<div dir="ltr">The problem in the example is that the expression you abstract over (fst p) appears<div>in the type of another (non-variable) expression (snd p). The fix is to also abstract</div><div>over the second expression:</div><div><br></div><div><div>good-with : (p : Σ A B) → H (fst p) (snd p)</div><div>good-with p with fst p | snd p</div><div>... | x | y = {!!}</div></div><div><br></div><div>/ Ulf</div></div><div class="gmail_extra"><br><div class="gmail_quote">On Tue, Nov 8, 2016 at 11:24 PM, Jacques Carette <span dir="ltr"><<a href="mailto:carette@mcmaster.ca" target="_blank">carette@mcmaster.ca</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">The page <a href="https://agda.readthedocs.io/en/latest/language/with-abstraction.html" rel="noreferrer" target="_blank">https://agda.readthedocs.io/en<wbr>/latest/language/with-abstract<wbr>ion.html</a> is extremely informative (thanks!), but the subsection <a href="https://agda.readthedocs.io/en/latest/language/with-abstraction.html#ill-typed-with-abstractions" rel="noreferrer" target="_blank">https://agda.readthedocs.io/en<wbr>/latest/language/with-abstract<wbr>ion.html#ill-typed-with-<wbr>abstractions</a> is a bit problematic in that it leaves off the most important part: how to fix it!<br>
<br>
This is motivated by a real breakage (still the categories library). Try as I might, I simply cannot hand-write the correct "with" incantation to fix the one that used to work pre-2.5, and now is broken [1]. The example is very much like the one in the documentation linked to above, just "more so" in that there are many more type-level dependencies.<br>
<br>
If someone could hand-write the correct helper function that fixes the documentation, hopefully that would be a big enough hint that I can figure this out myself.<br>
<br>
Jacques<br>
<br>
[1] last line of <a href="https://github.com/copumpkin/categories/blob/master/Categories/NaturalIsomorphism.agda" rel="noreferrer" target="_blank">https://github.com/copumpkin/c<wbr>ategories/blob/master/Categori<wbr>es/NaturalIsomorphism.agda</a> for those who are curious.<br>
<br>
______________________________<wbr>_________________<br>
Agda mailing list<br>
<a href="mailto:Agda@lists.chalmers.se" target="_blank">Agda@lists.chalmers.se</a><br>
<a href="https://lists.chalmers.se/mailman/listinfo/agda" rel="noreferrer" target="_blank">https://lists.chalmers.se/mail<wbr>man/listinfo/agda</a><br>
</blockquote></div><br></div>