home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!van-bc!questor!hfs!bhead
- From: bhead@hfs.questor.wimsey.bc.ca (Bradley Head)
- Newsgroups: comp.sys.next.programmer
- Subject: strlen() memory exception in Speaker/Listener - NeXT's doing
- Message-ID: <1992Jul24.150603.8020@hfs.questor.wimsey.bc.ca>
- Date: Fri, 24 Jul 1992 15:06:03 GMT
- Sender: bhead@hfs.questor.wimsey.bc.ca (Bradley Head)
- Reply-To: bhead@hfs.questor.wimsey.bc.ca (Bradley Head)
- Organization: Headfirst Software, Vancouver, BC, Canada
- Lines: 18
-
-
-
- When I specify a parameter type of (char *) in a msgwrap file, then
- proceed to use the speaker message passing a NULL pointer, the speaker
- sending mechanism produces an exception inside strlen(). Why didn't NeXT
- check for NULL prior to calling strlen(). Now I have to, and it makes the
- interface less than seamless. In University, when writing published
- functions, I was taught to check input parameters (especially pointers)
- prior to using them. strlen() doesn't check for a NULL string pointer
- before dereferencing it (which is why I wrote my own strlen()), neither
- does NeXT before using it inside Speaker/Listener. Now I have to throw in
- the check prior to the call. This wouldn't bother me so much if NeXT would
- DOCUMENT such a side affect.
-
- ...Brad
- --
- Bradley Head, bhead@hfs.questor.wimsey.bc.ca (NeXTmail :-)
- InStep Mobile Communications Inc., Vancouver, BC CANADA
-