As I understand it, the referer string is supposed to show where
a page is linked from.
The referer which Arachne reports is the last site Arachne
visited, regardless of whether the two sites are linked or
connected in anyway.
To illustrate, I just did 'arachne http://localhost' on the
command line and then looked at my httpd logs. Here's what I
see:
localhost - - [19/Jan/2001:17:05:11 -0500] "GET / HTTP/1.0" 200 2439
"http://www.angelfire.com/id/glenndoom/how2fix.jpg" "xChaos_Arachne/4.1.66;beta
(Linux i386; GGI; 800x600,HiColor; www.arachne.cz)"
What this log entry is telling me is that Glenn has a link on his
how2fix.jpg to the root home page on my machine, and that I clicked
on that how2fix.jpg to get to my machine. In this case, there should
be NO referer value reported.
Then I exit arachne, start it back up, and manually enter
"http://localhost/manual". Again, this should result in NO
referer, yet my logs show where Arachne last vistited:
localhost - - [19/Jan/2001:17:13:18 -0500] "GET /manual HTTP/1.0" 301 294
"http://localhost/" "xChaos_Arachne/4.1.66;beta (Linux
i386; GGI; 800x600,HiColor; www.arachne.cz)"
Compare, for instance, the log entry resulting from 'netscape
http://localhost'
localhost - - [19/Jan/2001:17:24:49 -0500] "GET / HTTP/1.0" 200 2439
"-" "Mozilla/4.75 [en] (X11; U; Linux 2.2.14-5.0 i586; Nav)"
and from 'amaya http://localhost'
localhost - - [19/Jan/2001:17:26:37 -0500] "GET / HTTP/1.1" 200 2439
"-" "amaya/V4.2.1 libwww/5.3.1"
and from 'lynx http://localhost'
localhost - - [19/Jan/2001:17:32:20 -0500] "GET / HTTP/1.0" 200 2439
"-" "Lynx/2.8.3dev.18 libwww-FM/2.14"
I know it's possible to disable referer, (and I have now done so)
but when it IS enabled, the referer should only contain the location
of the linking URL, if there was one.
- Steve
