Closed joneshf closed 5 years ago
This looks great to me :+1:
Do you think it's also worth mentioning the (<~<)
/ (>~>)
codec composition operators and that sometimes even profunctor
is not enough? (In that remapping that way only allows the migration to succeed, you need to dip down even deeper to construct a codec that can fail in a new way). I can write this part, just wondered if you thought that info belongs here too.
For sure! It took me longer than I like to admit that you have to use (<~<)
/(>~>)
to make migrations work (multiple days). Anything in that regard would be useful.
Eep, sorry about that! I should perhaps re-export those operators from this module too. I'll make a new PR with the further changes anyway. Thanks!
No worries. Using this package has been really nice. Thanks for making it!.
We discussed on Slack about documenting the motivation behind the
Data.Codec.Argonaut.Migration
module. Here's what I realized while discussing yesterday:Functor
instance forData.Codec.GCodec _ _ _
.Profunctor
instance forData.Codec.GCodec _ _
.I tried to capture those realizations. Let me know if there's different wording or if I missed the mark. More than happy to make any changes.