Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
Google Chrome also enables this tracking feature by default, but in the current Chrome 73 version it includes a 'Hyperlink auditing' flag that can be used to disable it from the chrome://flags URL. In the Chrome 74 Beta and Chrome 75 Canary builds, though, this flag has been removed and there is no way to disable hyperlink auditing." Firefox still allows this "feature" to be disabled (and disables it by default).
Posted Apr 8, 2019 17:34 UTC (Mon)
by hkario (subscriber, #94864)
[Link]
Posted Apr 8, 2019 19:28 UTC (Mon)
by rc (subscriber, #108304)
[Link] (15 responses)
Posted Apr 8, 2019 19:31 UTC (Mon)
by remicardona (guest, #99141)
[Link]
Posted Apr 8, 2019 20:25 UTC (Mon)
by MarcB (guest, #101804)
[Link] (12 responses)
Otherwise they will just continue to do tracking via redirects (and then, optionally, hide the redirects again via Javascript) and get *exactly* the same information.
As a demonstration of this, visit google.com and search for "lwn".
Then admire the nice, first link seemingly pointing to https://lwn.net.
Then use the "copy link to clipboard" function and paste it in the address bar.
Then realize, that the link actually is https://www.google.com/url?sa=t&rct=j&q=&esrc... (will likely vary).
Then ask yourself what "ping" would make worse (I can tell you what it would improve: Copying links would work again).
Posted Apr 8, 2019 20:42 UTC (Mon)
by mathstuf (subscriber, #69389)
[Link] (9 responses)
Posted Apr 8, 2019 20:53 UTC (Mon)
by MarcB (guest, #101804)
[Link] (5 responses)
Any such extension could also be written for the ping attribute and would be complete site-agnostic, simpler and robuster.
Posted Apr 9, 2019 8:06 UTC (Tue)
by alonz (subscriber, #815)
[Link] (4 responses)
Posted Apr 9, 2019 16:56 UTC (Tue)
by leromarinvit (subscriber, #56850)
[Link] (3 responses)
No offense meant against you personally, but the fact that people/companies feel the need to do such shady things is an indicator of a very sorry state of affairs.
Posted Apr 10, 2019 0:13 UTC (Wed)
by codewiz (subscriber, #63050)
[Link] (2 responses)
I'm a privacy advocate, but I don't see anything wrong with encrypting the url in response to user agents not honoring the ping attribute and stripping redirects.
Posted Apr 20, 2019 17:52 UTC (Sat)
by Tomasu (guest, #39889)
[Link] (1 responses)
Posted Apr 21, 2019 4:37 UTC (Sun)
by codewiz (subscriber, #63050)
[Link]
As other comments have already pointed out, we'd be better off with user agents honoring a privacy-respecting form of ping=, so web developers doesn't feel compelled to escalate it to JavaScript, encrypted urls and other opaque techniques that achieve the exact same result.
Posted Apr 10, 2019 7:23 UTC (Wed)
by LtWorf (subscriber, #124958)
[Link] (2 responses)
But IMO it is a grave security issue from browsers, because the preview of the link shows an URL, but then copying it or clicking it actually takes me to a completely different URL.
Back in the days they were always recommending to check the preview of the link to know if it was safe to click or not, and now browser vendors have made that unsafe to trust.
Posted Apr 10, 2019 8:11 UTC (Wed)
by roc (subscriber, #30627)
[Link] (1 responses)
Since JS was invented, scripts have been able to navigate the current page to some arbitrary URL with or without any user action.
> Back in the days they were always recommending to check the preview of the link to know if it was safe to click or not
That advice makes sense if you trust the page containing the link (e.g. results from a search engine you trust).
If you don't trust the page containing the link, there's no point in checking the link preview because the page can send you wherever it wants at any time, and that's been true for 20 years.
Posted Apr 10, 2019 8:13 UTC (Wed)
by roc (subscriber, #30627)
[Link]
This is a good argument in favour of <a ping> actually; it allows sites to stop obfuscating outbound links.
Posted Apr 11, 2019 14:56 UTC (Thu)
by benoar (guest, #52466)
[Link] (1 responses)
This is clever, as it allows (very) “dumb” UA to still work, and tracks only those who can follow redirect; but have the good side-effect of offering a tracking-free experience.
Of course, this is all with Javascript disabled (NoScript). Not quite a standard user setup, I admit.
Posted Apr 16, 2019 7:31 UTC (Tue)
by anton (subscriber, #25547)
[Link]
Posted Apr 15, 2019 4:17 UTC (Mon)
by marcH (subscriber, #57642)
[Link]
I guess the double negation didn't help:
> developers are removing the ability to disable ...
Now unlike the various and silly "disable_foo" in /sys/, I don't see how to simplify this one.
Posted Apr 8, 2019 19:51 UTC (Mon)
by roc (subscriber, #30627)
[Link] (23 responses)
In practice, having it disabled does not improve anyone's privacy. Web sites still track link clicks; they just change the link to point to a URL on their own server which redirects to the actual destination. This is slower and more user-unfriendly than "ping" (e.g. it breaks "copy link URL to clipboard"). We rightly decry "security theatre"; disabling ping is privacy theatre.
In fact, enabling "ping" by default (but some browsers still allowing users to disable it) would probably be a net increase in privacy. If all major browsers enable ping by default, then some sites would stop implementing the redirect workaround, and the small set of users who manually disable the feature would get increased privacy.
This is a classic case of the Web being a complex ecosystem and users being unable or unwilling to understand the second-order effects of browser decisions. Unfortunately there are enough vocal critics of "ping" pressuring Firefox to do the wrong thing that Mozilla probably can't afford to do the right thing here.
Posted Apr 8, 2019 20:06 UTC (Mon)
by nilsmeyer (guest, #122604)
[Link]
Posted Apr 8, 2019 20:10 UTC (Mon)
by brunowolff (guest, #71160)
[Link] (15 responses)
Mozilla doesn't do the right things because they get a lot of money from a company that wants tracking to be easier.
Arguably ping shouldn't have been implemented at all, because almost all sites that want that information don't want to let people opt out of it, so need to do it server side instead of client side anyway, leaving very few users of the feature.
Posted Apr 8, 2019 20:32 UTC (Mon)
by roc (subscriber, #30627)
[Link] (8 responses)
It's actually the opposite situation. It's impossible for users to distinguish tracking redirects from other kinds of redirects, if they notice the redirect at all. On the other hand, with "ping" a single line of user CSS (e.g. delivered by a browser extension) will show you which links are tracked --- before you click on them:
a[ping]::before { content: "(tracked) "; }
> Forcing redirects on the service, makes them pay more for tracking (to handle more requests or to proxy requests) which discourages tracking.
Using redirects or "ping", the site has to handle the same number of requests. There is almost no difference in server work required. There is no evidence disabling "ping" discourages tracking in practice.
> It also makes it more obvious to the user if they have redirects blocked or needing approval.
Essentially nobody blocks redirects. Redirects are used for many things, many of which are essential to the function of sites (e.g. OAuth), so blocking them breaks the Web. "ping" on the other hand is a much narrower feature which makes far more sense to block.
> Mozilla doesn't do the right things because they get a lot of money from a company that wants tracking to be easier.
As someone who worked there for a long time but doesn't now: this is just not true. You don't have any evidence for such a connection, and you are just making stuff up.
> Arguably ping shouldn't have been implemented at all, because almost all sites that want that information don't want to let people opt out of it, so need to do it server side instead of client side anyway, leaving very few users of the feature.
That may have been true up to now, but now that Safari and Chrome prevent disabling "ping" I expect sites to use UA-sniffing to use "ping" instead of redirects for those browsers ... leaving Firefox users with a worse user experience for no privacy benefit.
Posted Apr 8, 2019 21:23 UTC (Mon)
by brunowolff (guest, #71160)
[Link] (2 responses)
Lot's of sites work with redirects blocked. Whether it's too annoying to have on by default depends on the mix of sites one typically visits.
Google gives Mozilla money. Google is very interested in being able to track what people do on the web. Mozilla does some things with the design of Firefox that do not maximalize privacy. There may be other reasons they do that, but it is reasonable to conclude that it is likely Google's money influences those decisions.
I usually don't send a UA (it's officially an optional header), but pretending to be using a browser that doesn't allow disabling ping while actually disabling ping might be an easy way to avoid most link tracking. But that assumes the web site cares more about how the user preceives performance than being able to be able to track people, since the clients can't be trusted to honor ping, no matter what browser they claim to be.
Posted Apr 8, 2019 21:56 UTC (Mon)
by roc (subscriber, #30627)
[Link] (1 responses)
It is arrogant to demand that Mozilla set Firefox default preferences to suit the needs of the incredibly rare user who blocks redirects.
> Mozilla does some things with the design of Firefox that do not maximalize privacy. There may be other reasons they do that, but it is reasonable to conclude that it is likely Google's money influences those decisions.
That's not reasonable at all, given that there are many things Google would like (or would have liked) Mozilla to do that Mozilla doesn't do (e.g. adopt Chromium, abandon Tracking Protection, support PNaCl). You have no evidence that something nefarious is going on and apparently expect Mozilla to somehow prove the complete absence of such influence. If you want Mozilla to turn down Google money on principle and go broke "for the good of the Web", then you are misguided because that wouldn't be good for the Web at all.
Honestly I don't think your position is rational at all. If Firefox shipped ping by default you'd disable it, sites would use it, you could easily see which sites and links are trying to use it, and you'd deal with less Web breakage because you would have to do less blocking of redirects. What's not to like?
Posted Apr 22, 2019 9:41 UTC (Mon)
by oldtomas (guest, #72579)
[Link]
Hm. In my eyes, it is arrogant that Mozilla reduces their users to the lowest common denominator. Arrogance, it seems, is in the eye of the beholder.
(A proposal towards a compromise: enable ping by default, but highlight those links especially (perhaps by adding a little ⚠ next to it) with a short explanation in the hover text -- but I have the hunch that this won't fly with Mozillians, for... reasons).
> > Mozilla does some things with the design of Firefox that do not maximalize privacy [...]
I concur with that.
> That's not reasonable at all, given that there are many things Google would like (or would have liked) Mozilla to do that Mozilla doesn't do [...]
"Some things" != "all things". Let me apply Hanlon's razor here and assume you didn't notice your fallacy.
I don't think anybody's saying here that Mozilla is thoroughly evil. I consider it worth arguing with Mozillians, while I won't waste my time arguing with Chromians -- just hoping my life (or Chrome's) is short enough I won't have to.
Posted Apr 10, 2019 14:07 UTC (Wed)
by mathstuf (subscriber, #69389)
[Link] (4 responses)
Actually, I have uMatrix set up to block cross-domain redirects by default. It is only really a problem for shopping carts and OAuth stuff which bounce the browser around, but this is a known situation and I can reasonably predict when such stuff is actually going on to know when to relax the behavior.
Posted Apr 10, 2019 19:58 UTC (Wed)
by roc (subscriber, #30627)
[Link] (3 responses)
Posted Apr 10, 2019 20:33 UTC (Wed)
by mathstuf (subscriber, #69389)
[Link]
Posted Apr 15, 2019 10:39 UTC (Mon)
by ballombe (subscriber, #9523)
[Link] (1 responses)
Posted Apr 15, 2019 11:06 UTC (Mon)
by farnz (subscriber, #17727)
[Link]
Yes - specifically that this is an arms race between trackers (who understand that they won't ever get to track the 1% of people who will accept functionality degradation as a consequence of getting privacy) and privacy advocates (who want it to be easier to opt-out of unwanted tracking, and easier to see what tracking is happening).
Ensuring that a tracker can hit the 99% by using things like ping attributes that are both easy to identify and easy to filter makes privacy easier to fix; firstly because the 1% can turn off ping (easy) instead of having to deal with layer upon layer of JavaScript based obfuscation of the real link destination designed to help trackers, and secondly because it becomes easier to change browser UIs to identify tracking and inform "regular users" - e.g. by changing the hover behaviour for tracked links to include "(and notify tracker.example.com") for links with pings.
Posted Apr 8, 2019 20:47 UTC (Mon)
by MarcB (guest, #101804)
[Link] (5 responses)
The cost to the site operator is actually identical for redirects and ping. However, for the user, pings are cheaper, because the ping and the request to the link target can be done simultaneously while for redirects the client needs to wait for the first response to know the real target. Add to that the cost of the Javascript. Add to that the issue of broken link copying.
I doubt disabling redirects in general is actually feasible. Redirects are a core concept of HTTP and disabling them would break countless legitimate use cases (asking for permission would be quite annoying).
Posted Apr 8, 2019 21:07 UTC (Mon)
by brunowolff (guest, #71160)
[Link] (4 responses)
Posted Apr 9, 2019 15:41 UTC (Tue)
by nivedita76 (subscriber, #121790)
[Link] (3 responses)
Posted Apr 9, 2019 16:13 UTC (Tue)
by micka (subscriber, #38720)
[Link] (2 responses)
Posted Apr 11, 2019 9:56 UTC (Thu)
by mp (subscriber, #5615)
[Link] (1 responses)
Posted Apr 11, 2019 20:31 UTC (Thu)
by mathstuf (subscriber, #69389)
[Link]
Posted Apr 9, 2019 9:21 UTC (Tue)
by NAR (subscriber, #1313)
[Link] (1 responses)
I'm not sure how significant this would be. There's probably a huge inertia here, either in the "this is how it's usually done" form or in actual web frameworks. The frameworks need to be updated, the actual instances need to be upgraded, etc. before we see any of this.
Posted Apr 9, 2019 19:43 UTC (Tue)
by roc (subscriber, #30627)
[Link]
Posted Apr 13, 2019 10:46 UTC (Sat)
by oldtomas (guest, #72579)
[Link] (3 responses)
Oh veh. Mozilla. I wanna love you -- I gotta hate you. It seems that this "ad industry" way of thinking just oozes into the community member's brains without them really noticing.
Posted Apr 15, 2019 15:07 UTC (Mon)
by mathstuf (subscriber, #69389)
[Link] (2 responses)
> Oh veh. Mozilla. I wanna love you -- I gotta hate you.
So, uh, how is lynx these days? Less snarkily, what is the alternative you've chosen?
Remember, the real culprit here is Google for disabling the ability to not send the ping. Though I suppose an extension could do it as well if it really wanted to, it's just not an in-browser setting anymore.
Posted Apr 21, 2019 15:51 UTC (Sun)
by oldtomas (guest, #72579)
[Link] (1 responses)
Firefox. My main profile has javascript disabled (don't ask). For good measure, my /etc/hosts contains "things" (yes, I'll go berserk again when browsers start doing "DNS requests" via HTTP).
Sure, the measure is extreme, and perhaps Ghostery might fare better, perhaps not.
Many sites which I'm interested in still work (LWN is one of them). When something doesn't work... I consider whether it's worth to me to use some more permissive profile: most of the time it just isn't.
My beef with Mozilla is that it makes those choices more and more difficult -- e.g. making disappear this easy choice ("turn off Javascript) in the UI precludes more timid users from even experimenting with it... with the (umm...) justification that those users are Just Too Stupid to switch it on again (just an example).
> Remember, the real culprit here is Google [...]
Don't get me wrong: Google, for me, is in quite another category. I said I want to love Mozilla. I definitely don't want to love Google.
Posted Apr 22, 2019 15:52 UTC (Mon)
by foom (subscriber, #14868)
[Link]
You could look into "proxy autoconfig" (PAC) files. This mechanism allows you to write a simple JavaScript function to tell the browser how to access a given hostname (e.g. directly to the host, through an http proxy of your choice, etc). You can configure the browser to read it off a file on your machine, or via http. Every browser supports this, as it's commonly used in corporate situations.
Posted Apr 8, 2019 21:02 UTC (Mon)
by bokr (guest, #58369)
[Link] (3 responses)
lynx presumably doesn't do js or images unless specifically extended and configured to do so,
If it did get snuck in, how much damage could happen before it became a CVE and got corrected?
The question is what can <your idea of bad guys> do if they can insert involuntary action triggers
Ransom notes that you can't help acknowledging that your browser or gui email reader has seen? Nice ;-/
To get bad stuff to your html renderer, maybe more simply than owning the computers, buying an ad slot
BTW, what would LWN do if I posted this post in HTML mode and included a ping= ? Could I do my own analytics
Posted Apr 8, 2019 21:56 UTC (Mon)
by MarcB (guest, #101804)
[Link] (2 responses)
So what would change with ping in mails or forums is precisely...nothing.
Redirects in mails are a "best-practice" for quite some time, for various reasons. Professional ESP add them, because interaction with those links is a valuable indicator if the sending customer is clean (low interactions indicated that the mails are unwanted). Additionally it is a marketing analytics tool they can sell.
Phishing uses redirects as well, because the indirection can delay take-downs or slow down recognition by spam filters if multiple domains are used.
Posted Apr 11, 2019 10:55 UTC (Thu)
by bokr (guest, #58369)
[Link] (1 responses)
I was afraid someone could mistakenly enable ping= in other tags like <img ... >, not just <a ...>, so that it might have effect
Posted Apr 11, 2019 12:14 UTC (Thu)
by excors (subscriber, #95769)
[Link]
Posted Apr 9, 2019 1:38 UTC (Tue)
by flussence (guest, #85566)
[Link] (2 responses)
Unfortunately, the change is to make privacy worse for everyone in the same way Microsoft sending DNT by default did.
Posted Apr 9, 2019 2:17 UTC (Tue)
by roc (subscriber, #30627)
[Link] (1 responses)
Posted Apr 9, 2019 19:40 UTC (Tue)
by flussence (guest, #85566)
[Link]
Posted Apr 9, 2019 15:22 UTC (Tue)
by scientes (guest, #83068)
[Link]
If you spoof a chrome user-agent though, then don't send you the evil google redirect, which means you can copy urls no problem.
However, with a chrome user-agent youtube won't work anymore, because it will use chrome-specific extensions:
https://fossbytes.com/google-accused-of-sabotaging-micros...
good thing that everybody now is using the Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
botnet advertising company "engine" to render pages /s
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
As someone who has actually implemented a click-tracking solution — it is relatively common practice to encrypt the real target URL, so the link you get from the original site only includes an opaque blob, and only the click-tracking redirector can decode and decrypt it.
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
Thanks for the tip. Unfortunately, it does not work in my setup (Debian 8 Iceweasel with disabled JavaScript). I guess there is something additional in your setup to achieve that result.
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
Sadly what happens for me now is: I search for 'foo' using DDG, then in more than half of the cases I add !g to the search and usually immediately get what I was unsuccessfully looking for. This may be related to the general relevance algorithms used by both engines, not to a "filter bubble" per se, but click tracking is probably relevant either way.
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
some sites would stop implementing the redirect workaround,
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
from Mozilla presumably is careful about html rendering in preview formats etc.
and likewise mutt and emacs M-x eww, but who is keeping an eye out so it doesn't get snuck in somehow?
like ping= into email streams or html streams passing through their computers.
and preparing html ad material for automatic insertion -- I have seen lots of ads injected locally by isps even
as I browse a foreign web site, and guess the local isp is getting paid by the advertiser to insert their stuff.
(idk how all that is arranged, and who sanitizes what, and who determines the final hrml my browser sees).
on the reading of my post? ;-)
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
even where alt text is being rendered by a text-only presenter of html -- depending of course on where the presenter's html parser
comes from and what the implementers of that parser decided to do with ping= if seen in <any ...>.
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
Major Browsers to Prevent Disabling of Click Tracking Privacy Risk (BleepingComputer)
They still ping google searches in Firefox
