[Haskell-cafe] Call for discussion: OverloadedLists extension

[Michael Snoyman]

On Sun, Sep 23, 2012 at 5:51 PM, Chris Smith <cdsmith at gmail.com> wrote:
> Michael Snoyman <michael at snoyman.com> wrote:
>> That said, it would be great to come up with ways to mitigate the
>> downsides of unbounded polymorphism that you bring up. One idea I've
>> seen mentioned before is to modify these extension so that they target
>> a specific instance of IsString/IsList, e.g.:
>>
>> {-# STRING_LITERALS_AS Text #-}
>>
>> "foo" ==> (fromString "foo" :: Text)
>
> That makes sense for OverloadedStrings, but probably not for
> OverloadedLists or overloaded numbers... String literals have the
> benefit that there's one type that you probably always really meant.
> The cases where you really wanted [Char] or ByteString are rare.  On
> the other hand, there really is no sensible "I always want this"
> answer for lists or numbers.  It seems like a kludge to do it
> per-module if each module is going to give different answers most of
> the time.
>
> --
> Chris

Note that I wasn't necessarily advocating such a pragma. And a lot of
my XML code actually *does* use two IsString instances at the same
time, e.g.:

    Element ("img" :: Name) (singleton ("href" :: Name) ("foo.png" ::
Text)) [NodeComment ("No content inside an image" :: Text)]

(Courtesy of xml-conduit.)

To prove your point even further, with OverloadedLists we could
replace that `singleton` call with `[("href", "foo.png")]` and then be
using two `IsList` instances simultaneously as well (`Map` and `[]`).

Also, I use the `ByteString` instance of `IsString` regularly when
using `http-conduit` and `warp` (for all of the header values), and to
an even greater extent when hacking on the internals of any HTTP
library (whether `http-conduit` or something in the `wai` ecosystem).

Michael