Just yesterday I witnessed my friend type a url into the google search bar (on the results page). She found what she was looking for, so I guess that's alright.
[1] https://arstechnica.com/gadgets/2021/06/google-chrome-ends-i...
It's bad, very very bad
Be upfront and clear with your users. If you are sending data to a 3rd party in a way that is not 100% obvious, then it's bad UI leaning towards dark patterns. If you are a UI dev and it never occurred to you a user would do something like this, then that's typical. If after you learn that users are doing something you as the dev finds not normal, then maybe the UI needs to be tweaked. If you design the UI to confuse/trick people, then you are just an asshole.
Edit: Also I genuinely don't think collecting end-user data is the primary goal of the Chrome team with this change, although I agree it's a likely side-effect. I think they are trying to create a better end-user experience, even if you and I don't feel that it is better for us (nerds) in particular :)
If I was actually wanting to search for information about the URL provided, then the result is expected. However, if I just want to go to the damn url, it is not useful at all. Now, in this unified input that you seem complacent with, how do I avoid the step of search results when I just want to view a specific website?
Also, I didn't say google wants to kill the URL bar. I said google wants to kill the URL. So you wouldn't type in "example.com", you would do something else to get to whatever content you wanted from what used to be called "example.com".
I don't know what the replacement for URLs will be, and it seems neither does Google (yet), but I'm at least open to the idea that there may be a better idea.
Some further reading:
https://arstechnica.com/gadgets/2018/09/google-wants-to-get-...
I don't. As I replied to that post directly, the URL bar is not like the choke lever on a car, but more like the steering wheel or navigation input. It's the UI element that is used to communicate user's goals to the machine - not some incidental control of a subsystem that gets automated away.
I think you're sadly misguided and, frankly, scaringly naïve.
Not allowing a user to go directly to where they want to go is just bad. It's like ordering a Lyft to take you somewhere, but the driver takes you to 4 different places that pays Lyft money to artificially increase traffic to their place before ultimately taking you to where you wanted to go. It's just a waste of everyone's time.
I don't think you can jump to that conclusion.
For example I can come up with an idea that removes URLs without going through google. Imagine a new protocol where the user types in something like "john's pizza cafe; order a pizza". My local pizza joint is registered in a DNS-like system and handles the query directly with some commodity search-like system, never going through google; or the pizza joint offloads the query to some 3rd party which crawls their site and handles the resolution.
(This is just some dumb idea I came up with off the top of my head; please don't nit-pick it to pieces. Wait for Google's proposal to come out and nit-pick that.)
Every internet query going through google would be bad, sure. But google isn't proposing that... yet :)
that's not how this works. if you don't wont opinions, don't post an idea to the internet.
you've just replaced a working DNS system with a DNS system based on best guess or worse requiring a small biz to register with multiple DNS like systems. What happens when competitor beats you to those multiple registration points?
>But google isn't proposing that... yet :)
Why in the world would you think that google would do anything that did not use their system? That's just not based in reality one bit.
Yes, I said it's a really dumb idea. Stop wasting your time nit-picking it :) It was an example of how you could create an open, keywords-based system without requiring Google as an intermediary.
> Why in the world would you think that google would do anything that did not use their system? That's just not based in reality one bit.
Google creates open standards all the time.
https://http2.github.io/faq/#whats-the-relationship-with-spd...
https://en.wikipedia.org/wiki/WebP
https://en.wikipedia.org/wiki/VP9
I don't think their new URL scheme would catch on without buy-in from non-Google stakeholders, so there's plenty of incentive for them to come up with something in cooperation with those stakeholders, who would surely reject a strong dependency on Google's services.
Yeah, you replaced Google with another intermediary or a bunch of other intermediaries. That just shows you're refusing to see that the problem is one of why there should be any intermediary at all.
and to prevent spamming the hell out of this system, each user's local KNS client will prefer those sites for a given keyword which are "bookmarked" by the user's peers (more on the closer peers, less on the more distant ones).
this involves some kind of social-netowrking into this system, but don't panic: it is not meant to be a FB, more like a GPG-like web of trust. if your direct "friends" or "peers" or whatever identity which participates in the system (a blog, a yellow pages provider, a local newspaper publisher) visits a site as a result to a keyword and he mark the site appropriative to the keyword (ie. it is not a spam but a legit content for the given keyword according to his opinion), then your KNS client/search engine ranks the site higher. if a friend's friend marks a site this way then your search results still get ranked upper but not that much. and so on: every level more distant the ranking is weighted less.
"bookmarked" is not the best term here, because the users don't save the site to bookmarks or into "Favorites", just marks it appropriative. they might not have incentive to do so, but it can be automated by assuming the user judges a site appropriative for a given keyword by interacting with it, or not going back too soon, etc.
though it is not perfect because one can just stay on a site out of an interest other than the entered keyword. and it raises privacy questions as well: users generally don't want EVERY of their visits to be signaled on the net, however they are often happy to share a fair amount of their keywords/visits with their friend circles. for the remaining set of activities on the web, they may use an other KNS identity (profile/persona) which nobody, or an other set of people follows.
your also now tying search into a social platform by using "friends" or "peers". i want to find something regardless of what anybody else is doing or has done. if i wanted what my friends were doing, i would already know about it from talking with my friends.
i'm so sick of everything trying to be relevant to what others are doing i could vomit.
it's less like "registering" keyword–site pairs, more like broadcasting keywords about your site. normally only the site owner can broadcast keywords about the site (it's cryptographically signed).
if multiple site owners have the "cats" keyword on their sites, then you simply get more hits on searching for "cats".
the "social" aspect is there to help in search result ranking. individually for each clients. it's really a bad naming to call it "social" and call the participans "friends", i've just named them so to make an analogy to the well-known social networks. but you are right, i don't want to mimic today's "push what is trendy" idea.
i wanted to design this system to be as analogous to real life as possible.
this "rank higher what others like/suggest/approve" logic is analogous IMO to this IRL: you are searching for a car repair shop, but there is 50 car repair shop advertisements in the newspapers. you ask some people which one to go to. (this may interpreted as "which one is the best" or "you like the most" - but only that should be taken into account in my ranking algorithm that the entity claiming it's a repair shop is really one, not fake, ie. not something which people generally flag as spam). in this way the ones which are likely really repair shops get more relevancy points, and the spammy ones get less. but you don't neccessarily ask ordinary people about it, you may ask the government what are the accredited ones. in the algorithm, it translates to follow/subscribe to the government ('s KNS handle) and sort the search results according to their ranking (the may give each accredited car shop 1 point, and 0 for the rest).
this idea is based on the observation that IRL without central directory of "things" you often evade spam advertisement by looking around what is its reputation at others. it does not prevent you to discover lower ranked search results, of course.
this ranking algorithm is there exacly to prevent any company registering all keywords and effectively gaining benefit from it. it may increase the load in the keyword propagation network traffic, but that's an other layer to worry about.
I mean, I think the DNS was a brilliant invention; it saved you having to memorise the IP address of every site you visited. But talking someone through entering a URL over the phone is a bit like fixing printer problems remotely; it feels like the early '80s.
But I'd prefer real URL bars, rather than a field that doesn't know whether you're entering a URL or search-terms.
I'm remembering the difficulty of explaining what is a forward slash and what is a backward slash. After all, they both lean forward - one leads with its feet, the other with its head. It doesn't help if your contact is using a different keyboard from you. You have to keep asking them to read it back to you.
I still have to do this occasionally; it feels like swimming through treakle.
Edit: Oh, and then there's people who don't know what "colon" means. And "hash" and "pound" have different meanings depending where you are. At least in this country, punctuation is less and less emphasised. I had to teach my kids to punctutate myself, and I suspect I'm the only person they use punctuation for.
Leading with one's feet is commonly known as leaning backward.
I really hate what Google is doing in this area with their immense power over the web. It may be more friendly for those who just don't care how computers work and just want their Facebook button, but it's very hostile for those who do know how they work.
Normally I don’t even need a url though, I’ll just google “chase bank” or wherever it is I’m trying to go.
I know Google is generally supposed to be above maliciously modifying search results, but people definitely aren't above maliciously poisoning Google search results. I definitely wouldn't want to get to my bank's website by any source other than my bank telling me where it is. (I do rely on my history and/or my bookmarks to remember that once the bank tells me, but hopefully those are proof against external malice, or else I've got bigger problems.)
[0] One could argue that Google will correct spelling and URL errors for you. But on the other hand, it's likely to inject ads above the real destination, which are very likely fraud, so I'd say it's a wash at best there.