A few small fixes #553
A few small fixes #553
Conversation
|
When is fuzzy-completions called with non-simple strings? |
|
I'm currently adding some other communication protocols; eg. |
|
Then shouldn't it be doing all the appropriate coercions then? |
|
Well, wouldn't that mean most of that translation effort would be wasted? Most functions just accept non-simple strings already. Yes, of course I could put a special case for that one call in... but that's really awful, code-wise. |
|
As it stands, it's equally as awful to make a change for code that doesn't exist. |
| @@ -8,7 +8,7 @@ | |||
| ;;; are disclaimed. | |||
| ;;; | |||
|
|
|||
| (in-package swank/rpc) | |||
| (in-package :swank/rpc) | |||
There was a problem hiding this comment.
Helmut actually meant to write it like that. If we change it here, we should then change it in a bunch of other places.
There was a problem hiding this comment.
As it stands, it's equally as awful to make a change for code that doesn't exist.
Well, this is a single place that doesn't accept a string but has more narrow restrictions.
I think ensuring that type within the same library isn't wrong.
Helmut actually meant to write it like that. If we change it here, we should then change it in a bunch of other places.
I just noticed that even within swank it's not one-or-another. Should I amend my PR to change all over to keywords, as it seems to be custom (data are my QL libraries)?
Thanks for considering!