Quote:
|
Originally Posted by testpie
I'm asking this because i have been using a referer tracking script i coded myself on my website, The Pink List for quite a while now; but because it logs a unique referer as the full URI (i.e. http://www.site.com/whatever.html), it means that i have been having problems with linkdumps which use scripts to send out their traffic, as my script simply revisits the referer and looks for the title - and in the case of .../whatever/script.cgi, there usually isn't a title.
Basically then, would it be better to only log a unique referer by the base website address (i.e. http://www.site.com) and ignore any pages, as this also prevents one website from occuring multiple times within the referers section; or would this adversely affect the list by affecting websites with subdomains and etcetera?
|
Log both... Set up tables. 1 for the url base, 1 for the additional info.
Keep counts at the additional info point or at both points.
That way you can conrol it at the base level but pull counts from which actual pages are sending to you.