safelinks.protection.outlook.com/?url=https%3A%2F%2Fmail.haskell.org%2Fpipermail%2Fghc-devs%2F&data=02%7C01%7Csimonpj%40microsoft.com%7Ce68e6131af494519061c08d83887d962%7C72f988bf86f141af91ab2d7cd011db47%7C1%7C0%7C637321503404154244&sdata=w4QIMTGeMQ35bEo4xkOn2rIj8z02iMVje4Ar9AXM22k%3D&reserved=0.

As for a feature request: if your feature changes the language GHC accepts (most do), the right place to post is at https://githubcom/ghc-proposals/ghc-proposals https://nam06.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fghc-proposals%2Fghc-proposals&data=02%7C01%7Csimonpj%40microsoft.com%7Ce68e6131af494519061c08d83887d962%7C72f988bf86f141af91ab2d7cd011db47%7C1%7C0%7C637321503404164244&sdata=ogVVk5kRFXPc81dZ%2FiyGV9FbgQoQUtzazpFGcrAKUXA%3D&reserved=0. There is a description of how to proceed on that page. Proposals submitted there get debated within the community and then eventually sent to a GHC Steering Committee for a vote on acceptance or rejection. Then, we worry about implementing it. If you have a suggestion that does not change the language GHC accepts, you can post an Issue at https://gitlab.haskell.org/ghc/ghc/ https://nam06.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgitlab.haskell.org%2Fghc%2Fghc%2F&data=02%7C01%7Csimonpj%40microsoft.com%7Ce68e6131af494519061c08d83887d962%7C72f988bf86f141af91ab2d7cd011db47%7C1%7C0%7C637321503404164244&sdata=tg40q%2B9%2BW5KcVzHsPx9oCP7wdw5ipdZCOOdFnzUz3mU%3D&reserved=0.

I hope this is helpful!

Richard

On Aug 4, 2020, at 8:59 AM, Anselm Schüler (conversations subemail) mailto:conversation@anselmschueler.com> wrote:

First of all, in general, I’m new to mailing lists (as used for discussions) in general, so a question about that:

When subscribed to the mailing list, do you get every message, or are some discussions hidden?

Second of all, I’d like to know what kinds of messages are appropriate here. I’m not familiar with coding compilers or anything of the like, so I’m somewhat afraid of offering unhelpful comments or being just woefully underqualified to participate here.

For example, I have a feature idea in the back of my mind, which I imagine would be easy to implement (that might be wrong). Is it alright if I submit that here or should I use some other forum?

Thank you in advance for the answers.

Anselm Schüler

www.anselmschueler.com https://nam06.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.anselmschueler.com%2F&data=02%7C01%7Csimonpj%40microsoft.com%7Ce68e6131af494519061c08d83887d962%7C72f988bf86f141af91ab2d7cd011db47%7C1%7C0%7C637321503404164244&sdata=WJBJ3P%2BFUnTmVs9Z97DHLat9mHr6vC22l0mQ36nafGc%3D&reserved=0 mail@anselmschueler.com mailto:mail@anselmschueler.com _______________________________________________ ghc-devs mailing list ghc-devs@haskell.org mailto:ghc-devs@haskell.org http://mail.haskell.org/cgi-bin/mailman/listinfo/ghc-devs https://nam06.safelinks.protection.outlook.com/?url=http%3A%2F%2Fmail.haskell.org%2Fcgi-bin%2Fmailman%2Flistinfo%2Fghc-devs&data=02%7C01%7Csimonpj%40microsoft.com%7Ce68e6131af494519061c08d83887d962%7C72f988bf86f141af91ab2d7cd011db47%7C1%7C0%7C637321503404174237&sdata=9y99aKPjfvm7WIH5Xv9JDlN7TIXdVPJZF%2BvAgbUvBUU%3D&reserved=0

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

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

--

brandon s allbery kf8nh

allbery.b@gmail.com mailto:allbery.b@gmail.com

-- brandon s allbery kf8nh allbery.b@gmail.com mailto:allbery.b@gmail.com_______________________________________________ ghc-devs mailing list ghc-devs@haskell.org mailto:ghc-devs@haskell.org http://mail.haskell.org/cgi-bin/mailman/listinfo/ghc-devs http://mail.haskell.org/cgi-bin/mailman/listinfo/ghc-devs