André answered well with words. My only addition is an exampleassuming a well-behaved browser and no interference from cache. http://example.ORG/ contains:<img src="/internal.gif"/><img src="http://example.COM/external.gif"/><a href="/internal.html"/>Internal Page</a><a href="http://example.COM/external.html";><a href="http://example.COM/external.gif"/> Opening this page will create two log entries for example.ORG:- The referer for the page entry depends on how the page was opened.See previous post.- The referer for internal.gif will be "http://example.ORG/";. Theadministrator of example.ORG would consider this to be an internalreferer. Opening this page also creates a log entry for example.COM:- The referer for external.gif will be "http://example.ORG/";. Theadministrator of example.COM would see an external referer and beupset that another website was "borrowing" the graphic (and bandwidthand processing.) Clicking any of the links will create a log entry on the specified server:- Clicking the link for internal.html creates a log entry onexample.ORG with the referer of "http://example.ORG/";. Theadministrator of example.ORG would see an internal referer.- Clicking the link for external.html creates a log entry onexample.COM with the referer of "http://example.ORG/";. Theadministrator of example.COM would see an external referer and(usually) be happy that another website was linking to the page.- Clicking the link for external.gif opens the image in the browserand creates a log entry on example.COM with the referer of"http://example.ORG/";. The administrator of example.COM would see anexternal referer and probably assume the image was used on a pageuntil investigating -- opening http://example.ORG/ and discovering thelink. Each server sees only requests for files (page, images, etc.) on thatserver. The referer is an indicator of why the file was requested.The logs do not differentiate between embedded elements and anchorlinks -- opening the referer page is the only method to discover whythe file was accessed (if the referer page has not changed since therequest was made.) solprovider On 8/2/08, Paul Li <paulur@xxxxxxxxx> wrote:> solprovider,> I'm really appreciated your message and it helps me A LOT!!! Just want> to clarify the two cases when the referer is url :>> 1. "The referer is the URL of the page containing the link if someone> clicks a link from a website on the Internet">> a. the referer URL is the "website on the Internet" from which a user> clicks the link, and> b. the " website on the Internet" is aother website but not my website.>> 2. "The referer is the URL of the page containing an image or other> immediately downloaded file".> this url is of a web page on my my site.>> Is my understanding correct?> Thanks again!> Paul>> On Sat, Aug 2, 2008 at 12:46 PM, <solprovider@xxxxxxxxxx> wrote:> > On 8/2/08, Paul Li <paulur@xxxxxxxxx> wrote:> >> Some referer pages are "-", some are other pages in the same website,> >> and still some are pages of other website. If I want to find the> >> users' request history (visiting history), could I just ignore the> >> referer page but only check the request page?> >>> >> Btw, why the referer pages are different, like what I asked above,> >> some are "-", some are other pages in either the same or different> >> websites?> >>> >> Thanks,> >> Paul> >> > The referer is not the last page viewed when someone types an address> > into a browser's address bar. This would create a security issue> > (typed URLs should not surrender viewing history) and a functionality> > issue (the page in the referer should contain a link to the current> > request.) Analyzing the referer should tell you:> > - what images are contained in what pages (and if other websites are> > using your images.)> > - what pages link to each page.> >> > Note that cache makes Web server logs inaccurate. Returning to a> > previously viewed page may load the page from local cache without> > contacting the Web server. A gateway server can also cache pages and> > intercept requests -- companies and ISPs may cache static pages to> > reduce bandwidth.> >> > Assuming a page is being loaded from the Web server:> >> > The referer is "-" if someone types an address into a browser's address bar.> > The referer is "-" if someone uses a Favorite or Bookmark.> > The referer is "-" if someone clicks a link from within a local HTML> > file on their computer.> > The referer is the URL of the page containing the link if someone> > clicks a link from a website on the Internet,> > The referer is the URL of the page containing an image or other> > immediately downloaded file.> >> > The referer comes from the HTTP Headers of the request. Anybody using> > a tool allowing control of the headers (e.g. telnet, putty,> > HyperTerminal) can set the referer.> > solprovider
|