flee@dictionopolis.cs.psu.edu (Felix Lee) (09/01/90)
If you're running C News you may have received many duplicates
recently, all involving message-IDs that have more than one '@'
character in it, like:
<9008311926.AA17938.mullin@buista.SINet.SLB.COM@ASC.SLB.COM>
(Multiple '@' characters are illegal according to RFC822.)
The problem is that dbz lowercases after the first '@' sign, but
relaynews (and others) lowercases after the last '@' sign, so the
article is never seen in the history due to case mismatch.
The fix I'm using right now is, in rfc822ize() in libcnews/case.c:
*** libcnews/case.c.orig Wed Nov 15 12:43:18 1989
--- libcnews/case.c Fri Aug 31 17:19:25 1990
***************
*** 109,113 ****
prime();
! p = strrchr(s, '@');
if (p == NULL) /* no local/domain split */
p = ""; /* assume all local */
--- 109,113 ----
prime();
! p = strchr(s, '@');
if (p == NULL) /* no local/domain split */
p = ""; /* assume all local */henry@zoo.toronto.edu (Henry Spencer) (09/02/90)
In article <F=xd97?1@cs.psu.edu> flee@dictionopolis.cs.psu.edu (Felix Lee) writes: >The problem is that dbz lowercases after the first '@' sign, but >relaynews (and others) lowercases after the last '@' sign, so the >article is never seen in the history due to case mismatch. Sigh... This is an illegal message-ID anyway (as Felix noted), but it is definitely bad behavior. The latest code, to be released in a patch that should have been out some weeks ago and will be Real Soon Now :-(, leaves it all to dbz so it's all done consistently. -- TCP/IP: handling tomorrow's loads today| Henry Spencer at U of Toronto Zoology OSI: handling yesterday's loads someday| henry@zoo.toronto.edu utzoo!henry