Hacker News Reader

Comment by mibbitier | Hacker News Reader

I want advertising. It tells me about things I might like.

You're perhaps one of the very very few people who dislike being told about things, but thankfully you're in the minority.

The whole "do not track us" idea is ridiculous and a 'moot' issue. Cookies aren't really necessary, you can track people server side based on their browser make up, and they won't know they're being tracked. 99.9% of browsers are completely unique and identifiable back to the computer.

If you really don't want to be tracked (<0.01% of users), then use TOR, lynx, adblock, disallow cookies, etc etc etc

Look also at the recent EU cookie laws, and how ridiculous and needlessly cumbersome they have made websites that have addopted it. Endless clicking confirmation boxes / dropdowns to say it's ok for them to store a cookie. We do not need more of this madness.

View on HN

I want advertising. It tells me about things I might like.

Assuming that I wanted to be aware of relevant advertising, the signal to noise ratio is absolutely microscopic. Even sites that have massive data dossiers on me (Facebook and Google) routinely abuse my attention.

>I want advertising. It tells me about things I might like.

I hate advertising. Usually the ads are irrelevant and for things I do not and would not ever want, are often scammy, sometimes they carry malware, often get in the way of me retrieving the information I went to a site to see..

I bet if I counted every time I clicked on an advertisement in the last decade, that number would be less than 20.

That said, I agree completely the "tracking" worries are absurd and more borne out of FUD than any concrete privacy issue.

>Look also at the recent EU cookie laws, and how ridiculous and needlessly cumbersome they have made websites that have addopted it.

Due to shoddy implementation of a shoddy law - that does not reflect whatsoever on the concept. And cumbersome? Really? Could you point to an example site?

> And cumbersome? Really? Could you point to an example site?

It's definitely cumbersome. I've hit a bunch of sites recently that threw up an interstitial requiring me to "opt in" to cookies. This is obnoxious, and I go through the same thing on every device. It's not that one site doing it is especially cumbersome. It's the aggregate behavior.

I didn't realize that a new EU law was responsible for this until I read mibbitier's comment, though.

The worst implementations "drop down" a message at the top of the page to tell me they're using cookies. This invariably happens just as I'm clicking on a link. The link moves, and I click on some link I didn't want to click on.

There are no words strong enough to describe how shitty that is, and it's incredibly common, on big, widely used websites.

Recently became required by the EU for all websites to notify users of cookies, so don't expect the notifications to go away. Better implementations don't move content around the page though.

I expect us (UK) to withdraw from the EU in the next few years, so hopefully it'll become irrelevant.

>You're perhaps one of the very very few people who dislike being told about things [...]

The question is told by who, when, and at what cost?

If the answer is, by someone with a vested interest, all the time, and at the cost of someone accumulating massive amounts of very personal information about me, then no, I don't want to be told about things.

I really don't think many people see advertising as valuable information. They see it as payment for a service they want and that's what makes this issue so difficult. I don't like being tracked, but if I have to pay for every single service with my credit card, that's not going to increase my privacy either.

It's a real dilemma.

Cookies aren't really necessary, you can track people server side based on their browser make up, and they won't know they're being tracked.

In theory, maybe. In practice, no that's not true. The big web advertising companies all subscribe to the Internet Advertising Board's code of standards which prohibits anything like panopticon-style user tracking.

How are 99.9% of browsers unique?

Surely something like "Windows 7 running $LATEST_CHROME" must account for a very large chunk of traffic?

I assume this takes into account IP address also?

See: https://panopticlick.eff.org

Interesting , I just cleared cookies and tried that and apparently my browser is unique.

I guess it would be somewhat rare (I am running Ubuntu) but not that rare.

I'd be interested to know how this can be, since nothing in the info it gave back looked especially unusual.

That you're running Ubuntu is 99% of the entropy. Nobody uses Ubuntu (for certain values of nobody). The fonts you have installed. The versions of the plugins you have installed. &c.

I've long wanted to build a WebKit based browser that simply lies in response to all of those evil questions. Add it to my backlog of Important Projects, I guess.

But then a lot of websites wouldn't work properly because they are pushing IE specific CSS (or whatever) on you.

I haven't installed any weird versions of plugins, I assumed they would be tied to whatever chrome version I have.

And Ubuntu isn't rare enough for me to be the only person using it (especially on EFF).

Going to have to experiment with this a bit.

> nothing in the info it gave back looked especially unusual.

It doesn't have to be unusual, it just has to vary slightly from machine to machine. It's the specific combination of those slight variations that's unusual.

Panopticlick demonstrates a few things a website can gather. It's not just browser and OS version: https://panopticlick.eff.org/

I just tested at Panopticlick, and it told me my browser is completely unique. Just like it said the last 8 or so times I've been to Panopticlick.

So I question how meaningful that is. Yes, it's pulling enough info to uniquely identify a browser for now, but in a week I'll add a plugin, remove a plugin, change a setting, update Firefox, whatever and the tracking will be lost.

Edit: Also, browsers could be patched to randomize some of the information Panopticlick is using, like the exact way the HTTP Accept header is written, and the order in which plugins are reported.

Reply on news.ycombinator.com