• cybersandwich@lemmy.worldEnglish
    604·
    1 年前

    What this has done for me has highlighted how many things are tracker me and how badly those things are designed because they don’t fail gracefully.

    I had a telehealth visit link today that broke using this feature. So that’s nice to know. My virtual doctors appointments are being tracked by a third party.

    • Buffalobuffalo@reddthat.comEnglish
      392·
      1 年前

      Edit, looks like Firefox is smarter than me, ignore this.

      I don’t know what the link was doing, but just because FF thought it was “tracking info” does not mean it was nefarious. It could be used for authentication or security. I have not tested it, but I presume this would break a “reset your password” email link.

      • Knusper@feddit.deEnglish
        8·
        1 年前

        I’m rather certain, the way it works is that it removes parameters that are named like well-known tracking parameters. For example, most webpages use Google Analytics, so you see UTM parameters everywhere.

        A “reset your password” link could theoretically use a parameter that’s named utm_content, then it would presumably get removed by this feature, but I see no sane reason why one would name their password-reset parameter like that.
        In general, such tracking parameters are usually named in a way that it will rarely clash with other parameters a webpage may want to use, so for example they may have a prefix like utm_.

        • Buffalobuffalo@reddthat.comEnglish
          1·
          1 年前

          Looking at some comments on the linked post, I think you are right, and it would probably be fine for things like a password reset. I could play around with it, but my laptop is in the other room.

          • Knusper@feddit.deEnglish
            1·
            1 年前

            Stripping all GET parameters would break many, many legitimate webpages. 🫠

    • Coasting0942@reddthat.comEnglish
      21·
      1 年前

      Umm, your telehealth link was basically a one time password to log you in/authenticate you.

      This feature is for browsing the web where you shouldn’t have to identify yourself to visit a blog about Ravens. If you’re visiting your bank, a service you already use, etc, then the unique url was more for them to confirm it’s you because only you have that unique url.

    • deleted@lemmy.worldEnglish
      1·
      1 年前

      This is a good step forward for privacy. However, how it’ll handle data embedded in the URL like MVC?

      Also, if it does work well, it’s a matter of time until developers find a way to get around it and probably enhance and increase data collected in the process.

    • laurelraven@lemmy.blahaj.zoneEnglish
      1·
      1 年前

      It’s just the GET parameters it’s stripping, those can be used for all sorts of things to pass information to a website to be used as variables for all manner of innocuous things… They just get (ab)used by trackers more than normal web traffic since most of the other uses comes from a site that can pass that as a POST instead, which embeds the parameters in the request header rather than making the URL a mile long, and wouldn’t be useful (and could actually be problematic) to be shared with others as part of copying it

    • kakesEnglish
      31·
      1 年前

      As much as I like this idea in theory, in practice I would actually be pretty annoyed if ctrl+c did anything other than copy the currently selected text. I would like a keyboard shortcut, though.

  • 1984@lemmy.todayEnglish
    111·
    1 年前

    Why wouldn’t this simply be default behavior, and then they could add a “Copy link with tracking” menu item?

    • Mamertine@lemmy.worldEnglish
      4·
      1 年前

      I do that manually. In about 1 out of 10, what looks like tracking stuff is actually needed for the link to work. So I’d expect that copy without site tracking option to not work 100% of the time.

      • derpgon@programming.devEnglish
        3·
        1 年前

        Probably just removes known tracking GET parameters like utm_*. Just from a parameter name and content it is impossble to infer the use.

    • Smokeydope@lemmy.worldEnglish
      4·
      1 年前

      It might become the default a little later on, they want to make sure it works ok first as an experimental feature before pushing it as a default would be my guess.

  • phx@lemmy.caEnglish
    10·
    1 年前

    Hell yeah! Normally I try to do this manually, so this is a useful feature for me

      • voxel@sopuli.xyzEnglish
        1·
        1 年前

        it breaks some websites for me tho, such as the jetbrains YouTrack issue tracker. (including self-hosted); it just keeps reloading
        works on all other websites pretty much fine, but it’s definitely NOT perfect

  • N-E-N@lemmy.caEnglish
    5·
    1 年前

    Is this on Android yet? If yes, how do I use it, don’t see an option

  • thejodie@programming.devEnglish
    5·
    1 年前

    I saw it earlier. When I tried it, it still kept the ?utm=blah&rel=blargh stuff on the URL from FB. 🤷‍♂️

  • silence7@slrpnk.netEnglish
    5·
    1 年前

    These things also all strip off the tokens which make gift links work

    Edit looks like I was wrong on this

  • Doug Holland@lemmy.worldEnglish
    52·
    1 年前

    Firefox, or Mozilla, continues to be the only browser (at least among the biggies) that’s for the users, not the trackers and marketers.