[Openmcl-devel] paserve error in openmcl

Gary Byers gb at clozure.com
Mon Jun 14 10:45:12 PDT 2004



On Mon, 14 Jun 2004, Cyrus Harmon wrote:

>
> I'm not sure if this is a paserve problem or an openmcl problem, but
> I'm getting the following:
>
>
> 127.0.0.1 - - [Mon, 14 Jun 2004 17:02:59 GMT] "GET / HTTP/1.1" 200 -1
>  > Error in process listener(1): :SITUATION is an invalid initarg to
> INITIALIZE-INSTANCE for #<STANDARD-CLASS SOCKET-ERROR>.
>  >                               Valid initargs: (:STREAM :ACTION :CODE
> :IDENTIFIER).
>
>
> One thing I notice is that there is a case-mismatch in l1-sockets.lisp:

Unescaped characters in symbols are mapped to uppercase by the reader
unless the readtable-case of the current readtable has been changed.
>
> (define-condition socket-error (simple-stream-error)
>    ((code :initarg :code :reader socket-error-code)
>     (identifier :initform :unknown :initarg :identifier :reader
> socket-error-identifier)
>     (Situation :initarg :situation :reader socket-error-situation)))
>
> (define-condition socket-creation-error (simple-error)
>    ((code :initarg :code :reader socket-creation-error-code)
>     (identifier :initform :unknown :initarg :identifier :reader
> socket-creationg-error-identifier)
>     (situation :initarg :situation :reader
> socket-creation-error-situation)))
>
> I imagine this is case-insensitive and doesn't matter, but then again
> I'm never quite sure how when case-sensitivity can creep into things
> down in the depths of ccl... Assuming that isn't the problem, I'd chalk
> this up to a paserve problem if I didn't seem to recall some stuff
> changing around socket-error in the latest release. Any
> ideas/suggestions?
>

I seem to recall (perhaps incorrectly) that paserve tries or tried to
redefine the CCL:SOCKET-ERROR class to make it more like ACL's.  If I
remember correctly, this is an example of why that can't work. It may
have been the case that some conditions that were signaled as SIMPLE-ERRORs
a few OpenMCL versions ago are now CCL:SOCKET-ERRORs.




> Thanks,
>
> Cyrus



More information about the Openmcl-devel mailing list