When I load cl-async SBCL raises a bunch of warnings:
...
WARNING: redefining DEFTYPE type to be a class: CL-ASYNC:SOCKET-CLOSED
WARNING: redefining DEFTYPE type to be a class: CL-ASYNC:TCP-EOF
WARNING: redefining DEFTYPE type to be a class: CL-ASYNC:TCP-INFO
WARNING: redefining DEFTYPE type to be a class: CL-ASYNC:TCP-ERROR
WARNING: redefining DEFTYPE type to be a class: CL-ASYNC:TCP-RESET
WARNING: redefining DEFTYPE type to be a class: CL-ASYNC:TCP-TIMEOUT
WARNING: redefining DEFTYPE type to be a class: CL-ASYNC:TCP-REFUSED
WARNING: redefining DEFTYPE type to be a class: CL-ASYNC:TCP-ACCEPT-ERROR
I wonder if it would just be easier to redefine define-condition-alias as this instead:
(defmacro define-condition-alias (alias name)
"Define an alias for the specified condition."
`(define-condition ,alias (name)
()))
Yes, it adds a few more classes, but they are trivial, so maybe better do this instead of fighting the SBCL compiler?
If you think this is worthwhile, I would be happy to submit an MR. But I will wait to see what you say first.
When I load
cl-async
SBCL raises a bunch of warnings:I wonder if it would just be easier to redefine
define-condition-alias
as this instead:Yes, it adds a few more classes, but they are trivial, so maybe better do this instead of fighting the SBCL compiler? If you think this is worthwhile, I would be happy to submit an MR. But I will wait to see what you say first.