On Tue, Nov 30, 2010 at 4:05 PM, David Menendez <span dir="ltr"><<a href="mailto:dave@zednenem.com">dave@zednenem.com</a>></span> wrote:<br><div class="gmail_quote"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
<div class="im">On Tue, Nov 30, 2010 at 5:31 AM, John Lato <<a href="mailto:jwlato@gmail.com">jwlato@gmail.com</a>> wrote:<br>
>> From: David Menendez <<a href="mailto:dave@zednenem.com">dave@zednenem.com</a>><br>
>><br>
</div><div class="im">>> Is Pointed useful at all? The last time this discussion came up, I<br>
>> asked for algorithms which were generic over pointed functors (in the<br>
>> same way that traverse is generic over applicative functors) and no<br>
>> one could think of any.<br>
><br>
> It's useful for 'singleton' on probabilistic data structures (e.g. a Bloom<br>
> filter).<br>
<br>
</div>Again, I'm asking for clients, not implementations. Are there any<br>
useful, non-trivial functions which are parameterized by arbitrary<br>
Pointed functors?<br></blockquote><div><br></div><div>I consider "singleton" itself to be a client, albeit somewhat trivial. But as a result of the last discussion, I'm no longer sure that "singleton" is a proper use of "pointed" and I shouldn't have brought it up now. I'm still in favor of splitting Applicative for philosophical reasons though.</div>
<div><br></div><div>In any case there's the function posted by Ross Patterson, which seems to fit all your criteria.</div><div><br></div><div>John</div></div>