percentage no referrers

[maxcohen]

Hi there,

Can anyone tell me what percentage of sites is 'supposed' to leave a refer but doesn't?

With 'supposed to' I mean the percentage of surfers that block refers in any kind of way, so not the refers that don't appear because they came from bookmarks, newsletters, etc.

[cgrantski]

A lot depends on the audience of the site. Mainstream sites don't have this problem very much at all. Sites catering to people who for some reason think this is important will have more of it. Which kind of site do you want to know about? (You still may not get an answer you can rely on ...)

If you are talking about browsers that leave the referrer field blank for every request, then I can say that the sites I have data for, which are mainstream sites, very rarely have any of it - it would stick out like a sore thumb when I manually sort and parse a piece of a log. I think I've seen it maybe twice in the last three months.

I've always wondered what people are thinking when they go to the trouble of erasing referrers. What's the point?

[gregbo]

I've always wondered what people are thinking when they go to the trouble of erasing referrers. What's the point?

Some people block the referrer because they don't want to disclose information about themselves. This can be for legitimate or non-legitimate reasons. Some browsers don't send a referrer due to bugs or by design.

[cgrantski]

We're digressing, and I'm not putting down the people who want to block that information, but I can't think of an example of a situation where there could be bad consequences of having a record of what site somebody was on when they clicked on a link that led to a second site, other than non-legitimate links and sites. (Sorry about the complicated sentence.)

I mean, it's usually impossible to trace site logs back to individuals (I know it can be done when somebody has made a purchase on the site and there are quite a few other circumstances where it can be done with the cooperation of the site owner and working with certain site databases, but it's truly possible in only a minority of situations and it's not undertaken casually. I know it can be done with some pretty gigantic legal pressure on ISPs and so forth. But generally, no.). I also can't come up with any reasons why it would be bad to reveal that a particular site has a link to another site, unless bandwidth, images, etc are being siphoned off by the other site.

So, help me out here ... can you give me an example of when it would make a difference other than in the head of the person? I'm referring to activities and sites in the legitimate side of the internet.

I wouldn't mind at all having my mind changed about it. But my current and possibly naive take on it is that it's purely an exercise of privacy rights on principle rather than a way to avoid real and detrimental effects on somebody or something.

[larryn]

cg,

I couldn't find the page (my google-fu must not be working today), but I remember reading a page at MS that explains their logic for blocking some referrers, their logic being that passwords or other identity items in a query string could be passed between secure and non-secure servers.

Not much of an excuse/explination, but it could be a way to cover for bad programming.

Larry