Hacker News Reader

> they're just acknowledging the fact that 99% of users don't need to see 99% of the URLs they visit on a daily basis.

While we're talking about things we don't need, let's include this change.

The fact is, for most of the last two decades we've already had a UI where users who don't care to attend to the URL don't have to, and users who care to notice can. What does this add? Nothing. But it does take away some legibility for people who care, and discoverability for people who might learn to.

99% of people using web browsers really get no cues from the path? Cite please. URLs aren't high tech any more than the address to your house is, and my observation is that even non-developers who are simply experienced browsers pick up cues -- even from barely legible URLs mostly meant to be parsed by machines. You don't have to be a programmer to observe that typing a string in takes you to a page, or that the string changes when the browser loads a change, and put together the address correlation until you start to understand what a URL is without even really thinking about it.

Or at least, you wouldn't have to be a programmer to learn to make that connection based off of simple observation skills if we kept the current model. If we move to this new hide-the-URL UI, probably you would (self-fulfilling prophesy!).

And sure, the web has lots of URLs that don't provide a lot of easily parsed cues. In the interest of being a little less selective, though, let's look at a few others:

http://ask.metafilter.com/261148/Give-up-financial-stability...

https://www.google.com/search?q=percent+of+the+web+powered+b...

http://blog.ebay.com/ebay-inc-earnings-q12014-twitter-sessio...

https://twitter.com/paul_irish/

http://www.allenpike.com/2014/burying-the-url/

Do people need to see these things? Nope. Can they derive utility from being able to see them? Yes. And they do.

Well yes (even if HN cuts those off), but you should also consider that all of the example pages you listed have the title and intent of the page clearly listed right on the page themselves.

Legible URLs are important when you need to decide to follow the link; once you're on the page, it's less important because there's bigger and more legible cues about the content right on the pages themselves then.

> URLs aren't high tech any more than the address to your house is

When you stop by the local coffee shop, do you take note of what its address is?

The users have an idea what URLs are, they just don't care.

No, because I know that I drove down the same street, parked in the same parking lot, and walked through the same door. The URL should always be visible so that I can glance up and see if I am in a familiar place before putting in a username/password.

The point of this is that you see and compare the domain before putting in a username/password. That's probably more reliable than comparing a full URL - the difference between y0urbank.com and yourbank.com is much more obvious than the difference between y0urbank.com/?bunch/of/state=whatever and yourbank.com/?bunch/of/state=whatever

The component of the URL you check when avoiding phishing attack (the domain) is still displayed.

Good analogy - Now that lots of places don't display their address, whenever I'm trying to figure out where a small shop in a strip mall is (particularly in a busy retail corridor where there are layered strip malls) even gps only helps so much. I end up spending a lot of effort to find it.

Similarly, ever tried to find your bud's house the first time in a subdivision, esp. at night? Those places all look the same and you drive past 4 times before you figure out one of the house's numbers and deduce it. The pizza guy the other night thanked me for having my number clearly displayed and lit, for the same reason.

Of course once you get into a building, it still remains annoying - I mean it's still not really obvious how the internal room/suite numbering works. That portion of addressing is totally up to the architect, and a lot of times is only intuitive after you know the space you're in (if ever).

> When you stop by the local coffee shop, do you take note of what its address is?

Postal address? No, but I do know (for example) I'm on Main street, in that tiny alley one block west of 4th Ave, in the building just left of that weird giant sculpture of a bird.

I'm constantly aware of and can describe my relative location, even if I'm not always capable of expressing it using an absolute designation like postal address.

On the web, though, there is no such relativity. One website is not near or far from or above or below another. I cannot conceive of my location, much less express it, in any way but absolute address.

So: since I do not travel to reach places on the web, how do I know where I am? The address bar.

When you stop by the local coffee shop, do you take note of what its address is?

That's like clicking on a bookmark. But when you need to tell someone else (who is unfamiliar with the district) where that coffee shop is, or vice-versa, that information becomes really important. Hiding the path and showing only the domain is like telling someone "it's in California".

On the other hand, if the URL is displayed, then there will be many who take note of the fact that it changes whenever they click on a new link or go back/foward, and it makes them mentally associate "that piece of text" with "this page I'm looking at" - they don't even need to know the term URL to do that. It's a bit unfortunate that browsers don't have "Address:" next to it anymore, because that would've made this association so much easier (someone seems to have made the same observation almost ten years ago, although it was FF vs IE: http://cheeaun.com/blog/2004/09/address-label-for-address-ba... ). Having made that association, they can then tell you exactly where a page they're looking at is, and vice-versa.

Yeah, there are ugly URL. But those are the problem of the sites that use them. I don't see why all sites have to pay for their sins.

    ftp://ftp.scene.org/pub/music/artists/4-mat/

and

    ftp://ftp.scene.org/pub/demos/groups/dead_hackers_society/

vs.

    scene.org

and

    scene.org

? No, thanks.

The only problem I see with this is the confusion between the search box and urls for most users.

I noticed my 10 year old brother trying to find a song the other day in a peculiar fashion. He simply typed in youtube and "genre name." He click on the 2nd link of the search results. Then clicking on the 3rd item of a side bar linking to a playlist. He was navigating the web through links provided by google like we use the directory system.

In this case, knowing the specific url would be mind numbing and utterly useless. However, the distinction between google and the web is just too blurred for so many people.

However, I think drgath's point is ultimately correct. For some people, there really is no internet besides Facebook. For some no internet without Google. Even crazier: I've met someone who does not know the internet without Siri.

If we hold onto things like the directory structure, then we couldn't have "advancements" in user interface design like iOS. Could we eventually get a web without urls like iOS is an operating system without a visible directory structure to end users? I think it would be a big win. The directory structure was replaced by single purposed apps. What will url's be replaced by?

you described the nefarious walled garden metaphor.

The holy grail of companies. Full control over 95% of the users by sacrificing 5...

you are just dead wrong thinking this is good design. obscenely profitable? Yes.

Good design is made by serving the extreme 5% while accommodating the 95... Take it from someone who actually majored in product design and usability. The rationale for user interfaces is that the 95% will at some tasks be the 5percentile, and if you don't serve them, over time you lose them. People think ios is a hallmark of usability only because the market has too much canon fodder so the 95percentile seems infinite. But eventually enough people will be fed up by being unable to send a file from one app to another the way they want. And will move to whatever crap interface that at least have a file system that allows them to compete the task.

Good design is made by serving the extreme 5% while accommodating the 95...

Oh I can't upvote you enough, this isn't just for design, this is how advancements in tech happen in general. I remember when CVS was the dominant version control systems and old fogies didn't need this subversion nonsense. I read an essay in defense of svn that argued "just keep using cvs and don't worry about it, there's always going to be a minority that needs key features most people don't, let them have svn."

We're now two generations of version control down the road, both svn and git ate their predecessor's lunch by catering to the needs of the handful that were unsatisfied. Once the new thing works, most people eventually comes along because the key features turned out to be pretty nice, even if not necessary. That's how software progresses in general. Walled gardens exist to prevent others from making the next product that could eat the current one's lunch. Why else would it be verboden to "duplicate" iOS functionality?

> I'm using Canary with this option enabled, and all you have to do is click the domain box, then you can freely view, edit, and copy the URL.

Does Command+L still work? Because one click is too many.

yes, ^K and ^L still work the same as before

Does ^L still jump to the URL bar, activate the URL button, and select it? Since that's how I copy URLs anyway, I'm fine with that. ^K should just jump to the URL bar since it's now a search bar and ^L should jump to the URL/search bar and select the URL, as before.

With that one small change, I'd still be happy.

Yeah, that still works. Interesting, looks like Ctrl-K prepends a ? before you start typing, which is how you tell Chrome (and Firefox) that what you're typing should go to the default search provider instead of being used as a URL.

How's the weather over Googleplex today?

while you're at it you might uninstall chrome and install ... don't even remember how the AOL browser was called... But it obviously was superior to any modern browsers with their lowly urls instead of AOL keywords.