More on cookies

[Dorothy Day]

On Wed, 4 Sep 1996, Michael Haseltine wrote:

> 
> Setting Netscape to verify each cookie isn't a good solution, either. It's a
> nuisance to have to click a dialog box for each cookie, up to one for every
> graphic, 14 on the last page I looked at.
> 

A suggestion that's been posted to browser newsgroups is to allow the
cookie file to be created (or create it yourself), then make it
read-only. Then turn off the query on whether to accept the cookie, and
let it be sent. Since the file is read-only, the data won't stick, and
you won't have to constantly click Okay or Cancel.

I haven't tested this, and have no assurance that this will work over
time; surely some sites at least will start testing whether the data got
written to the file, and warn you that your cookie is crumbly. But the
less sophisticated uses should cruise on, no wiser to the foil.

Of course, your enhanced use of the site, dependent on retrieving your
previous uses, will also be rendered impossible. Think of it as similar
to a customized profile for a TOC service: invoke your settings, and get
the information you want more quickly. The server side learns and
modifies the profile (cookie file) as you modify your choices. Suppress
those settings, and make the choices afresh.

Transaction logs of IP numbers (or account names, if available) visiting
a web page will continue to collect such data whether your browser
allows cookies or not. If we want to counter that kind of data
collection, we'll have to fight it on a much larger front.

---
Dorothy Day			
School of Library and Information Science
Indiana University
day at indiana.edu