On 30/09/2026 17:09, Nathan Bossart wrote:
> On Tue, Sep 29, 2026 at 11:51:39PM +0200, Vik Fearing wrote:
>> On 29/09/2026 22:23, Zsolt Parragi wrote:
>>> And +1 for using %nonassoc instead
>> That's three now. I'll prepare a new patch.
> Thanks. One other thing I'd like to explore a bit more is whether to make
> IMPLIES "round-trippable," at least to see how much more complicated it
> would be. (I'm happy to do this part BTW).
>
Attached is a new patch set that makes IMPLIES non-associative, changes
the tests, and also survives a round trip. I put the latter in a
separate patch in case it isn't wanted after all.
I think I have incorporated everyone's feedback, but let me know if I
missed something.
--
Vik Fearing
