Google wants to get rid of URLs but doesn’t know what to use instead

Google wants to get rid of URLs but doesn’t know what to use instead(arstechnica.com)

18 points by badgers 7 years ago | 11 comments

ergothus 7 years ago |

Let's assume (mostly) good intentions from google. The article mentions that users don't understand URLs.

Pause to allow a moment to blame users for not trying. Then let's actually approach the problem constructively instead of blaming users.

As I see it, the URLs contain two parts that matter to the user: the domain and then everything after that. (Scheme they can either handle or has been made moot by https everywhere, port is rarely used for the average user, etc)

Domains actually ARE bad for the average user, because they identify a machine (or machine group) while the user cares about a company/individual identity. This is why it took 15 years to get people to stop assuming all websites started with "www". The domain works fine...for techies. It is poor for what users expect.

As for everything else...the path, query params, and hash are all interpreted by the end machine. Some sites try to make it human usable (my thanks), but even those are consistant from one to another. So to Joe or Jane User, these are meaningless.

It is easy to blame google and even easier to blame users. I love the url and hate efforts to hide it... but there really are tasks it does a poor job of that we have no particular good option for.

Rather than blame the url (or these other parties) we should figure out a good way to address (pun unintended but appreciated) these issues.

arpgy 7 years ago |

I don't know why we would assume mostly good intentions. URLs are a way to link directly to online data. Without such a direct pointer you would have to .. I don't know ... 'search'.

combatentropy 7 years ago |

Syntax highlighting is all I think they should do. If they want to color "http:" red to show that it's insecure, fine, but don't hide it entirely.

In 69, the version that just came out, they have begun to sometimes hide the subdomain. It threw me for a loop, because someone sent me a link to the mobile version of Wikipedia:

  https://en.m.wikipedia.org/

but the URL bar said just:

  https://en.wikipedia.org/

The page layout was very different, and I wanted to switch to normal. But at first it looked like I was at the normal URL and maybe Wikipedia had changed its look and feel overnight.

Thankfully you can disable this in chrome://flags, at "Omnibox UI Hide Steady-State URL Scheme and Trivial Subdomains." But this still doesn't bring back "http:"

Firefox too now hides "http:" but you can get it back in about:config, by setting browser.urlbar.trimURLs to false.

CompuHacker 7 years ago | |

Chrome and Firefox developers have been actively pushing the trimURLs concept for years and will actually fight you about it, claiming that users won't care or understand.

> https://bugs.chromium.org/p/chromium/issues/detail?id=41467

Misrepresentation of URLs by default is becoming increasingly common. The URL doesn't have to fit in the box on every device, and if I don't know what resource I requested and how I can't troubleshoot my problems.

pvorb 7 years ago | |

I agree that syntax highlighting is the only thing they should do and I think that hiding technical details will inevitably lead to its own security risks.

The same has happened with Microsoft hiding file extensions from users in Windows by default, which leads to malware hidden in party.jpg.com email attachments or similar things. It's one of the first things I turn off, when I'm on a new Windows machine.

pvorb 7 years ago |

I believe that getting rid of URLs will further take control away from users of the web towards search engines, which is likely Google's actual intension behind these efforts. If you have a full URL that you can write down on paper, there's still a way to get around the search engine. Once we lose the technical detail of URLs, that won't work anymore.

PaulHoule 7 years ago |

Is this AMP 2.0?

jamesgeck0 7 years ago | |

No. AMP addressed a problem that had been solved for years; we already knew how to make websites that load quickly. URLs are different; they have issues that nobody is quite sure how to solve. Any solution Google comes up with that isn't purely cosmetic won't go very far unless Mozilla, Microsoft, and Apple agree to implement it.

touristtam 7 years ago | | |

It's exactly that: an issue of standardization.

ocdtrekkie 7 years ago | |

Yeah, basically. They'll come up with a trust system of some sort that either, you can't afford that will be like EV certs, or they'll make a generic way to paste over the URL that anyone can use, and hence, will still be phishable.

Either way, it'll insert Google into another part of navigating the web and give them more control of the Internet.

pvorb 7 years ago |

URLs might have security issues, but they are one of the main reasons the internet became such a success.