Hello, I noticed that functions such as WRITE-SEQUENCE will signal a condition of type SIMPLE-ERROR if EPIPE is returned when writing. This is admitedly in non-standard territory, as only TYPE-ERROR is specified for WRITE-SEQUENCE.
I could wrap these calls closely inside a HANDLER-CASE/HANDLER-BIND and perhaps signal there another custom condition, but without printing the error, it's still awkward for code to determine the exact error cause (I don't see an ERRNO slot in the SIMPLE-ERROR condition object). I also wondered if such cases shouldn't actually signal a FILE-ERROR or perhaps even a C-ERROR, which could include an ERRNO slot? Currently these propagate to the top-level of Crow-httpd's ACCEPT-LOOP-THREAD which logs it and resumes, which is not a problem, except in the case where the configuration doesn't include :LOG-ERRORS in the debug options, in which case it'd be nice to not log this type of expected error while still logging any unexpected one. Even in the case of :LOG-ERRORS, it'd be nice to log it at a higher level including the client's IP address and session ID, which still requires to properly differenciate this condition to handle it before it reaches the top-level... I did not check how other implementations signal this yet. Thanks, -- Matt ------------------------------------------------------------------------------ Everyone hates slow websites. So do we. Make your web apps faster with AppDynamics Download AppDynamics Lite for free today: http://ad.doubleclick.net/clk;258768047;13503038;j? http://info.appdynamics.com/FreeJavaPerformanceDownload.html _______________________________________________ Ecls-list mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/ecls-list
