3084RE: [webanalytics] Re: How do we find how many visitors have directly typed our domain on Browser
- Aug 1, 2005Retention of the referrer or loss of the referrer is down to the browser and
code that executes there, not the server. There's nothing that CGI, IIS or
JSP can directly cause that will alter the referrer. It's owned by the
does a document.load(URL) or similar, and this will break the referrer, but
that's code running on the client and isn't server side.
doesn't take the referrer with it, redirecting on the server with a proper
redirect message does take the original referrer with it.
I suspect many in this forum have experienced breakage of referrers and have
blamed redirects. I suspect the redirect was only indirectly (ahem) the
culprit. The real problem is that redirected URLs don't get logged in
the page is executed, so the tracking service never even sees the page.
> From my experience, Apache redirects do retain the referrer.
> I don't know about IIS. However, redirects created
> programmatically (like JS, CGI, JSP, etc) are much more
> likely to drop the referrer.
> --- In firstname.lastname@example.org, "john marshall" <jmarshall@c...>
> > Tom,
> > That's not correct. Server redirects do not affect the referrer.
> Examine a
> > page of Google results and look at the URLs for the paid ads. You
> will find
> > they go to a page within the Google domain so that Google can count
> > the click, and a redirect is passed back to take the browser to the
> target page.
> > At the target page the referrer with the keywords is available.
> > The RFC says the referrer is 'the last page the user viewed' and not
> > last URL the browser read from'. I'm paraphrasing of course.
> > There's an exception to this. A version of Apple's Safari browser
> broke the
> > referrer by not adhering to the spec, so the referrer was
> indeed some
> > useless junk from the guts of adwords in the above example.
> Thankfully that
> > browser has quietly wilted.
> > John Marshall
> > www.clicktracks.com
> > _____________________
- << Previous post in topic Next post in topic >>