5 ms·
I don't really know how I feel about having the browser mess with URLs without the user engaging it deliberately. It feels to me something that should perhaps b
by eythian 5y ago
I don't really know how I feel about having the browser mess with URLs without the user engaging it deliberately. It feels to me something that should perhaps be approached with caution. On the other hand, it does make sense. It's a tricky one.
- ChefboyOG 5y agoI think an automatic alert (which can be set to ignore by the user) which flags such links, and offers the option to turn on a config flag which enables URL manipulation like this, would be a good compromise.
- bonestamp2 5y agoYes, maybe some flag that comes up at the end of the address bar to indicate that it sanitized your URL. Then you can click the flag to see details about what it changed and have the option to navigate to the original URL.
- tinyhitman 5y agoMaybe prompt the user with a "use clean" or "use as-is" button (and "cancel"). and maybe a custom option; where you can toggle what is cleaned and whatnot. and provide a "remember for this domain".
- rakoo 5y agoFirefox already positioned itself when it gave the user a possibility to block tracking cookies and fingerprinting techniques. It's engaged even further now with Site Isolation. If utm_* query arguments are used solely for tracking, then it only makes sense that Firefox goes the next step
- sodality2 5y agoIt's possible it can be used for other things however, so it's a slippery slope. I block them anyway but maybe not default?
- IgorPartola 5y agoConversely, if you know that using utm_ parameters will break for a large number of your users, you just won’t use it, no?
- michaelmior 5y agoThis is a fair point. At this point, it's unlikely to break anything. But if it became the default, then any sites that do use them for something important would likely stop. Although there's nothing preventing sites from just renaming these parameters and modifying tracking code to keep tracking anyway.
- rakoo 5y agoIt's still used only for tracking only, so I'd say it's the right time to block them by default before they start being used for something else.
- michaelmior 5y agoI've seen some sites that use those parameters to activate specific features on a landing page for example. Pretty rare, but it does happen.
- freeAgent 5y agoMost of the technologies used to track people on the web have other uses as well, so this isn’t a unique case.
- darkwater 5y agoWell that's exactly the kind of job that an (opinionated) User Agent should do for you. <aybe configurable, maybe not. You can always change your agent (so, browser) if you don't like its opinion.
- arsome 5y agoI'm not so sure, by this logic we should have ad blocking by default as well, however that's a recipe for getting your browser banned by popular sites.
- kroltan 5y agoPopup blocking is basically standard and expected, don't see why ad blocking couldn't become so too.
- zo1 5y agoPopup blocking is definitely not standard and expected. There is a 90% chance of every website you visit to show a popup and not be "blocked" by the most privacy-conscious browsers. But. They're technically not "popups", they're just divs overlaying over the content that you were served but can't see. Or they're little slide-banners that nag you about signing up for a newsletter email or agreeing to tracking cookie non-sense. Oh and let's not forget about the popups asking you to allow "notifications" from this site, or to allow "location info" to be shared.
- matheusmoreira 5y ago> that's a recipe for getting your browser banned by popular sites Good luck with that. They have no choice but to believe whatever data the browser sends them, data that we control. If their precious content leaves their server at all they've already lost.
- Nextgrid 5y agoAd blocking by default is absolutely the way to go. Spoof the Chrome user agent if this actually becomes a problem (which would help with fingerprinting anyway). This is a bit like antivirus software authors worrying about being "banned" by the virus creators.
- matheusmoreira 5y agoIt's exactly the kind of thing user agents should do. If it's good for the user, they should do it by default for everyone.
- smithza 5y agoIt is at best neutral for the user. Sometimes these source trackers help companies know that affiliate links are more often drivers of traffic. Other times it helps with A/B testing because they discover the main logo was more often clicked than the "click me" button or whatever.
- leipert 5y agoBut how does this help me as a user? Affiliate links are often hidden, and depending on the system might even lead to higher prices, because the shop is offsetting the affiliate program cost. Whether the company does A/B testing, what does that have to do with me? That also can be implemented without external trackers and just be set in a session. So I would say it’s a net-positive.
- calvano915 5y agoCash back portals often require click through on affiliate links.
- ______- 5y ago> On the other hand, it does make sense. It's a tricky one. All attempts by Mozilla to bake-in addon-like behavior so we don't have to install 'yet another damn addon' is welcoming, but as with any of these features, they come with caveats already present in the addons. For example, Firefox's HTTPS-Only mode (that is basically the HTTPS-Everywhere addon) breaks some sites, and also their anti-tracking feature will break some sites too. But then again: if a site is serving HTTP only then they're doing it wrong (with the exception of captive portals). As for the anti-tracking feature: I rarely see sites asking me to disable my AD-Blocker, and when I do I never give-in, no matter how desperate I am to see hidden content.
- woko 5y agoExactly. Every time I have used an add-on like ClearURLs, I have had issues at some point due to some zealous clean-up of URLs which breaks a redirection. Typically, I don't want the browser to mess with my browsing if I am on the websites of my bank, a shop, etc.