SHORT VERSION:
Complete case-insensitivity in handling URLs can cause problems.
Even though host names are not case sensitive, the path can be.
REBOL got severely baffled by a server response redirecting a GET
of
http://xxx.yyy.zzz.com/Search/
to
http://xxx.yyy.zzz.com/search/
(and it baffled me for a while as well!)
GORY DETAILS:
This one took some hair-pulling ... (OBTW, names have been changed
and long lines rewrapped for all The Usual Reasons.)
I got the following response
>> print read http://xxx.yyy.zzz.com/Search/
** User Error: Error.
Target url: http://xxx.yyy.zzz.com.com/search/ could not
be retrieved. Circular forwarding detected.
** Where: print read http://xxx.yyy.zzz.com.com/Search/
for a page I can retrieve with my browser (Netscape 4.5).
A bit of command-line HTTP hacking reveals the following
# telnet xxx.yyy.zzz.com 80
Trying...
Connected to xxx.yyy.zzz.com.
Escape character is '^]'.
GET /Search/ HTTP/1.0
HTTP/1.0 302 Found
Server: Netscape-Enterprise/2.01
Date: Wed, 09 Feb 2000 20:17:08 GMT
Location: http://xxx.yyy.zzz.com/search
Content-type: text/html
Content-length: 222
<HTML><HEAD><TITLE>Found</TITLE></HEAD><BODY><H1>Found</H1>
This document has moved to a new
<a href="http://xxx.yyy.zzz.com/search">location</a>.
Please update your documents and hotlists accordingly.</BODY>
</HTML>Connection closed by foreign host.
After discovering this, I found that (of course)
>> print read http://xxx.yyy.zzz.com/search/
works just fine.
-jn-
