caml-list - the Caml user's mailing list
 help / color / mirror / Atom feed
From: Stefan Holdermans <stefan@vectorfabrics.com>
To: Abdallah Saffidine <abdallah.saffidine@gmail.com>
Cc: caml-list@inria.fr
Subject: Re: [Caml-list] How to use GADTs across modules in OCaml without raising warnings?
Date: Tue, 13 Oct 2015 08:06:39 +0200	[thread overview]
Message-ID: <B4753904-8571-4D43-AB3E-F12996C5ED7D@vectorfabrics.com> (raw)
In-Reply-To: <CALx9x4dgS_1_SFGu3Qn77UH8dw957A-s2h44Gri4UbRRBRQ6FQ@mail.gmail.com>

Abdallah,

> gadt1.ml:
> 
> type
>  never
> 
> type _ t1 = A1 : never  t1 | B1 :
>  bool t1
> 
> type _ t2 = A2 : string t2 | B2 :
>  bool t2
> 
> let get1 : bool t1 -> bool = function B1 -> true
> let get2 : bool t2 -> bool = function B2 -> true
> gadt2.ml:
> 
> let get1 : bool Gadt1.t1 -> bool = function Gadt.B1 -> true
> let get2 : bool Gadt1.t2 -> bool = function Gadt.B2 -> true
> when I compile using ocaml 4.02.3 (ocamlbuild gadt2.native), I get a warning 8 about the function Gadt2.get1 not being exhaustive. I am quite puzzled that Gadt2.get1 raises a warning while Gadt1.get1 and Gadt2.get2 don’t.

I suppose that, from the perspective of ‘Gadt2', ‘Gadt1.never’ is an abstract datatype that *could* have been implemented by ‘Gadt1’ as

  type never = bool

Perhaps there’s a design priciple emerging here? Concrete GADTs do not mix well with abstract type arguments.

Cheers,

  Stefan

  reply	other threads:[~2015-10-13  6:06 UTC|newest]

Thread overview: 7+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-10-13  5:27 Abdallah Saffidine
2015-10-13  6:06 ` Stefan Holdermans [this message]
2015-10-13  6:09 ` Philippe Veber
2015-10-13  6:15   ` Ben Millwood
2015-10-13  9:45     ` Jeremy Yallop
2015-10-14  2:13       ` Abdallah Saffidine
2015-10-13  6:10 ` Jacques Garrigue

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=B4753904-8571-4D43-AB3E-F12996C5ED7D@vectorfabrics.com \
    --to=stefan@vectorfabrics.com \
    --cc=abdallah.saffidine@gmail.com \
    --cc=caml-list@inria.fr \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).