Over the past couple days new tracking cookies associated with demdex.net* have started appearing, seemingly skirting around "do not track" browser settings.
They also appear associated with companies (eg, postmedia.demdex.net) and financial institutions (eg, td.demdex.net).
If you don't want to be tracked by this organization, you'll likely have to explicitly block such in your browser's privacy and security area. Or you can try Ghostery or some other tracking cookie blocker.

* demdex.net is a domain used by Demdex which is an advertising company that is part of a network of sites, cookies, and other technologies used to track you, what you do and what you click on, as you go from site to site, surfing the Web. Over time, sites like demdex.net can help make an online profile of you usually including the sites you visit, your searches, purchases, and other behavior. Your profile can then be exchanged and sold between various companies like demdex.net as well as being sold to other advertisers and marketers.

I know you're wed to Firefox, but have you ever looked into Tor Browser Bundle?

Edit: Have you got a static ISP with your dial-up?

I hadn't but just had a boo; not likely my cup o' tea.


It's a dynamic IP connection.

After I manually blocked all the demdex.net-associated cookies and accessed the affected sites, all seems well. But it's another reason to carry on my practice of poring over my cookies at least once daily.

Am I misunderstanding the nature of a dynamic connection? Doesn't it mean that you can't be tracked because you're a moving target?

I've never seen the demdex.net cookie, but I've got a bunch of other recurring ones blocked in my hosts file.

I use Cookie ($15 shareware), and I love it! I've got it set to clear all unwanted cookies every 60 seconds, and the only time I've ever run into trouble with that was when it cleared a shopping cart before I completed my purchase, so now I just turn it off when I'm shopping.

One would think so — I certainly did — which is why I posted the "warning". Demdex (at least) seems to be able to set a cookie despite non-tracking preferences; manually blocking such offending seems to fix the problem.

Perhaps ganbustein and/or tacit and/or others more fluent on the subject could chime in.

> Demdex (at least) seems to be able to set a cookie despite non-tracking preferences; manually blocking such offending seems to fix the problem.

If your dynamic connection prevents your being tracked, why the paranoia?

Because I'm not certain about it. Ergo the request for the experts to chime in.

If the theory is a good one, you'd think that all ISPs would offer dynamic IP numbers at a premium charge.

Depending, I might bite.

[sotto voce] Dynamic IPs also skirt around pay walls, depending on how many a given ISP uses and how heavy one's usage is. But don't tell anyone. Shh. [/sotto voce]

Now that I've had a chance to muse about it ...

My sense of the situation – rightly or wrongly – is that a dynamic IP connection is actually static for the duration of the session, but the next session would have another IP address assigned. At least that's the way my ISP works and rotates through a couple dozen IP addresses.

Once a cookie gets set (and isn't deleted after a session), it's available at the next session. So if it's a tracking cookie, it's irrelevant whether the ISP provides static (permanently assigned) or dynamic IP addresses – the browsing history can be tracked.

May one of the resident wizards confirm my musings or disabuse me of them, in whole or in part.

> So if it's a tracking cookie, it's irrelevant whether the ISP provides static (permanently assigned) or dynamic IP addresses – the browsing history can be tracked.

It may depend on the cookie's criterion for identifying its location... If it uses your IP address, each browsing session will look like it's a different computer, but if it uses a unique machine identifier, you're correct?

I don't know what information tracking cookie services use for their nefarious purposes, but many of the so-called paywalls (appear to) simply monitor usage by a given IP address (ergo, different IP address, different user).

As you've said, let's see if anybody kicks in with some authoritative info.

Gerber is correct that a dynamic IP address is "leased" for the duration of a logon session, or for a preset interval of time. IPSs charge more for a fixed IP address, often MUCH more because that IP address has to be dedicated and hardware resources must be permanently allocated to that IP address. During the course of a day a given dynamic IP address may be used by multiple users thus reducing the ISP's physical hardware requirements. There is also an assumption on the part of ISPs that a fixed IP address is likely to be used for web hosting and the probability is it will use a lot more bandwidth.

Artie, I think you may be confusing NAT (Network Address Translation) with a dynamic IP address. NAT is a technology used in routers and does provide some security on local area networks. Rather than my attempting to describe NAT in detail I will refer you to this Wikipedia article.

Gerber ?!?!
Methinks it's a tad early for too much wassail.
But have a happy ... nonetheless.

As an addendum to this discussion check out the Adobe Privacy Center / Analytics and on-site personalization services.
Adobe offers hosted services as part of Adobe® Digital Marketing Suite which sets tracking cookies containing 2o7.net and omtrdc.net (Omniture) domains.

The following, taken from Adobe's specs, may be of particular interest:

When a company uses Adobe's analytics and on-site personalization services, that company tells Adobe what type of information it would like us to collect. Examples of the type of information a company may ask Adobe to collect are:
• The URLs of the web pages you visit and the time spent on them
• The URL of the page that showed the link you clicked on that brought you to that company's website
• The searches you have performed, including searches that led you to that company's website
• Information about your browser and device, such as device type, operating system, connection speed, and display settings
• Your IP address, which Adobe may use to approximate your general location
• Information you may provide on that company's website, including information on registration forms
• Whether you clicked on an ad
• Items you've either purchased or placed within the shopping cart feature on that company's website
• Social network profile information, including photos, fan and like status, user IDs, age, and gender

Paranoia might be an understatement.

Gerber, I have no idea where Gerber came from, I still haven't had a drop of wassail. My apologies

Hmmm? now I know it came from the spelling checker. It changed Grelber to Gerber when I hit the Post button on this message.

In addition to a source of cookies, Omniture (Adobe Analytics) is the name of a web beacon which, in addition to its tracking activities, is somehow also required for the proper functioning of many web pages. While it can be blocked with utilities like Ghostery, such action can be self-defeating when it also blocks certain pages' essential content or functionality. Paranoia, anyone?

Btw, it looks like Gerber discontinued the wassail, or someone there had altogether too much of it, to the (hopefully temporary) detriment of other afficionados. Until supplies have been restored (and even after), I wish you a very Merry Christmas!

Meanwhile, Grelber took over the baby food business, or was that just an hallucination?

Now that just be plain weird.
I bet there's a copy writer and/or printer out there who's polishing up his/her résumé (because there ain't no Gerber job to go back to).

As an aside ... If I were going to do it, I'd take over the pet food (especially the cat food) business — far bigger with hefty profit margins. Next time you're in any grocery store, check out the shelf space devoted to cat and dog foods compared with baby foods; it's usually 5:1 or greater.

Thanks, joemike, but either I was my usual obscure self or you misunderstood my question.

I was asking about the behavior of a cookie placed during one (dynamic IP address) browsing session and not cleared before its "host" website was revisited in a succeeding session (i.e. one with a different address).

Will there be a correlation of addresses, with info gathered during the second visit being appended to that gathered during the first, or will the info gathered from each address be unique to it?

Thinking about it, though, it doesn't seem like it would matter, because with addresses being passed around from user to user, if "n" different users visit the same site, the info accumulated by its cookies will be a mish-mash of info gathered from all of them, and if there's only one user visiting a particular site, over time, that user's info will be splintered among "x" different users identified by "x" different IP addresses.

Either way, it seems like grelber's dynamic address makes him untrackable to a very large degree, apparently subject only to which address he's using during a visit to a site that only he visits?

If that's the case, I wonder why ISPs don't allow users to opt for dynamic addresses (at less cost to the ISPs), which, on the surface, appear to be more desirable for many (most?) users?

There are at least two comments possible to put your cookie & IP# musings in perspective. The first one is about cookie content other than IP #. This could include identifiers like hardware IDs, user name, location etc., to name just a few. The second comment pertains to data gathering channels other than cookies, like the various trackers every web surfer attracts like smells do flies. Between these two there is more than enough bandwidth to collect perfectly distinguishing user info to circumvent the limitations of dynamic IP numbers. And I bet it's not even that hard to distinguish between different users sharing the same hardware.

Of course, I'm not saying that these channels are used all the time by every web site, but since it's so easily possible you cannot exclude it offhand. Unfortunately, for the sake of the argument here, it's impossible to qualify this surveillance without knowing the exact content of all data gathered.

Because fixed IP addresses cost the ISPs more money. With fixed IP addresses each user has to have dedicated hardware allocations that cannot be used by any other users. With "leased" (dynamic) IP addresses 100 users may be served by as few as 25 or 30 IP addresses. Thus reducing the ISPs cost 70%.

My son and I have the same speed service (60 Gbps) from the same ISP but for business reasons he has a fixed IP address and I have a leased IP address. He pays $200 more a month for his internet connection. (Of course in the even of an outage the ISP provides my son with on the spot repairs on a 24x365 basis. I have an outage it can take 24 to 48 hours before a technician will appear at my location.)

Thanks.

Oh, well...no panaceas, but a confuse the issue factor, even if it is kinda porous, still seems like it may be useful as an additional layer of protection.

So my question still stands.

You explained that in post #32297, which is what prompted me to ask the reverse of the question you answered.

And the answer is still the same. It costs the ISPs more money to provide fixed IP addresses so what do you see as the ISPs incentive for charging less for the more costly fixed IP address?

But the question is...


I don't doubt that they could charge enough for the cheaper service to keep it at least as profitable as the more expensive service.

If, as suggested, dynamic IP addresses are ISPs' defaults, then the only "option" is for static – at a premium. The cost of the former is probably dictated by market conditions in every arena where service is offered.

We're gettin' closer!

Why isn't dynamic service offered, i.e. why can't users opt for it, in all markets?

I thought that the gist of the conversation is that it is by default (as the lesser-priced service which is the ISP equivalent of a "loss leader").

Are you claiming that you don't have such (by default)?

Your post prompted me to investigate, because I've never seen a dynamic IP address offer from Verizon.

As far as I know, I've had a static address since I signed up for Verizon DSL, and I don't recall being offered a dynamic option at the outset (although that was a looong time ago, so I wouldn't bet on it).

The office that presumably deals with such stuff is closed for the weekend, so no more 'til Monday or so.

Update: Well, I'll be damned! My IP address is dynamic.

Now, do I want to get into turning my modem on and off regularly for the minimal, if any, benefit to be derived?

I take it that your DSL is "always on".
My standard dialup via fax modem, as slow as it is, "resolves" that problem.

It's always on. (I was under the naive misapprehension that toggling AirPort would change my address.)

My modem is a nuisance to get to...on the floor, out of the way, and when I toggled it just now I had a devil of a time getting it back on, so (my insurance notwithstanding) I'll be "static" (for all intents and purposes) unless somebody's got a helluva reason to go dynamic.

Edit: I wonder how the pool of available "dynamic" addresses is affected when everybody's on DSL and, chances are, their modems never get turned off?

ISPs, including Verizon, typically set a time limit on an IP lease after which the modem will automatically and invisibly to you request and receive a new IP lease. When your network is up and running 24x385 it is not uncommon that in effect you renew the lease on the same external IP address, but that is NOT guaranteed. By-the-way your local router typically uses the same dynamic (leased) internal IP address scheme for your Local Area Network.

So to summarize there is no benefit to be gained by restarting the modem to lease a new IP address or renew the lease on the same IP address you already had. The ISP does a more than adequate job of taking care of itself and NAT does a decent job of shielding your computer from outside malefactors.

That's an unusually long year.

I can't even blame that one on the spell checker, so I will just say it sometimes feels that long.

Thanks for the explanation; I can now track my IP address (Edit: which changed when I toggled my modem yesterday), see what happens, and report back...eventually.

It seems, then, that with broadband routers that a (very?) significant percentage of users (Edit: probably) never turn off there may be almost no savings to an ISP as respects dedicated hardware allocations, because the pool of dynamic addresses must be almost as large as the pool of users.

The ISP's percentage in using dynamic addresses seems to be that users who require static addresses must request and pay for them.

I have been using MacScan for tracking cookies and spyware. See this web site:

http://macscan.securemac.com

I'm very happy with it.