Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 



Forgot your password?
typodupeerror
×
Firefox Mozilla Privacy

Firefox's Optional Tracking Protection Reduces Load Time For News Sites By 44% 207

An anonymous reader writes: Former Mozilla software engineer Monica Chew and Computer Science researcher Georgios Kontaxis recently released a paper (PDF) that examines Firefox's optional Tracking Protection feature. The duo found that with Tracking Protection enabled, the Alexa top 200 news sites saw a 67.5 percent reduction in the number of HTTP cookies set. Furthermore, performance benefits included a 44 percent median reduction in page load time and 39 percent reduction in data usage.
This discussion has been archived. No new comments can be posted.

Firefox's Optional Tracking Protection Reduces Load Time For News Sites By 44%

Comments Filter:
  • e-commerce (Score:5, Interesting)

    by SumDog ( 466607 ) on Monday May 25, 2015 @09:10AM (#49767703) Homepage Journal

    I worked in e-commerce once. Our client had fifteen tracking pixels in the final page of the checkout process! It added a good 10 ~ 20 seconds to that page. That was on top of all the Adobe Omniture garbage.

    I refused to pulled crazy triple shifts after I the Thanksgiving break and was let go. I was so glad. It was totally not worth it and unemployment felt awesome after all that rubbish.

    Also, fuck TOMS shoes!

    • That was on top of all the Adobe Omniture garbage.

      That makes an awfully expensive garbage!

    • by antdude ( 79039 )

      Are you still unemployed? :P

  • ...but I already have Ghostery for that, so it doesn't offset the news about ads in the default home page.
  • by Lumpy ( 12016 ) on Monday May 25, 2015 @09:19AM (#49767751) Homepage

    Add adblocking on top of that and you will double those numbers.

    The advertising industry is ruining the internet.

    • by Anonymous Coward

      Lack of forward thinking ruined the internet. The utopia is dead, capitalism will win out. There was no monitization strategy envisioned during the birth of the internet, and therefor it's not surprising the monitization strategy which companies are using offends the technologists sensibilities, since there was never to be one in the first place.

      Until you accept it and start acting like the greedy, commercial actor which you are, someone with better hair and a flashier car will always promise you what you w

    • The advertising industry is ruining the internet.

      I challenge you to describe a plausible alternative.

      • by Rob Y. ( 110975 ) on Monday May 25, 2015 @10:01AM (#49767917)

        Yes, the public has 'spoken', and the Internet shall be ad-supported and otherwise 'free'. That doesn't mean that internet advertising has to be as intrusive as possible - just because it can be. Certain kinds of internet advertising is probably effective enough without tracking your every move. Even Google was pretty good - and financially successful - when it simply tracked your search queries and used aggregated data to produce good search results. The results may be marginally 'better' (i.e. personalized) today, but that's got plusses and minues. In any case, I wonder how much more revenue personalized searches generate for Google than before. You still have to click on the ads for them to make their money...

        As far as other sites go, I imagine they're all sitting on huge troves of tracking data that they can't begin to figure out a use for - except maybe to sell it to somebody else which Google itself does not do, btw.

        • I don't think the public has spoken. They're just not aware what's going on and the potential harm. Would people accept it if they had to click an "OK" button telling them they were being tracked each time a site tracked them? I don't think so. Sites added the tracking with very little public discussion or disclosure.

      • No advertising.
        • I asked for plausible alternatives. Please explain how "no advertising" is such.
          • by Anonymous Coward

            For one thing there will be far less junk sites like ehow who thrive off of SEO spam for ad clicks. We'd also get rid of all the other shitty clickbait sites like Gawker and Buzzfeed. Finally it would bankrupt Facebook and put Fuckerberg in the poor house. Seems well worth it to me.

          • Apparently, it means, 'no salary for jones_supa' by the look of it. I agree, that isn't plausible. For you.

            The Internet historically has hated fucking spammers and advertisers. That hasn't changed. And the tools are in need of continual sharpening.

          • Its called offering the customer something in exchange for money. Companies like GE and Exxon Mobil have been using this strategy with moderate success.
        • by tepples ( 727027 )

          True, advertising is not the only way to fund the creation of articles and operation of a website. It's just that

          Subscribers can read the full text of all comments posted by tepples.
          [ Log in ] [ Buy a subscription with PayPal ] [ See other ways to gain access ]

      • by pereric ( 528017 ) on Monday May 25, 2015 @10:15AM (#49768005) Homepage

        Well, Wikipedia seems to work pretty fine without commercial ads (they do some fundraising sometime). And Open Streetmap seems to do fine, as are the plethora of services built upon it. Sometimes NGO:s and individuals do stuff and share it just because they want it done. Finding sponsorship or donations for the hosting fees are a minor problem then.

        • OpenStreetMap is slow, because running geodatabase queries efficiently without strong infrastructure is a hard problem.
      • by JamesTRexx ( 675890 ) on Monday May 25, 2015 @11:17AM (#49768375) Journal
        Static advertising.

        No more audiovisually distracting intrusive advertising burning bandwidth and CPU to peddle things you've already bought or looked into.
        Newspapers and magazines had people managing advertising themselves, picking relevant products and the way it's presented. Why can't websites manage it like they do and take responsibility for it?
  • RequestPolicy

    • by gstoddart ( 321705 ) on Monday May 25, 2015 @09:33AM (#49767791) Homepage

      + NoScript + Ghostery + AdBlock + Block 3rd Party Cookies

      For Chrome: ScriptSafe + Chostery + HTTP SwitchBoard + Disconnect

      The internet is full of shit which needs to be ruthlessly blocked.

      • + NoScript + Ghostery + AdBlock + Block 3rd Party Cookies

        For Chrome: ScriptSafe + Chostery + HTTP SwitchBoard + Disconnect

        The internet is full of shit which needs to be ruthlessly blocked.

        Safersurfing for persistent cookies not stored in the usual places.

      • Only after a page doesn't work in a text browser.

        I use links2 as my main browser and only use Iceweasel+NoScript if the page is interesting enough to read. Haven't used an adblocker in a while since NoScript seems to work well enough for me.
      • Don't forget EFF's Privacy Badger [eff.org].
      • For Chrome: ScriptSafe + Chostery + HTTP SwitchBoard + Disconnect

        Just FYI ScriptSafe is abandoned...

      • hmm... that will not get you very far. You don't seem to deal with LSO cookies at all, adblock is now as closed source as it gets (unlike adblock plus) and you are not addressing fingerprinting either (extensions like rubberglove, privoxy/proxomitron and disabling font enumeration in firefox).

        soon enough there will be a university degree in safe/private browsing.

      • Ghostery + Adblock are redundant with RequestPolicy + Noscript. And I prefer Self-Destructing Cookies, since it can get rid of all sorts of persistent storage objects and doesn't tend to break things. It deletes all cookies set while on a tab when you close the tab, unless you whitelist them.
        • by allo ( 1728082 )

          ghostery, but not adblock.

          adblock: adblock privacy gives you ghostery
          noscript: gives you a lot of ghostery

          adblock blocks ads, even from the same host. And noscript can allow scripts per domain, while adblock (privacy) blocks some depending on the path.

      • Re: (Score:3, Informative)

        by firewrought ( 36952 )

        And under Firefox, don't forget to tweak your about:config:

        dom.storage.enabled = false # DOM storage is cookies reborn
        plugins.enumerable_names = "" # Useful for fingerprinting
        network.http.sendRefererHeader = 0
        network.http.sendSecureXSiteReferrer = false
        geo.enabled=false
        general.useragent.override = "???" # May not be worth it.

        If you don't need them, WebGL and WebRTC are just big security holes:

        webgl.disabled=true
        media.peerconnection.enabled=false

        Not privacy-related, but...

        network.prefetch-next = false # Don't load pages without asking (esp. at work)
        network.http.pipelining = true # Improve load performance.

        • by allo ( 1728082 )

          for useragent.override use a popular one, not a random one. you may be the only person with "???" as UA, but not the only one with " Mozilla/5.0 (Windows NT 6.3; rv:36.0) Gecko/20100101 Firefox/36.0"

      • by allo ( 1728082 )

        plus Policeman

  • Faster! (Score:4, Funny)

    by Anonymous Coward on Monday May 25, 2015 @09:24AM (#49767765)

    Wow slashdot now loads faster!

  • Bullshit ... (Score:3, Interesting)

    by gstoddart ( 321705 ) on Monday May 25, 2015 @09:28AM (#49767771) Homepage

    I'm going to call complete bullshit on this.

    Because disabling 3rd party cookies and setting cookies to "ask before setting" will probably also have the same effect, because once you hit a site you've said "no" to, you never get asked again.

    This is saying "our awesome tracking protection is faster than promiscuously accepting all cookies and running scripts".

    The slowness comes from letting 3rd party tracking sites set cookies and run scripts ... which modern browsers seem to treat as the default, or letting any crap set cookies or run scripts.

    Their tracking protection isn't magic, it's just blocking crap. Some of which can be blocked by default anyway.

    And it's a setting which Mozilla backed down on enabling by default because advertisers whined at them.

    • Re:Bullshit ... (Score:4, Insightful)

      by Anonymous Coward on Monday May 25, 2015 @09:37AM (#49767819)

      It's not bullshit, it's an ex-Mozilla employee discussing just how bad the situation is. Turns out Mozilla don't have the clout to fix the situation without resorting to compromises we wish they wouldn't have to, but it's not bullshit. Just try it out yourself.

      • It's not bullshit, it's an ex-Mozilla employee discussing just how bad the situation is.

        Fair enough, but TFS makes it sound like this is a speedup due to how super awesome the tracking protection is, as opposed to the default crap of letting everything from a zillion other domains run by default.

        Run everything from every cross domain crap and advertising the crux of the problem here, because ads and other tracking crap have fucked up the internet to the point that dozens of other domains get to know every p

        • by Anonymous Coward

          Perhaps. But I look at it this way. The people willing to not turn the settings on are paying the bills. I would tell everyone about it. If no script did not pretty much screw up ~10% of the time. Even this site touches no less that 5. You get on some sites and there are 50+ domains it touches. You have no idea which one is actually serving product or advert or tracking.

          I have seen similar numbers in my testing. Throw in a proxy server and some DNS caching and the numbers half again.

          I went from abou

          • Another 5 or so by not using my ISPs DNS server.

            And if you're using GoogleDNS or OpenDNS, you're back on the tracking bandwagon.

            And maybe this is the way it should be... have the DNS providers be the tracking clearinghouse, and serve only relevant ads in a way that doesn't get in the way of the actual site content.

            The fact that Apple sticks "Safari Reader" in the Safari browser says something about how bad things have got... not only do you end up loading a bunch of stuff you don't actually want that takes time/bandwidth, the end result is often bad enou

    • Re:Bullshit ... (Score:4, Insightful)

      by tburkhol ( 121842 ) on Monday May 25, 2015 @12:33PM (#49768829)

      The slowness comes from letting 3rd party tracking sites set cookies and run scripts ... which modern browsers seem to treat as the default, or letting any crap set cookies or run scripts

      When Newegg includes a 1px image from criteo.com, criteo is no longer a 3rd party. When newegg directs "promotions.newegg.com" to edgesuite.net, then edgesuite is no longer a 3rd party (and in a way that is much more difficult for even clever ad blocking software to detect).

      The point they're trying to raise here is that including all of those web-bugs and their associated cookies does impact the visitor experience, and FF has a system to reduce it. You can take this from the user perspective: here's an easy way to speed up the web, without having to figure out which of the adblocking plug ins are really legit. You can look at it from the host perspective: if web bugs make your whole web site feel much slower, then maybe the analytics aren't worth it. There are a lot of people who just don't think about why their internet is slow. Every time someone stands up and says it takes longer to load all the ads on most pages than the actual content, a few more people will understand the cost of "free' web pages.

    • Re:Bullshit ... (Score:5, Insightful)

      by tlhIngan ( 30335 ) <slashdot.worf@net> on Monday May 25, 2015 @12:40PM (#49768881)

      The slowness comes from letting 3rd party tracking sites set cookies and run scripts ... which modern browsers seem to treat as the default, or letting any crap set cookies or run scripts.

      Their tracking protection isn't magic, it's just blocking crap. Some of which can be blocked by default anyway.

      Well, the reason it's faster is you avoid making extraneous HTTP connections which can be slow by slow servers.

      A lot of ad and tracking servers stall out the browser, and because everyone uses them, they're overloaded. The browser might have everything it needs to render the page, but all the tracking stuff stalls out the renderer so you get only the headers. You can easily increase the speed if you tell the renderer to ignore those tracking objects and the network stack to not retrieve that content.

      Slow ad servers are the bane of the internet - why ad companies don't purchase more bandwidth and capacity is beyond me.

    • by Anonymous Coward
      It's not bullshit at all. I once did a Cookie RFC compliance survey on the top 10,000 Alexa sites and wrote a Perl script to log their complete HTTP responses to a single root page request (i.e.: GET / HTTP/1.1). I was disappointed to see that the majority of them were still using Set-Cookie headers (badly) formatted as per the original Netscape Cookie Proposal circa 1994, not even compliant with any RFC known to man. I was also disappointed to see how many Set-Cookie headers were sent out by sites just in
  • A little late (Score:5, Interesting)

    by Runaway1956 ( 1322357 ) on Monday May 25, 2015 @09:40AM (#49767829) Homepage Journal

    I've already switched to Pale Moon, in part because Pale Moon loads sites much faster. I also benefit from reduced CPU usage, from about 60% to about 15%. Memory usage has also dropped, although less dramatically than CPU usage.

    HELLO FIREFOX!! You started life being the leanest, meanest, most efficient browser in the world! It's time to get back to your roots!

  • by guises ( 2423402 ) on Monday May 25, 2015 @09:45AM (#49767863)
    I had to look this up [mozilla.org]. For anyone else wondering: this is one of those hidden FIrefox features which is only available to people who know about it ahead of time, through the about:config interface. If you're one of those people who isn't in the club, the boolean you search for is "privacy.trackingprotection.enabled".

    [Insert rant about FIrefox's god-awful UI and severely lacking menu system.]
    • What do you mean lacking menu system.... ohh, right, you must be on Windows (or *nix, no idea how they do it these days).

      A lot of UI/Designers forget that you can't just make a pretty OS X like app in Windows because there is no persistent menu bar across the top to hide all your features. Being forced to dig through that side hamburger menu thing is a PITA.

      • by guises ( 2423402 )

        there is no persistent menu bar across the top to hide all your features

        There used to be. Is that gone? I wouldn't know - I stay as far away from MacOS as I do from GNOME, and for the same reason: if you want to do anything which even slightly deviates from what the UI designers planned for you to do (like enabling tracking protection) then you either know the secret handshake or you're out of luck.

        Actually, the thing which has irritated me the most about Firefox lately is the lack of configurability of shortcuts. Ctrl-w closes a tab while Ctrl-q quits the program... right

    • by allo ( 1728082 )

      The irony: Look at the two urls belonging to the feature "SAFEBROWING_ID" ... oh yeah, there it goes again with the unique id (which is used for phishing protection as well).

    • by Eythian ( 552130 )

      They covered all that in the article. Not just how to turn it on, but why it's not on by default/exposed in the UI yet. Seriously, you can't fail to RTFA and then be all "I had to look this up" and "you need to be in a special club to use this" when it explains what's going on right there in front of you.

      • by guises ( 2423402 )
        I don't think so. The links in the summary were to the paper and to explanations of the results of the paper. As it so happens, in the third link in addition to the paper results there was also an explanation of the feature but there was nothing in the summary that said, "This is the link you click on to find out how to do this."

        You're right that I didn't click on the links, but since the links didn't give any indication that one of them was providing the information that I was after I'm not willing to ta
  • If you block "player.ooyala.com", the page loads much faster. Turns out I'm blocking some analytic tracking thing.

  • I read the paper and it is ok (nothing groundbreaking) on the technical side. But I was shocked to see broad conclusion of political or economic nature that were not supported by any argumentation in the paper.

    In particular, the first sentence of the conclusion: "The Internet’s principal revenue model leads to misaligned incentives between users, advertisers, and content providers, essentially creating a race to the bottom."
    I guess we'll just take your opinion for it.

Never test for an error condition you don't know how to handle. -- Steinbach

Working...