You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The whole experience for writers improved quite a bit over the years. Basically we no longer have to define artificial types we can now write:
module M : sig
type t
module Sub : sig
(**/**)
type m := t
(**/**)
type t
val get_m : t -> m
end
end
And at render time that m reference is turned into a t that links to M.t. However from a reading perspective it can get confusing:
That second link goes to M.t.
Should maybe odoc render the fully qualified name in this case ?
For reference here's what happens if one does not hide the substitution, it's then quite clear what's happening but m links on the substitution and another name is still introduced:
The text was updated successfully, but these errors were encountered:
Would a tooltip with the complete path help? I'd rather something we can apply universally than just attempt to detect types that render the same but point to different targets (would we do the ambiguous check per value, per signature, per include?)
Doesn't really help with reading, I can already hover on the link and it shows me the destination. Basically for now I opted not to hide substitutions, it gets too confusing otherwise.
I'd rather something we can apply universally than just attempt to detect types that render the same but point to different targets (would we do the ambiguous check per value, per signature, per include?)
The idea would be to apply it on any hidden substitution regardless of whether it renders the same as something else.
The whole experience for writers improved quite a bit over the years. Basically we no longer have to define artificial types we can now write:
And at render time that
m
reference is turned into at
that links toM.t
. However from a reading perspective it can get confusing:That second link goes to
M.t
.Should maybe
odoc
render the fully qualified name in this case ?For reference here's what happens if one does not hide the substitution, it's then quite clear what's happening but
m
links on the substitution and another name is still introduced:The text was updated successfully, but these errors were encountered: