Very little changes. You can look at the changes to Parser.y on the Phabricator revision. In order to get this parsing sans any shift/reduce conflicts you have to introduce fexp2 as well as fexp, which makes it so that dollar-less lambdas/do blocks can only be applied as the last argument.

-- Andrew

On Mon, Sep 7, 2015 at 6:39 PM, Edward Kmett <ekmett@gmail.com> wrote:

Reserving judgment on the merits of the proposal, I would be curious to see the net impact of such a change on the grammar that GHC parses. How much changes?

-Edward

On Mon, Sep 7, 2015 at 9:15 PM, Mike Meyer <mwm@mired.org> wrote:
On Mon, Sep 7, 2015 at 8:08 PM Dan Burton <danburton.email@gmail.com> wrote:
+1 people who like it can use it and people who don't like it don't have to use it.

Yes, but they still have to be able to read code that uses it, deal with it in coding standards, deal with how it affects their tool chain, and otherwise have it affect them.

"If you don't like it, don't use it" is a bogus argument, and should be shot on sight.

_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe@haskell.org
http://mail.haskell.org/cgi-bin/mailman/listinfo/haskell-cafe