summaryrefslogtreecommitdiff
path: root/doc/forum/functions_that_yield_properties/comment_3_76f4a92cf26ae2fcc3152a0f1a19f516._comment
blob: 7b1954bb0674631f98fc331f38261dfceb31dcb1 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
[[!comment format=mdwn
 username="spwhitton"
 subject="comment 3"
 date="2016-06-05T06:13:05Z"
 content="""
> The type of this will be somewhat more complex than the one you gave, but it should work.

GHC's inferred type is not something I can understand, and I suspect that it is far more general than it needs to be.  In this sort of situation, are their strategies one can employ to write a sensible type signature?  I think that the only thing I need to restrict is avoiding trying to ensure properties with info.

> You might be able to finesse this by using a monoidial value and get the description of mkp mempty.

Could you expand a little on this suggestion, please?  I want to be able to use unmodified core properties like `User.accountFor`, and that takes a non-monoidal `User`.

> Or, you could do something like this to tie the knot. I don't know if this is a good idea (it might even <<loop>>), but it illustrates the core problem nicely; to get at the Info, we need a Host, but to get a Host, we need to already know its properties.

This seems to work!
"""]]