I am going to disagree since I believe that people should have a basic knowledge of the tools they are using. I am not saying that the bar should be particularly high, but being able to differentiate where something is and searching for something is a realistic expectation.

That being said, I think developers have made some things more difficult to understand than they should be. While people understand what an address is, calling it a URL adds an extra layer to learn (even if it is just terminology). URLs would also be easier to understand if they were more consistent from the end user's perspective, much like someone's mailing address, rather than an engineer's perspective.

I would also argue that the omnibar is designed from the engineer's perspective rather than the end user's perspective. The engineer said something to the effect of, "we can differentiate an address from a search query based upon the format of the input". It sounds good on paper. It sounds like it makes things easier for the user. Yet all that it really accomplishes is creating confusion for the end user since it fails to differentiate two concepts.

> believe that people should have a basic knowledge of the tools they are using

The whole point of intuitive design is to realize that you shouldn't make your users "think", insofar as making effortful cognition (i.e., Kahneman's "Type 2" thinking).

Good design should leverage built-in assumptions of how the world, per users' intuitive mental models, to create affordances that guide users to the desired interaction and behavior.

If a design promotes incoherent behavior, it's not the user that's wrong; getting users to change intuitive behavior requires training and cognitive effort.

Sadly, why vim and emacs are still difficult for newbies to pick up. Yes, the learning curve is (arguably) minimal; that's not the point. It's that there is a learning curve at all.

> The whole point of intuitive design is to realize that you shouldn't make your users "think", insofar as making effortful cognition (i.e., Kahneman's "Type 2" thinking).

Defined like this, this makes "intuitive design" a very destructive trend.

There's only so far you can go trying to "leverage built-in assumptions of how the world, per users' intuitive mental models". Computing world is not like everyday experience in many important ways, and trying to pretend it is only burdens the user with inconsistent mental models that don't allow them to navigate and grow their intuition.

The way I see it, it is the failure of UX design that it tries to hide the nature of computers - instead of making it apparent, so that the users can learn and internalize the important axioms of this different world. Intuition needs to be calibrated to the underlying reality for it to be useful.

> Yes, the learning curve is (arguably) minimal; that's not the point. It's that there is a learning curve at all.

There's few things in the world that have no learning curve. Almost everything is learned. Almost everything that's important in life is learned either through direct training, or picked up over time from exposure. This is where every UX designer inventing their own "simplifying" abstractions is hurting users: nobody is training them, but software is so inconsistent nowadays that they don't get to learn from exposure either.

You're thinking like a HN reader, rather than the intended audience for the chrome omnibar (i.e., non-specialists)

They don't want the abstraction punctured to see the inner workings...

Anything less than that creates a Norman Door conflict, whereby their inner mental models conflict with the intended interaction, where they want to pull the door, when the door only works by pushing (yet the door handle basically beckons to be pulled)...

Whether it's a good thing or not is besides the point. It's how people work, if you want to reach them at scale.

> They don't want the abstraction punctured to see the inner workings...

> Anything less than that creates a Norman Door conflict, whereby their inner mental models conflict with the intended interaction (...)

Here I disagree, because the underlying assumption is that those "intuitive" interfaces are exploiting an existing mental model of the users, like door handles in the physical world. This is not true, because computing is new, and there is no obvious preexisting mental model shared by non-specialists. Part of the job of UI is to enable the users to build correct mental models in the first place - and to the extent they introduce bad abstractions, they fail at this job, leaving users with bad mental models that don't correspond to reality, and don't let them make sensible predictions.

Note: I don't disagree with Norman. I disagree with how his work is used to justify current UX design trends.

> Here I disagree, because the underlying assumption is that those "intuitive" interfaces are exploiting an existing mental model of the users, like door handles in the physical world.

There was a subtle joke about this in Star Trek: The Next Generation: Wesley Crusher, on a visit to a starbase to take the Starfleet Academy entrance exam, stood in front of a door confused until Picard used the handle to open it for him. Wesley had grown up on starships where the doors opened automatically as you approach, and had never seen a door handle before.

According to Wesley's actor, this wasn't in the script and they came up with it right before doing the scene (and it made it into the episode).

> and there is no obvious preexisting mental model shared by non-specialists

Strong disagree. Looking at the URL bar in Firefox vs Chrome is a great example of how Chrome nailed an "obvious preexisting mental model shared by non-specialists".

We can look at things like the design of door handles, and clearly see how the shape of the human hand influences the design. If we had tentacles, door handles probably would look more different.

The same thing applies to human minds (as opposed to hypothetical alien minds, or AIs). Humans all have a hippocampus, a prefrontal cortex, and other universal biological factors in cognition which influences HOW thoughts are formed.

You didn't really make an argument for why "brains seem to have parts" leads to "Chrome's omnibar makes sense". The reason Chrome's design is "good"--which I put in quotes, as I absolutely agree that it is "destructive"--is because people are lazy: it is the same reason why co-mingled recycling is "good" and why a world in which trash and recycling were the same bin would be even better, and why a world in which trash, recycling, your toilet, and YOUR OUTGOING MAILBOX were magically all a single slot would somehow be the best thing ever and UX designers would shit themselves with glee over how clever they are for training users that there is a single hole that everything should go into and it is the job of an army of people on the other side of the hole to make that work most of the time... except for the occasional letter that gets recycled and the occasional empty box of cereal that gets mailed back to the cereal company :/.

Well, if that wasn't the worst slippery slope argument I've seen in a while.

That's like me saying "dogs and cats are both mammal pets" and you jump in and say "WELL JELLYFISH IS AN ANIMAL WITH DNA TOO SO OBVIOUSLY PEOPLE CAN'T DISTINGUISH BETWEEN JELLYFISH AND DOGS/CATS". Uhhh, no, that's a massive jump to an extreme conclusion. There's a BIG difference between someone typing "facebook" into a URL bar, and them putting their trash in their mailbox.

I'm not saying you should put literally everything including "change settings" or "access your bookmarks" in the omnibar, merely that combining the URL bar and the Search bar was clearly a common sense UI idea, and even Firefox bowed in after a while.

Your job as a developer isn't to force the user to learn more things. Your job as a developer is not to be arrogant and insist that the user must understand some technical model they don't care about. If they want to go to BankOfAmerica.com, and they typed in "BOA", your job is to make it easier for them to do their banking, even if they're massive idiots when it comes to tech. Human brains have a universal structure, which merely means you need to target the lowest common denominator when it comes to design.

As an aside, this is why Apple products historically are known for being very easy to use by the elderly. I certainly am grateful for the fact that my grandparents were able to learn to use Facetime and looked forward to using that software before they passed away. I'm not sure if Apple's ease-of-use is still true, for what it's worth- they've added a lot of complexity in recent years.

---

Addendum:

This is HN, full of very smart tech people who also happen to be terrible at empathy, so I'll give the usual spiel on empathy (which is usually disregarded by people with overly large egos): The person who typed "BOA" into the URL bar might be a tech idiot, but that tech idiot may just be a really good neurosurgeon, and that person may just prefer to spend their brain cells dealing with their own specialty. Humans are good at specialization, and I'd rather my brain surgeon not know or care or spend too much of their energy in understanding DNS, or the difference in how Chrome distinguishes URLs from search queries. For that matter, I'd rather not have to teach my grandparents the difference between a URL and a search query, either.

If you (the reader) is anything like the average tech guy, there's a good chance you're physically unhealthy in multiple ways, and in an alternate universe there's an internet forum of doctors lacking empathy who are shaking their heads and saying "why can't our patients simply spent a bit of their brain cells and time on eating correct amounts of the precise macronutrients, identifying each of the 13 major muscle groups and stretch and exercise them every day, actually doing the correct amount of cardio where they track if they hit 160bpm and sustain that amount (under age 45), sleep at the precise time each day for optional functioning of the central nervous system, etc". There's probably a mechanic forum that's complaining about tech guys who own cars and saying "why can't they just do a blackstone oil test to know what's the optimal amount of mileage for their car to do an oil change, why don't they routinely fill up their tires to the labeled air pressure to maximise fuel economy, etc". None of what these doctors or mechanics are suggesting something /unreasonable/ per se... but most of that stuff just aren't priorities for an average person to learn and deal with. In conclusion, you can be extremely smart in the tech industry (or any other field) and still be an absolute idiot in so many ways, and that's okay- but you need to keep in mind that the same thing applies to your users for their fields of non-specialty.

In summary: https://xkcd.com/2501/

Full disclosure, I started writing this after I took my sleeping pills for the night, so ignore any grammatical issues or bad structure

This isn't a slippery slope: this is an analogy. Your example of settings makes no sense as you have failed to draw a similar analogy and your example of bookmarks betrayal your own understanding of the system you like as bookmarks are in fact accessible via the omnibar :/. All of the things in my example were of a similar form, and frankly it would be convenient if anything you wanted to get rid of all went down the same hole: if you actually could solve that problem well, it arguably would be good UX.

The only reason the recycling examples seem absurd is because 1) it simply isn't that difficult to remember "trash goes in a black in and recycling goes in a blue bin and mail goes in... a different blue bin ;P" and 2) it is obvious how hard it is to separate the two results in the general case and the ramifications seem silly to tolerate given the low effort that the user has to take on to prevent it.

The exact same thing is true of the omnibar: there are a bunch of heuristics that often work well enough and people who are particularly lazy and don't want to learn the difference between the various bins don't have to because we figure it out for them... most of the time. I personally think that people shouldn't be so lazy, but it seems like the entire goal of "UX design" is to empower lazy people to "not have to think" (which is very different from "intuitive": after all, "the only intuitive design is the nipple; after that, everything is learned").

The reason then I think it is "destructive" goes further, though, than "people who are lazy do not deserve our sympathy": it is because, as someone else pointed out, there is a cost to leaking a user's intent to the search engine when they wanted to type in an address (in the same way that mailing someone's trash out randomly might be a privacy violation). Google doesn't give two shits about this: they want all of your information. That Google is training people to conflate the process of searching for something with then process of navigating to something is great for them and sucks for Firefox as Firefox kind of has to follow or be screwed and yet it seems a bit antithetical to their mission (though I don't think they even know what their mission is anymore).

(FWIW, the true original sin is the address bar itself: the address bar is a status display of where you are... making it editable is arguably sort of "intuitive"--using the typical kind of "just so" stories that allow people to claim random things are intuitive because the story was flowery enough--but frankly it just makes no sense and causes all kinds of problems. If you fix this, and return to the paradigm that some of the earliest browsers have, I think some of the other things the omnibar does are less awkward, and the "out" for true navigation wouldn't be affected as navigation would feel entirely unlike searching: it would become intuitive that these are separate concepts.)

(Also: the fact that I am unhealthy is not because I don't want to know how to be healthy or because I am not prioritizing understanding it... it is because most of the time I can't for the life of me figure it out even in consultation with doctors, and when I do know things--such as that I need to walk more every day--I don't prioritize doing it because I injured myself a couple years ago and I am now somewhat depressed over it. I point this out because it is very different from not knowing or not caring to know: a life un-analyzed and without curiosity at every level of scale and in every interaction would seem to me to be a life wasted. If you truly don't know anything about how a car works, I recommend spending some time to learn: it's fun! Tim isn't like the mechanic ever figured out what was wrong with my car until I finally pulled the error code of the ECU myself and rubbed it in his face. Learn to play at least one musical instrument of every class so that you can extrapolate the rest. Study linguistics as early and as often as you can in life so you can at least appreciate the beauty of other languages even if you don't choose to become fluent. Avoid hiring people to do things for you until you know how to judge if they did a good job, and decide how you wish things to be done before letting other people tell you what they like and then don't back down!)

Oh it was definitely a slippery slope argument, in the form of an analogy. You analogy went as far as saying trash and mail are the same, you are aware of that? Clearly, that is absurd and is using a case of an extreme. Mixing trash and mail, or trash and recycling, and sorting them later is obviously an absurd process. Mixing bottles and cardboard in one recycling bin is not absurd. But your analogy was clearly attempting to conflate all of the above, and I'd be going easy on you if I didn't bother to point that out.

Clearly the Bookmarks "star icon UI" is physically located in the omnibar, but when I said "settings and bookmarks", that clearly wasn't what I meant- I meant modifying the settings and bookmarks through typing a command in the omnibar the way you'd type a URL or search query. Bookmarks are not in the omnibar the way you can type "INSERT INTO bookmarks SELECT current_website FROM current_window" into the omnibar to add a bookmark. You can't change the settings "UPDATE settings SET setting_value=black WHERE setting_name=theme" by typing into the omnibar.

However, neither of my previous 2 paragraphs really refute any of your points, they're just clarifications. Pointing out that you made a slippery slope fallacy in an analogy doesn't refute your point, and whether Bookmarks are considered "in the omnibar" or not is orthogonal to the discussion. I agree wit the paragraph on "the true original sin", but that's a different note as well.

So, returning to the main issue: I think the cost issue is a valid point, but then everything has a cost. Even putting bottles and cans in the same recycling bin would incur a cost. Then the end question becomes "is it worth incurring this cost?". For a lot of people, they are willing to pay the taxpayer cost to have their recycling be sorted between the bottles and cans and cardboard etc. Do you have a separate recycling bin for glass bottles, aluminum cans, plastic bottles, cardboard, etc? Or just one recycling bin? Similarly, there is a cost to giving google a tiny bit of additional information- but not much. Plus, they already have your entire browsing history stored on their servers already. Try going to some webpages on Chrome, going to a different device, and check the browser history on the 2nd device- it'll be synced with the first device. So on Chrome, there's no additional privacy cost to combining the url and search bar. Even if you use firefox, there's not a lot of information being leaked with the omnibar that accidentally googled something that was supposed to be an intranet site. For the consumer that doesn't know the difference between a URL and a search, they probably aren't running some sites on their local network anyways.

Considering how low the cost is, I'd say a sizable chunk of people who oppose merging the address bar and the search bar do it out of a purely principled stance where they cannot accept any cost greater than 0, not even 0.00001. I'm not sure if this is a logical position to take, although it makes sense as a philosophical perspective, at least. The other chunk of people are just using it as an excuse to back up a philosophy of "people who are lazy do not deserve our sympathy", which makes even less sense logically (again, see: brain surgeons, grandparents) or philosophically.

On that note, I absolutely hate the quote "the only intuitive design is the nipple; after that, everything is learned". It's one of those cynical quips that just makes the person saying it seem smart, but actually leads you away from the truth. People massively overestimate the concept of a tabula rasa, and not all knowledge is learned. One example I can use against (a good percentage) of tech people is exercise at the gym. If you've ever done squats/bench/deadlift, you know the importance of having a good form. If your form is wrong, it will "feel wrong"- this is something that even absolute beginners at a gym notice, even if they don't have a trainer telling them exactly what is the right form. Another one, (a more interesting subject on the state of the human brain), is observing how people react to certain drugs like acid. Amusingly, novices to psychedelics are always convinced that whatever thought they've just come up with is some groundbreaking revelation, when it's usually something that's been thought of many times before.

> such as that I need to walk more every day--I don't prioritize doing it

Isn't this inherently more lazy and/or worse? This would be like someone who knew what a URL is, and then purposefully choosing to search up items in the Firefox URL bar anyways. At least not bothering to learn is excusable- again, I don't really need my brain surgeon to know how a URL works and understand making a DNS query for an infranet site works. But philosophically, this seems like throwing your trash in your mailbox and being okay with it because it would disappear anyways (excuse my slippery slope analogy, heh). And also I can possibly argue that you don't know that much about your health, anyways- for a human body to stay healthy, it needs more than just walking. Putting heavier loads to build muscles (maybe not specifically for hypertrophy, but certainly not just LISS cardio) is important, especially for your back. Also, sustained high intensity cardio is important for your health as well.

> If we had tentacles, door handles probably would look more different.

I would love to see a virtual reality app that plays with different organism layouts, and solving the usability issues.

There is a lot of great sci-fi that tries to consider how a world would be designed for non-humanoids, or mixed aliens, or the reverse archaeological problem of calculating how aliens look from what they have left behind. I am struggling to think of any sci-fi that makes you deeply think about usability.

"The Galaxy, and the Ground Within" by Becky Chambers (Book 4 in the Wayfarers Series) not only deals with a mixed alien usability environment, it also passes an alien Bechdel test of sorts.

The whole Wayfarers series is amazing in my opinion.

"A Closed and Common Orbit", the second in the series) is my favorite book ever.

Re vim: As a hobby developer (a perpetual newbie) the learning curve is an issue for me simply because I don’t yet understand the value in overcoming the curve. “Efficiency” is not one of my programming goals, and so reasons like “vim is great - it’s super fast” have no appeal. So, remembering 80 unintuitive shortcuts just so I can do what I do in Kate, but a little faster? Eh…

The value is that your fingers stay in the same place for typing, and for editing/commands. But also, once the commands aren't just in your actual memory, but are partially in your muscle memory, you don't even really have to think about what key sequence you're using; it just happens. Your fingers know.

The advantage of both of these factors combined is that you don't have to think about editing, or the user interface. You don't have to break the flow of thinking about your code. You type code, you edit code, you rearrange code, and you only ever have to think about the code. You don't have to spare brain cycles to think about editing, your fingers just do it.

It's like learning to touch type. There are two fingered typists who can type almost as fast as I can touch type. Or even six-fingered typists who might be able to go a bit faster. But even though they know where the keys are, they still need to keep an eye on the keyboard to type. They still need to think about the keyboard. But once you learn to touch-type, you don't need to think about typing any more. You just... type. Being able to keeping your eyes on the screen all the time just allows you to concentrate on your code, rather on the way you're entering it.

By turning some conscious actions into muscle memory, it leaves your conscious mind able to spend more time on the things you want to think about.

This is exactly what I'm aiming to do with my guitar playing. To internalize the instrument so much that I can only focus on the music and not have to think about where and how I should use my fingers. I practice hard to be able to achieve this.

The thing is, I see the value in my guitar playing. But when I code, much of my time is spent thinking about what I'm coding, and much less for the actual typing and editing, so I don't really see the value for vim. That's why I finished the tutorial, but then never used it in actual coding. Purely my own opinion, of course.

I found doing a few vim tutorials valuable as a reminder to look up some basic text manipulation in other editors. Like selecting a text rectangle with the keyboard, or cut and paste a single line.

Or that a text search is a perfectly reasonable way to move the cursor seven steps right to the next right paren.

> You don't have to spare brain cycles to think about editing, your fingers just do it.

> Being able to keeping your eyes on the screen all the time just allows you to concentrate on your code, rather on the way you're entering it.

I don't see anything here that is special or unique to vim or emacs.

> “Efficiency” is not one of my programming goals, and so reasons like “vim is great - it’s super fast” have no appeal.

Then think in terms of friction - friction that prevents you from doing what you want.

The reason for pursuing this kind of efficiency in an editor is because the task of writing and editing text files is only means to an end as a programmer: your goal is to encode the program you have in your head in a form the computer could run. To persist a castle in the sand you're building. Your editor is usually the limiting factor on how fast can you do that. And this matter because the faster you go, the less friction you feel, the more time you have for the important part - thinking. If you can do this fast enough, cross a certain efficiency threshold, your mind will collapse the task, erase the perception of typing.

Think of what happens when you learn to drive, truly learn. Your car becomes an extension of your thoughts, not something you consciously give inputs to. Now imagine driving by having to type in commands like "accelerate" or "turn left". Now imagine the same, but through submitting paper forms.

Thanks. I think I’m just not at a programming level where my brain is working faster than my fingers. But I get it. I can relate in the following case: I like to write fiction, and lately I have been writing children’s stories that I’ve told my daughter as I put her to bed. I’ve found that I can invent and tell a pretty decent story in the span of 10 minutes. When I later write up that story, it takes me 1.5-2 hours. Adding in punctuation seems to slow me down the most, in part because those non-story elements distract me and mess up my flow. I’m testing whether it’s faster to use a voice to text app, and then edit in the punctuation at the end.

Have you ever seen an experienced user demonstrate some use cases? I'm in the same boat as you but that convinced me because some of it was like magic. I don't think the value is in uniqueness of function, right, but the sort of swiftness of expression you can get.

More like having you workbench set up just perfectly, less like having irreplaceable tools. But the start up cost is what keeps me out - learning it, trying to integrate it into other tools that I don't fully understand anyway (VSCode in my case), etc etc.

Only brief glimpses. And I have no doubt it’s amazing in capable hands. My interest in learning vim is likely solely rooted in some vague jealousy of these skills. I keep having to remind myself to be happy with what I have and stop spending what little time I have gazing at greener pastures.

I take a middle ground - I agree that intuitive design is good especially for the typical non-techy user, but I also love designs that marry that with expressing the truth of the underlying system and not getting in the way of people who do actually want to leverage their understanding of the conceptual model of a program for power use. Most UIs fail to achieve both, and end up either too dumb to be useful or too complex to be widely usable.

> I believe that people should have a basic knowledge of the tools they are using.

I strongly disagree with this.

Does anyone remember what connecting to Wifi looked like? I'm talking 15+ years ago. Here's how it would go:

1. Click a network

2. Choose an option from WEP, WPA, WPA-PSK, WPA2 and WPA-PSK

3. Be prompted with a either a password, a key selection or a box that generates a bunch of keys. Enter what you think the password is somewhere;

4. If it doesn't work, return to (2) and try something else.

5. If it works, you win.

That was literally how it worked. And then Apple came along and just prompted you for password. Why? Because the OS can figure out the rest so why ask the user?

You can argue "the user should have basic knowledge of the tools they're using". After all, if they're connecting to a WEP network, they should be aware it's not secure.

But users don't care about any of that so there's no point in asking them. You're just forcing users to make decisions they don't care about, don't need to know it and aren't interested in. For what?

That's what great product design and interface design is.

Let's also refer to Joel Spolsky's classic "Choices" [1] where he quotes:

> Every time you provide an option, you’re asking the user to make a decision.

and then goes on to (correctly) say:

> That means they will have to think about something and decide about it. It’s not necessarily a bad thing, but, in general, you should always try to minimize the number of decisions that people have to make.

[1]: https://www.joelonsoftware.com/2000/04/12/choices/

>You can argue "the user should have basic knowledge of the tools they're using". After all, if they're connecting to a WEP network, they should be aware it's not secure.

>But users don't care about any of that so there's no point in asking them. You're just forcing users to make decisions they don't care about, don't need to know it and aren't interested in. For what?

Because if it's not secure, and the lack of security results in someone stealing the user's password and wiping their account, the user will indeed care about it.

The user does in fact care about the consequences of the thing he "doesn't care" about; he's just incapable of connecting those. It's like a user not caring about whether there's gas in his car because he knows nothing about cars. He'll care about it when the car stops running.

They should have basic knowledge, but they won't. The majority of people in the world can barely use computers well at all, and we need to work with this reality.

https://www.nngroup.com/articles/computer-skill-levels/

They can barely use computers because:

- computers are uselessly complex even for simple tasks

- UI changes are pushed by vendors every other year or so, despite providing zero interesting features for users

- there's no IT education, and when there is it's focused on writing a CV on Microsoft Word and sending email, nothing more

All three points could be addressed. And having a proper UX that empowers users with ease can help. I don't think all web users need to know what "URL" means or even what's http://, but they could certainly understand the difference between "location" and "search criteria" if UI was good. Even a simple help box could help: "If you know the address you're going to, type it here. Otherwise, click the search button". The kind of UX that's similar enough to phonebooks and maps so users aren't confused.

It's not remotely as simple as you make it out to be. People have spent decades trying to make this stuff easier.

I'm not saying it's simple. I'm merely pointing out that industry players have other interests at heart than those of their users. They are worried about their position and income, not about empowering people.

There is always a tradeoff between making something better for power users and making it bullet proof for (ab)users. Indeed, chrome strikes a pretty good balance.

Also

> bar should be particularly high

hehe

Why can't we teach this in schools? My high school curriculum covered the basics of the UI elements - window, menu bar, URL and such - in addition to OpenOffice and the basics about internet (this was 1998)

Why should the Internet be accessible only to people who have undergone a high school curriculum that teaches it? That already excludes a large portion of the population.

Why exactly should the interrnet accessible for all people at all? When you are too lazy or dumb to use a browser like it was in 2000, you will too think that facebook or insta are the internet. Thats the problem on today's internet. It's too accessible for too many ignorant people.

I don't see the point of your argument. I don't see any reason for excluding people who "think that facebook or insta are the internet" from it. If you desperately require more powerful tools to support the way you use the internet then there are powerful tools available; why use the lowest-common-denominator tool that is the browser?

I couldn't resist adapting your argument for food:

Why exactly should food be accessible for all people at all? When you are too lazy or dumb to grow your own food like it was before the coming of Jesus, you will think that McD's is the agriculture industry. That's the problem with today's agriculture. It's too accessible for too many ignorant people.

To a first approximation, school can't teach anything. Most of your high school class don't remember what DNS is.

DNS? Is that a high school level subject?

I was going to say something about how you needed to know what DNS is to know how URLs work and how all this is easily teachable at the high school level, but I realized I'm just out of touch. Amend my comment to say that the previous poster's classmates can't tell apart a search bar and a URL bar either.

I also question if you need to know what DNS is to know how URLs work. "This name is how your computer finds the website you are looking for" is sufficient for some level of knowing what a URL is and how it works.

And URLs, canonically, don't require DNS to function.

To a first approximation, you are talking out of your arse.

> That being said, I think developers have made some things more difficult to understand than they should be. While people understand what an address is, calling it a URL adds an extra layer to learn (even if it is just terminology). URLs would also be easier to understand if they were more consistent from the end user's perspective, much like someone's mailing address, rather than an engineer's perspective.

I really don't think they're more difficult than they need to be to be functional. Users' mental model of "an address" is incoherent and often context dependent. Both a URL and an email address are an "address" to most users. The really lost ones will refer to the search terms they use to get to a web page as the "address".

I don't really see URLs as different from addresses. Their in descending size order rather than ascending like physical addresses, but the rest of it is fairly consistent. Most people spend drastically less time thinking about URLs than physical addresses though. We teach children how to break down physical addresses, and we do not teach them to break down URLs. It's entirely unsurprising URLs are more confusing.

> It sounds like it makes things easier for the user. Yet all that it really accomplishes is creating confusion for the end user since it fails to differentiate two concepts.

Phrased differently, it frees them from the requirement to be able to differentiate those two concepts. Much like how my calculator can give me the square root of a number without me needing to know how to actually calculate that.

Amazing that something as simple as a hierarchical address can be so complicated at the same time.

Like why is the domain big endian and the path small endian? Leads to endless confusion and has been a source of many fraud attacks.

On the other hand a post address is equally unnatural. Street name-> street Number-> Postcode is not in consecutive order of significance. Yet everybody handles such without problem.

A big factor is also unreadable query strings like ?q= instead of simple paths like /search/.

We learned about URLs in my grade school, about 20 years ago.

Why should people have a basic knowledge of their tools? The world would be a better place?

What's the alternative? You want soldiers that don't know how to reload their gun? Drivers that don't understand how to use the steering wheel? Writers that don't know what a pencil sharpener is for?

I designed[1] the Chrome Omnibox, and appreciate a lot of the conjecture in this thread and other places - I wish I was as devious as many people assume we were.

The short version is that while many users can differentiate, it's more a problem of WHEN they differentiate - Chrome's entire design philosophy was around speed, and the omnibox was designed so that we had a simple destination that a user could focus on when they had decided to go somewhere else, but before they had decided WHERE they wanted to go to or HOW they were going to get there, and then make it so they wouldn't have to think about the HOW. If their search term autocompleted to a URL so they don't have to go through search, that was a win. This same philosophy was what drove putting the most visited pages on the New Tab Page (which was a blank page in most browsers at the time).

Some users plan out their actions before they take them, but I wanted to make everything as streamlined as possible so you could interact with Chrome as you thought about what you wanted, and not making you wait until after.

[1] I was the designer and occasional engineer, but the entire team takes the credit - everyone was behind the idea from before we even started work on Chrome, and many smart people deserve more credit than I do for how good it ended up being.

How do you feel that your invention was used to harm users and empower a quasi-global monopoly on search engines for the profit of tech billionnaire psychopaths and secret services? I don't think you had to have bad intentions (just like physicists and the atom bomb), but do you think the web is better for users without search boxes or search engine choices, and with Google forcing centralized AMP proxying down everyone's throat and therefore trying to actually completely remove URLs?

I don’t understand what exactly you’re mad at the guy about? He made an ultra-usable textbox… millions of people use it every day instinctively because it’s a great tool. You can’t crucify him because of your own personal qualms and completely ignore the flip side of the coin.

I'm not mad, just looking for retrospective insight. I don't know about you, but i've learnt great lessons by making mistakes, and i've found historical examples like Albert Einstein's reflections to be eye-opening.

Everybody does mistakes, and i'm not here to throw a stone. But i'm looking for lessons to be shared and learnt.

Please, if we were to call out every developer who works or has ever worked in AnyEvilCorp® then almost no one comes out alive, maybe except RMS

I've called out people in the past on works they seemed proud of. In this case, parent commenter explicitly acknowledges that their work had unintended consequences that they did not foresee. So i'm asking for critical lessons/insights they may have gained on the way.

Calling out is "hey you're doing such bullshit". "How do you feel that what you did had this consequence?" is opening a debate to share experiences, not "calling out".

So the excuse is everybody else jumped from the bridge too?

Speak for yourself.

It's a valid question to ask.

On the other hand, I despise this exact change in Firefox. When I want to search, I use the search bar. When I want to enter an address, I use the address bar. I have a clear picture of what I want to do, and it is annoying to no end if the browser confuses one for the other. For example, by auto-completing a search term to a URL instead (see: computers don't know what users want). That's why I personally prefer the ability to make that choice explicitly.

I get it, I'm not the kind of user this article is about. And maybe I'm just insufficiently exposed to the other paradigm to have become "fluent" in it. But I just wanted to point out that there definitely are users whose mental model differentiates search terms and URLs.

Thank you for posting this! I agree.

Address bar: I know what I want and I want it to only search in bookmarks and history.

Search bar: I don't quite know what I want and I want to open an internet search for it. And possibly give suggestions from actual search results.

Separating the address and search bars is the first thing I do when I set up a new Firefox. I hope it will stay around but I'm somewhat bracing myself for it to go away eventually.

>For example, by auto-completing a search term to a URL instead

Press space? Not every use case has to be addressed by UI elements, especially when we're talking about edge cases

Especially considering if it really grinds your gears, you can have your search and address bar be discrete pretty trivially.

FYI in Firefox you can revert to the correct, two-box implementation by checking "Settings > Search" . You can get two boxes back, and you can turn off search suggestions in the address bar.

Yeah, but I want that horizontal space to the right of the nav field for other things, so I configure to combine nav and search.

The power user thing in most "omnibars" that's been around for a long time at this point is that if you prefix search queries with a question mark ? you get an explicit search behavior (with your default search provider) and for instance auto-complete doesn't show recent URLs anymore.

Firefox (and Edge; don't know about Chrome as I don't use it) still have the concept of "Search Keywords" that you can prefix searches with to get into search modes for against specific non-default search providers. For instance in the Firefox bar you can type @wikipedia and go directly to a Wikipedia Search mode.

In Firefox at least you can focus address bar for search with default engine with Ctrl-K or Ctrl-J.

For anyone wondering why there's two shortcuts for the same action:

> If you have enabled Emacs-style text editing shortcuts in GNOME, they will also work in Firefox. When an Emacs text editing shortcut conflicts with the default shortcuts (as occurs with Ctrl+K), the Emacs shortcut will take precedence if focus is inside a text box (which would include the address bar and search bar). In such cases you should use the alternate keyboard shortcut if one is listed below.

https://support.mozilla.org/en-US/kb/keyboard-shortcuts-perf...

>Users don't have the mental model to differentiate search terms and a URL.

I don't think good software always aims for the lowest common denominator. It's not an unequivocal good that users' lack of education is catered to. Infantilizing users makes them more dependent on services, and less able to navigate the internet.

I feel like this is the same argument as "kids these days don't know anyone's phone number!" If it works it works

Phone numbers are stored in your phone's contacts app, a tool that does the remembering for you. Making an input box do two different things and then adding a heuristic to guess what the user wants might be handy, but can go wrong. And then, as the linked post demonstrates, users get used to that and treat every input box as if it can do both of these things. It's a race to the bottom. Now your users "got stupider" and you need to come up with even more stuff behind the scenes to fix it.

Nobody ever questioned that you have to learn driving a car. How to handle it, what would break it, traffic rules etc. Why don't we demand that cars become so smart that a 16yo can just buy one, jump on all the pedals and turn all knobs and be a safe while doing so because the car will prevent anything bad from happening? We somehow accept that driving a car is something non-trivial that has to be learned, has certain rules that need to be followed, etc.

> Nobody ever questioned that you have to learn driving a car.

At the same time, nobody ever questions what you don't learn to drive a car. You have one acceleration pedal; you don't have separate controls for the fuel/air mixture and ignition timings. You don't manually control the differential. You have one brake pedal, not separate ones for the front and rear brakes, and automatic anti-lock braking is even a useful safety feature.

Modern cars even have warnings for tire pressure, so the driver has less need to manually check.

Modern cars are so smart that a 16 year old can buy one and drive it without having to know how to rebuild the engine, unlike cars of a hundred years ago.

The trick is that we have a long cultural history of deciding just what aspects of car driving are non-trivial and need to be learned and what aspects can be automated successfully (or even beneficially). Consumer computing is still in its relative infancy, and it's a moving target.

You have an automatic transmission, so you don't have to know how to use the clutch (a thing people used to complain about the kids not learning)

But you don't confuse your automatic transmission for something which has another function -- perhaps your ignition.

Just US things... No one in Germany likes automatic transmission.

Oh Idunno... It's getting harder and harder to find even German cars with manual transmissions.

I think the more 'magic' they bake in to help non-technical users, the more it makes it harder to understand how the thing works for everyone else. As you allude to, cars have some fundamental rules that don't tend to get broken. The steering wheel, accelerator and brakes for example are generally all in the same place and operate the same between models, brands, updates.

It's just as straightforward to interpret the browser bar as only being a search bar that accepts any textual representation of intent from the user and does its best to interpret that intent (that's what search engines do, after all). And if that intent looks like a URL [0], then it might interpret that as your intent and load that URL. It's just one of many cases that are expected from a search engine–other might include showing a map for a query that looks like a location, doing arithmetic calculations for queries that look like arithmetic expressions, etc.

[0] Or a domain...even the most tech-savvy among us probably types in domains without examining the difference between a domain and a URL or how software might detect whether a given string looks like a domain

Except conflating the address and search bar, then adding search suggestions, has objective problems. It's a bad thing that typing URLs in your address bar leaks them to Google/etc.'s search completion service, allowing search engines to track which URLs people begin visiting (not just searches they make).

> "kids these days don't know anyone's phone number!" If it works it works

My passenger had my phone number memorized. I went to visit him when he called me from jail. I used his gmail password to get his contacts' phone numbers to try to get someone else to bail him out, but it was only $300. After a few days I paid up to bail him out. He missed his next court date. I went to the bail revocation hearing a couple months later, and got $200 of that $300 back (iirc).

I make an effort to regularly exercise my mental phonebook, on his account.

Who are your lifelines? - https://www.taxiwars.org/p/who-are-your-lifelines.html

That story reminds me of this homeless guy I helped out one night. He was wearing a suit, which is why I believed him, actually. He said he was a pilot, who just moved into a house but locked himself out with his wallet inside. So he attempted to break into his own house. The neighbors called the cops and he was arrested. Now he was “homeless” until he got things sorted out. I ran into him a few days later, he taught me how to fly and became a friend that I still email from time to time.

Are these blobs of text being generated by some language model or something? They seem like wildly incoherent generated stories 100% unrelated to the content of the article or this thread.

>Here's another example of this: IIRC one of the most frequently searched terms on Google is "facebook".

one of the other most frequently searched terms on google is "google", because people don't understand that the chrome omnibox performs a google search, and they want to go to google to perform a search from a dedicated "search" box rather than a URL input - essentially, they want what the firefox UI gave them.

if you're trying to draw any conclusions about user preferences from user behaviour, you have to accept that you're going to be wrong about a significant portion of your users. one workflow might be "better" for some portion of your users, but it won't mesh with the mental model of some other portion of your users. different groups will form different expectations and habits, and a single UI will not completely satisfy the expectations of every group.

>they want to go to google to perform a search from a dedicated "search" box rather than a URL input - essentially, they want what the firefox UI gave them.

I would not even be surprised if that kind of user would search "Google" on the dedicated "Search box" because they are more confortable performing the search on the actual Google website than on a UI element inside the browser toolbar.

> one of the other most frequently searched terms on google is "google",

Perhaps one of the most frequent, but surely nowhere near the majority. I'm guessing it's still an incredibly tiny portion of searches. The fact that Google Search also handles this case, for the likely tiny portion of cases where it occurs, seems totally reasonable to me, and hardly an indictment of any browser's UI paradigm.

I don't mean to indict any browser's UI paradigm.

My point was just that you shouldn't draw too many conclusions from user behaviour because user behaviour is often dumb and a poor signal of what users actually want.

the people who don't understand the difference are largely unperturbed by this workflow. which means it satisfies their expectations, even if it isn't optimized for them.

> Tech-savvy people will just type "facebook.com"

Tech-savvy people will type f, see that their browser completes "acebook.com", and press enter. That's a form of search, it's just more directly provided by the browser and unambiguously serving the user.

(Ideally, tech-savvy people will type f and see no signs of facebook whatsoever, but that's an orthogonal issue.)

> (Ideally, tech-savvy people will type f and see no signs of facebook whatsoever, but that's an orthogonal issue.)

I'm tech-savvy and I use Facebook all the time. :-P

Going off-topic here, but outside of Facebook's issues with privacy, I think Facebook is a fine platform. Most of the issues people have with Facebook (again, ignoring the privacy issues) are self-inflicted. You have a racist uncle and don't want to see his bigotry in your feed? Unfollow him. Or in my case, just unfriend him entirely. You want to follow your government representatives to see what's on their mind? Great, just don't engage with the morons in the comment section.

Facebook is a great way to keep up with what your friends are up to. A friend of mine recently went on a trip to Scotland, and I enjoyed his posts on what he was doing. When I got married, it was an easy way to share the photos, since nobody uses Flickr these days.

That works for issues like "annoying uncle" in the US, I don't think, for example, Rohingya in Myanmar could just un-friend the ethnic cleansing that Facebook contributed to.

https://www.nytimes.com/2018/11/06/technology/myanmar-facebo...

I think the biggest issue with Facebook I've learned recently is that on the margin their algorithms push you towards seeing things that increases their engagement numbers. I think it's hard to understand the effect on you from having every recommended item slightly not what you were asking for in the direction of making you mad.

> Most of the issues people have with Facebook (again, ignoring the privacy issues)

This is _massively_ shifting the goalposts. From a personal user perspective, the privacy issues are pretty much the entirety of the problem with Facebook.

But you're also missing the fact that Facebook _itself_ is A Bad Thing for the world, even if it is locally useful as a product. The parent commenter is presumably saying that it would be a good thing if Facebook did not exist (I agree), not just that it would be a good thing for each individual person if they didn't use Facebook. Even if _you_ unfollow your racist uncle, he's continuing to use the platform to hear and spread more racist invective.

The privacy issues don't actually bother me much because they're just being used to serve ads that I'm going to block anyways. I'm also sort of jaded when it comes to online privacy and operate under the assumption that every website and every device I use is trying to collect as much data about me as they can, and that the battle for privacy has been lost. I'll just run uBlock Origin to block the ads and most of the tracking.

> Even if _you_ unfollow your racist uncle, he's continuing to use the platform to hear and spread more racist invective.

They will ALWAYS have a platform, whether its Facebook, Parler, Gab, or Voat. It doesn't matter to me if they're on the same platform as me if I don't see it.

You and JoshTriplett are arguing different (and non-contradictory) points. You are saying that "the experience of using Facebook is pleasant and worthwhile for me, after taking some steps to alter my experience of it". They are saying "Facebook is a net-bad for society and we would all be better off if it didn't exist". These can both be simultaneously true. Overall-bad things can be pleasant for some users some of the time.

---

> The privacy issues don't actually bother me much[...]I'll just run uBlock Origin[...]

That's nice (and I, too, have my own ways for mitigating some of the worst bits of Facebook), but the very fact that you are having to take steps to "protect yourself" from Facebook demonstrates that it is a flawed platform. The vast majority of users are either going to be unaware of those issues, or unable to mitigate them. This is (and I want to emphasise this, because it runs contrary to a common attitude on HN) *not their fault*. It is not reasonable to expect the average computer user to know how to install extensions, nor which ones are important and trustworthy. It would be a wonderful world if they did, but . The fact that we technically-minded folks _can_ mitigate the effects of Facebook's undesirable behaviour for ourselves, does not excuse that behaviour.

(I also think it's a little unimaginative to think that Facebook's tracking is _only_ used to serve ads, and that you will always be able to mitigate all impacts of that tracking - but that's veering into speculation)

> It doesn't matter to me if they're on the same platform as me if I don't see it.

It should do. The harm of extremizing communication is not (only) to the people that see the communication - it is in the outcome of that positive feedback experience. The end result of "communication that you don't see" is "real-world actions that you do see". Would you similarly say "It doesn't matter to me if the KKK are gathering to plan lynchings, so long as they hold their meetings behind closed doors and hand out their flyers on streets I don't frequent"?

Facebook is the largest and most accessible social platform in the world. There _is_ a difference between "Hate groups can collaborate and gather members on Facebook" and "etc. etc. on Parler/Gab", because the former allows them _much_ greater access to fence-sitters, the undecided, the less-politically-aware, and so on. Everyone and their aunt is on Facebook, but (mostly) only those who already hold a particular ideology find themselves on the alternative sites. If hate-groups can only recruit from among those who already agree with their credo, they're going to find it harder to grow - which is a good thing.

> You have a racist uncle

I didn't have a racist uncle - at least not until he started spending too much time on Facebook and reading every piece of racist garbage the algorithm fed him.

I didn't have an anti-vaxxer mom, but guess what.

I type f, see Facebook, hit enter, and by the time I hit enter, Firefox swapped the history result with some search autocomplete that's not helpful at all.

There's an incredibly helpful setting for this. You can change it so the search suggestions are below your bookmark/history search.

Helps me tremendously in that exact issue.

Autocompleting strings is something that allows you to take a shortcut (if it guesses correctly) but is completely ignorable if you never use it. Like, seriously, if you never looked at the bar while you are typing, it behaves the exact same way as if it had never been implemented.

Meanwhile, the omnibar changes old UX behavior to allow Google to MITM requests if people don't recognize it and update how they work. And most people didn't.

I don't think this is correct at all. Google's decision was intentional to capture every request, even ones who are simply searching for a known domain. Users adapted to google's pattern and learned things the wrong way.

This is only one example of the damage caused by very-low-friction, addictive interfaces. Users have increasingly lost the capability to navigate hierarchical interfaces and increasingly use web search instead of looking around for the right button

They could do that anyway

> Users don't have the mental model to differentiate search terms and a URL

Why don't they? This is a somewhat serious question.

Part of it, in my opinion, is we are moving too fast, and giving up on users to quickly. With enough education, people will understand. But enough education = at least a generation or two, introducing it in schools at a young age, etc. But we just threw our hands up and said "oh well, they don't get it, lets make it simple".

URL = phone number. Search box = Phone book. They are two very different things, and having them both in the same box I feel confuses some people even more.

But it's not phone number vs phone book. It's phone number vs automated operator that will connect you where you are trying to go.

I am technical and I often just search instead of URL just because it requires less precision.

Google is just DNS with a few extra steps.

Ok, it's going to find a book on the shelf by its author/title or dewey decimal number, vs. looking it up in the card catalog by subject (and then going to the shelf).

That should appeal to the older crowd.

I think a good analogy is an automatic vs manual car. Most people prefer to have one pedal for the accelerator. But you still have many car enthusiasts (web enthusiasts) that know how cars (the web) work and want to control the mechanics behind the scenes. I wish it was still an option provided to users.

Google has a financial incentive for the Omnibar though. After searching for "facebook" users would click the first link, which might also happen to be an ad for Facebook. Effectively, turning a profit from the laziness of people not typing .com.

In the UK most cars are still manual. Is this really not the case in the US?

Having a clutch gives you a lot of fine grained control

Chrome's omnibar was designed to allow Google to MITM every web request, not to simplify things for the user.

Meanwhile FF still allows people to choose at least.

> But users don't care about any of that. Users don't have the mental model to differentiate search terms and a URL. Chrome's decision was correct. It's surprising how long Firefox stuck with their bad (IMHO) design.

I think you're making the mistake of thinking of "users" as a monolithic, lowest common denominator group. There are users who think like you describe, and there are users who do not. Also catering to people's existing (perhaps oversimple) mental models ensures those models never change.

One of the PR issues with Firefox is that it has/had many features that suited the latter group, but decided to drop them to ape Chrome.

>Think like a user, not an engineer.

So your technology, too, can become a melange of idiotic decisions where you leak things into a search engine!

While you're at it, only feed your kids ice cream and pizza because that's what they want to eat. Think like a kid, not an adult.

Its really easy for engineers to get a feeling of superiority because they 'get it' and to wonder why everyone else is so stupid, how could they use this thing they don't understand etc. etc. etc.

You've got weak spots in the tech stack too. I'd wager there are tools that you use without knowing even about their existence. The point isn't to do dumb shit, the point is to have empathy and consider the perspective of other humans who are just as intelligent as you but knowledgeable in different domains.

Even as a technical user, I Google "natwest" instead of typing in "natwest.com". Much safer than typing "natwest.co" and getting caught by phishing sites.

I was going to chime in with the same thing. I do this as well, particularly while browsing on mobile. It's an extra click, but I can leave off at the very least the TLD and get better spell checking.

"It's surprising how long Firefox stuck with their bad (IMHO) design."

Not considering Mozilla's "business model" is forwarding searches by default to Yahoo/Google for a price. Would be surprising if the search engine website contracting with Mozilla did not have an expectation that the Firefox UI would be conducive to submitting as many default "searches" as possible, whether intentionally or not. (User could type an address in the search bar and it would be forwarded to Google. She now has a Google cookie.) Would also be surprising if there was not a similar expectation that the Firefox settings UI make it relatively unlikely that users would change the search engine defaults.

Chrome's design is good for Google. More so than Firefox's design is good for [search engine partner]. For users who do not want to be accidentally sending queries to Google, the "Omnibox" design sucks. It also is incorrect in that it defeats the notion of what is and what is not a valid domain name or a valid URL. It does not educate users about the www, it allows them to stay ignorant. Would not be surprising if that is how Google views its users and prefers that they remain unaware of what is happening.

For example, look at how Google descibes the NID cookie. They are not lying but they are repeatedly suggesting that its most signifcant purpose is a user's configuration preferences. However its primary purpose is for advertising. A fully informed user with choice over consent is apparently not what Google wants.

https://policies.google.com/technologies/cookies?hl=en-US

The primary purpose of the "omnibox " is not to benefit the user (although it may do so for some users), it is to benefit Google's "business model" of collecting data about users and selling online ad services.

Even engineers do this. A couple of years ago I watched a colleague - a very talented software engineer - do a Google search.

He typed "google" into the search/URL bar and hit Enter.

This took him to a search results page, with Google as the first entry.

He clicked on Google.

This took him to the Google home page.

He entered his search query into the search box on the Google home page, and hit Enter.

Now he had the search results he wanted!

This can only happen on a browser where you haven't visited Google before, right?

On any "real" use-case scenario, typing "google" into the omnibar and hitting enter, should take you directly to google.com.

I can just type g and hit enter on my browser, and it still takes me to google.com

This is is an unpopular opinion on a place like Hacker News, but it is correct.

Users want access to the information they are looking for as quickly and frictionlessly as possible—the omni bar does just that.

ctrl+k and I know I'll be searching in a web search engine ctrl+l and I know I'll be opening an URL and if the domain doesn't exist it will fail

I can't really count anymore how often omnibar just searched for some local domain I tried to type in, in a search engine, instead of just telling me that domain doesn't exist because I made a typo or whatever.

Having to type https:// in front of every URL is also not very pleasant, especially on mobile phones.

Also browsers assuming that anything outside of PSL is to be searched in a search engine right away instead of resolved as a domain, is another annoyance for this user. At least the browser could be less stupid and consider local search domains part of PSL, but no.

So whether omnibar is more of a hindrance than help depends on the user/use case.

I also kinda like the distinction that ctrl+l is only searching locally in bookmarks/history and ctrl+k is public search.

Firefox has has an "omni" bar for a long time. Firefox could search your bookmarks and history as you type in an URL (in addition to keywords that could trigger searches). It was one of the key reasons (in addition to tabs) that I switched to Firefox back in the day. (Or am I misremembering this?)

It was back in Firefox 3 (released a few months before Chrome was announced) that Firefox’s address bar got the last significant pieces of that functionality and was dubbed the AwesomeBar.

I think that at that time Firefox still struggled with single-word searches from the address bar, treating them as domain names; if so, that was probably the last material difference (as distinct from social) before Firefox could reasonably remove the search box by default.

Yes, Firefox was one of the first to define the "omni" bar experience, but one of the slowest to remove (by default) a dedicated search box even after they proved most users were using the "omni" bar.

Google wanted all user input leaking to them. They got it. When I first saw this feature it felt like a natural progression...a real innovative feature. I didn't need it though. In the end, browsers probably should not have knowledge of search engines, for various reasons.

> Here's another example of this: IIRC one of the most frequently searched terms on Google is "facebook". Tech-savvy people will just type "facebook.com" but users will just search for "facebook" and click on the (hopefully) first result. That happens a ton.

I've actually switched to searching even for URLs I know, if they don't autocomplete from history. My typing accuracy isn't 100%, and I'm confident the top Google result will be the correct domain even if I make a typo (rather than some phishing clone squatting on an adjacent name).

Firefox actually had both, you could always search with the "main" bar. That's why having two has been so infuriating over the years, if I ever had to use Firefox, I've removed the search bar immediately.

However, one thing that Firefox failed to fix over dozens of versions have been one-word-searches, which Chrome handled correctly while Firefox tried to resolve that one word as a domain and didn't even fall back to search if that failed. That single thing has brought me back to Chrome within minutes every time I've wanted to try Firefox again.

Prefix one word searches with a question mark and then space ("? ") will always search in Firefox. Old Power User trick from decades ago even when Firefox still had a split search box by default. (There's also search keywords to force it to search with a non-default search provider such as starting a search with "@wikipedia " to search Wikipedia.)

Yes, I removed the search box and used search keywords long before they changed it.

For the single word searches, I've never noticed such a problem. It should try to resolve it first, in case it exists on your LAN, and falling back to search works fine for me.

On the other hand, I can't think of any advantage myself to having these things be discrete boxes, and I know what a search term and a URL is (as I'm sure most users do, give them some credit here). Corner cases rarely happen and if I think something might look like a url but is a search term, you can prepend the term with whitespace. Ultimately having them not be discrete boxes means I can see more of the url before it folds, which I prefer.

Corner cases happen every day where I work. Specifically, we have an internal domain, and many users, rather than typing finance.pdq.xy.pg (made up domain) we like to type finance to go directly to the correct server. But we can't because 9 times out of 10 chrome/safari/edge interpret as a search. I either have to remember to put in a trailing slash, i.e. finance/ or type the FQDN. Thats the only reason I use firefox--I can separate search from urls.

Perhaps this is a difference in incentives? Chrome was explicitly owned and developed by Google, a search engine. Google makes money when you search for things, rather than using the address bar, so they're incentivized to reward this behavior. (I've always suspected this is why Chrome's history search is 10 times worse than Firefox's, but that's another story).

But Firefox was derived from mozilla, where the URL bar would search if what you entered wasn't a URL.

> It's surprising how long Firefox stuck with their bad (IMHO) design

You can still enable the separate search bar in the settings if you want.

I find it interesting that the discussion about this feature is only about design/interaction principles.

I mean, maybe this is right and a merged search/URL bar is the best call from a design standpoint. I don't mind having one.

But do you really think that's why Google implemented a feature in Chrome that sends more people to Google?

It was, I'm sure, a revenue decision first and foremost, and Google is probably delighted when people identify it as something else.

The same logic applies just as well for any browser vendor who makes money off of search partners.

> Tech-savvy people will just type "facebook.com"

browsers used to do these these things (trying to add 'www.' or '.com' etc. if the domain doesn't resolve) automatically in the location bar (which wasn't the search bar, so there was no ambiguity):

https://support.mozilla.org/en-US/kb/search-web-address-bar#...

> IIRC one of the most frequently searched terms on Google is "facebook"

I wonder if this holds true for both PC and Mobile, because it is easier to perform a search and trust the search engine to deliver the link, rather than finish typing ".com".

Firefox did away with the search bar a long time ago. For a while, it was possible to bring it back through about:config, but I don't think it's possible to make the url bar a clean url bar anymore (this is an invitation for corrections). It's annoying because I navigate to pages via URLs, and all I don't want to send those to a search engine for "suggestions" (though, a locally-indexed history is nice). All too often, I'll type in a url and it will send that to a search engine, and then the page I wanted to go to is buried under ads. Useless. Especially now that Firefox is moving to paid ads in the "url" bar... blech.

> Firefox did away with the search bar a long time ago. For a while, it was possible to bring it back through about:config

The search bar still exists in Firefox if you want it to btw. I use it daily.

Interesting.

I didn't know you could add the search bar but when you do, the address bar still says "Search with <your default search> or enter address".

Also see sibling. I use Firefox daily and I have both a search bar that I use when I really want to search and I also don't see any "Search" stuff in my address bar. When I type "n" I can hit enter because it finds news.ycombinator.com in my history. All the next results are other things from my history which have an n in it.

I have browser.urlbar.showSearchSuggestionsFirst=false

My mistake, I meant to say url bar. Last time I tried this, the url bar had all the functionality of the search bar, which misses the point. Am I remembering wrong?

The settings screen has the option "show search in address bar results". You can turn that off.

It's the perfect setup really, you can turn it off, but if you want it you can type "? <search term>" into the address bar and it'll go to default search - or use one of your engine short codes as the first word "@google hackernews"

about:config -> keyword.enabled -> false (to stop sending possible search terms automatically to your default search engine, this includes when you type a URL firefox mistakes for a search term)

about:config -> browser.fixup.alternate.enabled -> false (to stop firefox trying a DNS lookup/connection with the term you typed with www. before, .com after)

This is something... but it isn't perfect. The address bar is still not purely an address bar. It still defaults to sending a query to a search engine (I like to use this to search for partial URL matches, for example, navigating git repos that I use frequently). That's what the search bar is for. Bah humbug. Even with all these options set, FF still encourages confusion between URLs and searching.

I'm not sure I understand.. there's an internal search (tabs, history, bookmarks) which sounds like what you're doing - there's no search engine used for this. Or do you mean the Search suggestions (Settings -> Search -> uncheck "Provide search suggestions" to disable, some of these suggestions are FF generated)