3 ms·

Back then you could test your UI on people that were a completely blank slate, having possible never used a computer before. You could focus on making the best

Back then you could test your UI on people that were a completely blank slate, having possible never used a computer before. You could focus on making the best possible UI there is. Now, users come already with certain habits and expectations developed by years of of computer use. This means, even if you really cared about the user experience of your customers you would only be able to make incremental improvements to the status quo or else be be rejected for being too different. Not that certain standards are not also valuable but it still a sobering thought that we are in a local optimum that we will never be able to escape from. (And I fear most products are not even hitting the local optimum but are getting worse when it comes to UI design but that is besides the point.)

phendrenad2 4y ago
The beautiful thing is, you can still just make a UI that mimics Windows 95, and get all of the benefits of that research.
vintagedave 4y ago
> Back then you could test your UI on people that were a completely blank slate, having possible never used a computer before. You could focus on making the best possible UI there is. This is a really great point. Like many others on this thread I miss "old" UI design, and like many posters can explain why. There's been a large transition to hiding UI, not marking interactable UI, and not spacially locating related UI elements. Recently I watched one of my nieces use her phone. She was trying to do something, frustrated and telling me, but had not yet asked me to help. (I offer to help, but not to do something for them - to guide but let them achieve their goal.) I realised she was essentially tapping the screen looking for something that responded to taps. She didn't know what would react. Now, she knows lists can be scrolled, buttons etc can be clicked... but does she? Because there are no buttons, often, today and the button she needed, which I could see onscreen, was just text, no border indicating clickability. I've heard anecdotally that the big UI designers in companies like Apple don't mind this because 'everyone already knows how to use their phones' and Macs / Windows. This ignores the young -- but we often ignore the young and expect them to learn -- and it ignores the elderly, those with bad eyesight, those who have not before now had the financial opportunity to own something, and so forth. There are lots, and always will be lots, of new people. And I think that designing such that it can be used only by those who already know how to use it is a failure. Not just because there are always those who don't know, but because designing such that it can be used by anyone is more optimal. And finally, that designing such that it can be be used by _anyone_ means _everyone_ can use it. Phrased another way, another golden quote I cannot attribute is, 'accessibility is for everyone.' Which is a way of saying, designing for the non-average user who may have difficulties makes usage much easier for the average user too.
- SebastianKra 4y ago
  > And I think that designing such that it can be used only by those who already know how to use it is a failure. Why? Cars aren't a failure because they need to be learnt, neither is reading, 10-finger typing, math, ... We teach these things because they make other tasks more efficient. Similarly, interfaces are converging to a somewhat complex, but actually quite consistent, set of conventions that allows you to accomplish an infinite set of tasks. You might argue that we could achieve the same thing with less complex conventions, but I disagree. According to the article, subjects struggled with nested folders, yet now, hierarchies are a hugely important concept.
  - retrac 4y ago
    On a general purpose device, I really am a believer that software should teach the user how to use it, or document how to use it. How else would you do it? A manual? Which is done in software these days, anyway. Making as much as possible discoverable in the interface is good design. The more of that which can be done, the less that needs to be learnt explicitly. Maybe not everything can be, but 40 years ago the Mac applied this principle consistently to its interface and it introduced tens of millions of people to computing who were otherwise put off by having to read a manual first.
    - SebastianKra 4y ago
      > How else would you [learn to use it]? Randomly tapping elements on the screen and seeing what happens is a perfectly acceptable learning strategy for a child. From there, you might learn that an accent color can represent links; That a row of buttons, where one is emphasised, might be a tab-bar; That holding on an element will open a secondary menu; That little arrows represent folded content; that a list on the left might change content on the right; That certain icons always lead to similar functionality; Side note: OP's neice was wildly tapping the screen, because she likely learned that nowadays nearly everything is interactive (Go on YouTube and find me one element that can't be tapped).
      - 4y ago
KronisLV 4y ago
> This means, even if you really cared about the user experience of your customers you would only be able to make incremental improvements to the status quo or else be be rejected for being too different. Isn't this a good thing, though? Creating software that most current users out there can be easily onboarded with and be productive and familiar with quickly, without too much work put into it. I recently did an article on using Linux as my daily driver for a week and admittedly picking a really boring desktop environment (XFCE) was a pleasant experience: https://blog.kronis.dev/articles/a-week-of-linux-instead-of-windows https://blog.kronis.dev/articles/a-week-of-linux-instead-of-... Sure, things were occasionally broken for technical reasons, but at least the UX itself wasn't what was causing problems or friction. It's very much like using something like LibreOffice (with the default dropdown interface, or the optional ribbon one) after other office packages - it feels close enough to what you're used to not to be completely jarring, even if there might be other issues with it. > Not that certain standards are not also valuable but it still a sobering thought that we are in a local optimum that we will never be able to escape from. (And I fear most products are not even hitting the local optimum but are getting worse when it comes to UI design but that is besides the point.) In my eyes, not challenging the status quo and making incremental improvements is a good idea for the majority of developers out there (even in web development, where you have your navbars and your sidebars and so on), let the trendsetters take the big risks and take what is proven to work in the long run from them, without being impacted negatively by their failures. I do, however, support the idea of usability studies and looking for what might be better than what we currently have, but redesigns for the sake of being "modern" or "trendy" just seems like designers going for the fun low hanging fruit, like large companies just changing around all of their product icons, instead of working on, say, accessibility, text contrast and so on.
- willnonya 4y ago
  The problems with using Linux generally aren't desktop environment issues but more the diskljpinted nature of Linux. XCFE is my preferred Linux desktop though, flexible without a lot of baggage.

phendrenad2 4y ago

The beautiful thing is, you can still just make a UI that mimics Windows 95, and get all of the benefits of that research.

vintagedave 4y ago

> Back then you could test your UI on people that were a completely blank slate, having possible never used a computer before. You could focus on making the best possible UI there is. This is a really great point. Like many others on this thread I miss "old" UI design, and like many posters can explain why. There's been a large transition to hiding UI, not marking interactable UI, and not spacially locating related UI elements. Recently I watched one of my nieces use her phone. She was trying to do something, frustrated and telling me, but had not yet asked me to help. (I offer to help, but not to do something for them - to guide but let them achieve their goal.) I realised she was essentially tapping the screen looking for something that responded to taps. She didn't know what would react. Now, she knows lists can be scrolled, buttons etc can be clicked... but does she? Because there are no buttons, often, today and the button she needed, which I could see onscreen, was just text, no border indicating clickability. I've heard anecdotally that the big UI designers in companies like Apple don't mind this because 'everyone already knows how to use their phones' and Macs / Windows. This ignores the young -- but we often ignore the young and expect them to learn -- and it ignores the elderly, those with bad eyesight, those who have not before now had the financial opportunity to own something, and so forth. There are lots, and always will be lots, of new people. And I think that designing such that it can be used only by those who already know how to use it is a failure. Not just because there are always those who don't know, but because designing such that it can be used by anyone is more optimal. And finally, that designing such that it can be be used by _anyone_ means _everyone_ can use it. Phrased another way, another golden quote I cannot attribute is, 'accessibility is for everyone.' Which is a way of saying, designing for the non-average user who may have difficulties makes usage much easier for the average user too.

SebastianKra 4y ago
> And I think that designing such that it can be used only by those who already know how to use it is a failure. Why? Cars aren't a failure because they need to be learnt, neither is reading, 10-finger typing, math, ... We teach these things because they make other tasks more efficient. Similarly, interfaces are converging to a somewhat complex, but actually quite consistent, set of conventions that allows you to accomplish an infinite set of tasks. You might argue that we could achieve the same thing with less complex conventions, but I disagree. According to the article, subjects struggled with nested folders, yet now, hierarchies are a hugely important concept.
- retrac 4y ago
  On a general purpose device, I really am a believer that software should teach the user how to use it, or document how to use it. How else would you do it? A manual? Which is done in software these days, anyway. Making as much as possible discoverable in the interface is good design. The more of that which can be done, the less that needs to be learnt explicitly. Maybe not everything can be, but 40 years ago the Mac applied this principle consistently to its interface and it introduced tens of millions of people to computing who were otherwise put off by having to read a manual first.
  - SebastianKra 4y ago
    > How else would you [learn to use it]? Randomly tapping elements on the screen and seeing what happens is a perfectly acceptable learning strategy for a child. From there, you might learn that an accent color can represent links; That a row of buttons, where one is emphasised, might be a tab-bar; That holding on an element will open a secondary menu; That little arrows represent folded content; that a list on the left might change content on the right; That certain icons always lead to similar functionality; Side note: OP's neice was wildly tapping the screen, because she likely learned that nowadays nearly everything is interactive (Go on YouTube and find me one element that can't be tapped).
    - 4y ago

KronisLV 4y ago

> This means, even if you really cared about the user experience of your customers you would only be able to make incremental improvements to the status quo or else be be rejected for being too different. Isn't this a good thing, though? Creating software that most current users out there can be easily onboarded with and be productive and familiar with quickly, without too much work put into it. I recently did an article on using Linux as my daily driver for a week and admittedly picking a really boring desktop environment (XFCE) was a pleasant experience: https://blog.kronis.dev/articles/a-week-of-linux-instead-of-windows https://blog.kronis.dev/articles/a-week-of-linux-instead-of-... Sure, things were occasionally broken for technical reasons, but at least the UX itself wasn't what was causing problems or friction. It's very much like using something like LibreOffice (with the default dropdown interface, or the optional ribbon one) after other office packages - it feels close enough to what you're used to not to be completely jarring, even if there might be other issues with it. > Not that certain standards are not also valuable but it still a sobering thought that we are in a local optimum that we will never be able to escape from. (And I fear most products are not even hitting the local optimum but are getting worse when it comes to UI design but that is besides the point.) In my eyes, not challenging the status quo and making incremental improvements is a good idea for the majority of developers out there (even in web development, where you have your navbars and your sidebars and so on), let the trendsetters take the big risks and take what is proven to work in the long run from them, without being impacted negatively by their failures. I do, however, support the idea of usability studies and looking for what might be better than what we currently have, but redesigns for the sake of being "modern" or "trendy" just seems like designers going for the fun low hanging fruit, like large companies just changing around all of their product icons, instead of working on, say, accessibility, text contrast and so on.

willnonya 4y ago
The problems with using Linux generally aren't desktop environment issues but more the diskljpinted nature of Linux. XCFE is my preferred Linux desktop though, flexible without a lot of baggage.