>>>>> "bill" == bill <[EMAIL PROTECTED]> writes:
bill> I just encountered the problem with cgi-lib that I mentioned
bill> earlier... Here is what I received when trying to log into the
bill> RWI:
bill> <html> <head> <title>cgi-lib.pl: Unknown Content-type:
bill> application/x-www-form-urlencoded,
bill> application/x-www-form-urlencoded </title> ...
Ah. The infamous doubled content type bug. IE 4.something as I recall.
I forget what the resolution of this one is, but it is in the archives
for dev-list. I think we ended up with a patch which changed the regexp
slightly to ignore the extra cruft...
Yup, here is the patch: (ignoring $Id$ and $Revision$ changes)
:; rcsdiff -u -r1.1 -r1.2 lib/cgi-lib.pl
===================================================================
RCS file: lib/RCS/cgi-lib.pl,v
retrieving revision 1.1
retrieving revision 1.2
diff -u -r1.1 -r1.2
--- lib/cgi-lib.pl 2000/04/13 00:13:29 1.1
+++ lib/cgi-lib.pl 2000/05/26 03:34:49 1.2
@@ -79,7 +79,7 @@
if (!defined $meth || $meth eq '' || $meth eq 'GET' ||
$meth eq 'HEAD' ||
- $type eq 'application/x-www-form-urlencoded') {
+ $type =~ m#^application/x-www-form-urlencoded#) {
local ($key, $val, $i);
# Read in text
======================================================================
One would've thought they'd have fixed this in the RWI, if there were
using cgi-lib there, too.... It's only been seven months now since
the work-around for that bug was posted here. [hint, hint]
But no, the version of cgi-lib.pl in 2.23 also has not been patched
around that bug. ['nother hint]
-JimC
--
James H. Cloos, Jr. <http://jhcloos.com/public_key> 1024D/ED7DAEA6
<[EMAIL PROTECTED]> E9E9 F828 61A4 6EA9 0F2B 63E7 997A 9F17 ED7D AEA6
