[m-rev.] for review: Require subtypes to repeat existential type vars and constraints exactly.

[Zoltan Somogyi]

On 2024-11-13 12:10 +11:00 AEDT, "Zoltan Somogyi" <zoltan.somogyi at runbox.com> wrote:
>> 
>>     The reason for the restriction is simple:
>>     without it, how can one decide whether
>>     the argument of @samp{f2} is universally or existentially quantified?
>> 
>> It would be a reasonable design choice for the existential quantifier
>> to shadow the universal quantifier, making the argument of `f2'
>> existentially quantified. I suggest we delete that paragraph.
> 
> That design choice is one that one *could* make, but one that one
> *shouldn't* make. I will reword the last two thirds of that sentence
> to give the reason as shadowing being both potentially confusing to readers,
> and trivially avoidable.

Attached is the diff with the replacement wording.
Most of it is Peter's suggested diff.

Zoltan.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: DIFF.uev2
Type: application/octet-stream
Size: 1964 bytes
Desc: not available
URL: <http://lists.mercurylang.org/archives/reviews/attachments/20241113/33559766/attachment.obj>