6 ms·

As a designer, I agree completely. The problem is: interaction design really is a discipline in its absolute infancy. It's not graphic design. It's not engin

by state 10y ago

As a designer, I agree completely. The problem is: interaction design really is a discipline in its absolute infancy. It's not graphic design. It's not engineering. It's not architecture. Pretty much anyone can just pick it up and start doing it. And the people who decide to ship shitty designs you're complaining about are more concerned with 'pageviews' or 'installs' (and in turn, their bonus) than actual usability. Compare the process by which an app or website design makes it out into the world to how a patch makes it into an open source project. Anyone can start programming, but to be accepted as a 'good' programmer you're inevitably vetted by other, better programmers. The ad-hoc peer-review of OSS actually works reasonably well. Most importantly: it can be very critical. Most designers I meet today have laughably thin skin. Try comparing the most successful projects on GitHub to those on Dribble. Dribble makes my skin crawl. Anyway, you're right. Most contemporary interaction design is absolutely awful. It's going to take time for the rules of the discipline to evolve. We'd probably see this improve a lot if the apps and sites we used weren't optimized for 'engagement'. They just look like slot machines designed by Portlandia characters.

knieveltech 10y ago
I would agree that interaction design for the web is in it's infancy inasmuch as it still displays a tendency to stick it's fork in the power outlet when left unsupervised. Otherwise I disagree strongly. All of the design lessons learned leading up to the .com implosion have had to be collectively re-learned at least twice (remember jello mold layouts? eyeroll) in the last 15 years. The root problem: designers either refuse to make peace with the fact that layout and control interfaces have been largely solved problems for close to 20 years now, or they fail miserably at conveying these limitations to potential clients.
- state 10y ago
  I'm not saying progress isn't happening — I'm saying that there's a big difference between a discipline and what interaction design is in 2016. I think these incremental steps are good, but it's going to take the equivalent of the Bauhaus to solve this problem. You need a strong foundational movement to anchor people's work. You're right that the problems have been solved, but they aren't being upheld very well.
- flukus 10y ago
  A third problem, they have to accept that they don't have complete control, only suggestions.
Silhouette 10y ago
It's hard to disagree, but I think the problem is more fundamental: way, way too many people working in tech assume that because Google and Facebook and Apple and Microsoft are huge and financially successful overall, that means they know what they're doing and should be emulated. This has been true in recent years with the trendy but awful design. It's not just the typography. The obsession with flat design and giving up all the distinctiveness and affordances of other styles might even be worse. It's also been going on for a long time with assuming if Google says it in its guidelines for webmasters then it must be good, as opposed to being good for Google's search engine. It's even going on with software, specifically the various libraries and tools these organisations produce. Some of those libraries and tools are competently executed, but few are as remarkable or transformative as the hero worship might suggest. This culture of emulation naturally leads to the kind of bland, homogeneous, poorly conceived work that is all too prevalent in our industry today. How could it not? Ironically, I was just reading another HN post about a jobs board for older geeks, and the related discussion about the endemic problem with ageism in our industry. What we are talking about in this discussion is literally what happens when you hire a bunch of very young, very inexperienced, very untrained people, given them very snazzy hardware for their very good eyesight and their very dexterous hands to work with, give them effectively unlimited resources but very limited accountability and management, and leave them to run the show. Anyone who wants to do better certainly can do it, just by studying relevant fields like graphic design and usability and by acquiring the necessary skill and experience (or, for businesses, by hiring people like that). But unsurprisingly, the people who are willing to go against the grain tend to be those who have been around for long enough to have other ideas and to have seen the same sorts of mistakes made before, and who wants to listen to them when you can go watch someone from the Apple team wax lyrical about their new barely legible font, someone from the React team misunderstanding basic MVC, or someone from Google who has never run a small business web site in their life telling everyone what makes good quality content?
- freelancezombie 10y ago
  Amen.
- nercht12 10y ago
  Eh... I think that's true in part, but I've also met people who really love that flat Metro design. It's hideous in my book, but it's trendy, yes, but that's because some people like it. More so than that, looking at website design from a monetary perspective: It's fast, cheap, and easy. It takes alot less time to put together a sleek modern website that utilizes the flat but nifty new CSS3 features than sitting down at the drawing board and creating something truly beautiful... which, when you're done, may or may not turn out how you dreamed. With websites in high demand, it shouldn't be any wonder than a typical but FAST-and-CHEAP-to-produce site will win out for the majority of cases.
adrianratnapala 10y ago
Everything you say might be right, but there is something extra to jonstokes' comment: ametures and nerds often produce more functional sites than real designers. The mechanism here is that whatever esthetic and psychological principles are in play, an interface must also present the needed info and functionality. Craigslist and HN do that in a more-or-less bullshit free way. UX designers can get caught up in other concerns. In fact there is a systematic bias: since we know that users don't like too much information, UX pros will want to hide info and functionality. But since no one likes boring designs, they also want to add cooling-looking (i.e. distracting) stuff.
- flukus 10y ago
  I find that once you're trying to hide info or functionality it's because you don't know the users workflow. Understanding workflow is the key to usability, at least with LOB apps, trying to cram everyone's needs into a single table or building UI's that are essentially skins over database tables make for some awful UI's.
  - tluyben2 10y ago
    > cram everyone's needs into a single table or building UI's that are essentially skins over database tables make for some awful UI's ... which, after you learn how to use them, often work much faster than the alternatives. There are anecdotes a-plenty here like our local (tiny) bank who held out replacing their terminal based with their Windows version because in tests the keyboard based, archaic, every-on-one-screen application was far more productive than the new one and this particular application is only to be used by seniors in this bank who, because of the size, are not replaced / added very often. The Windows version, although not even that well designed, has been created to be faster to learn for new employees and the screens have been designed to be aesthetically pleasing and therefor sparsely layed out over many more pages. The old staff hates it, the new staff (not very many) is just used to this kind of easy to learn, nothing to master kind of thing. They don't have to be exclusive though; what happened to the 'expert mode' in apps/applications? That button you ram if you are an expert and suddenly 100 screens are packed into one, keyboard navigation is switched on and I can do my work much faster than clicking HUGE easy to learn, text heavy buttons? In a throw-away society, all this interaction design makes sense; no-one is at a job long enough to really get into software and you don't get through the trial period if you don't learn using it fast enough. In that regard you need big empty pages with videos explaining what everything does and a boned out interaction process that has been carefully thought out. I think we need that if the budget is there, but I think we need the expert mode as well and that, as far as I have seen in my life time, (interaction) designers are not very good at packing dense nor is it needed; if you are an 'expert', you'll cope (or don't care in some cases) about the (interaction) design.
freelancezombie 10y ago
Designer here, how can we solve this? I'd love to have my work peer reviewed because design is so subjective and I feel like I'm probably making elementary mistakes despite learning for years.
- nercht12 10y ago
  In case you haven't read this yet: http://www.joelonsoftware.com/uibook/fog0000000249.html http://www.joelonsoftware.com/uibook/fog0000000249.html It's old, but still useful.
- tracker1 10y ago
  Serious suggestion... Open up a text editor, and lay out your functionality in 80 columns wide, and 25 lines high... This is how it was in the really old days with text interfaces... everything comes up fast, you can create boxes to isolate information, but literally can only fit so much text on the screen... Now, take this, and turn it into a relatively scaling gui... for mobile, go 40x50. There is something to be said from experience as an ansi artist and a sysop in the bad old days before the web.
  - anotheryou 10y ago
    One big problem I see is, that power users need something quite different than first time users. You can make some compromise by tucking away advanced features without removing them, but in general it's still an issue. First time users are probably what you focus on, the rest just has to be a tad better than the competitors. First time users need to be convinced by pretty design and a super simple linear workflow that solves just one problem. Power users couldn't care less about design (SAP is still alive). They love it feature packed and want to access as many options as possible with few clicks. The extremes are icons without labels, color coded shortcuts on every keyboard key¹ or command line interfaces. ¹Avid-Keyboard for Video editing: http://www.drted.com/IMG_1225.JPG http://www.drted.com/IMG_1225.JPG color coding in general is mostly ugly, but helpful. edit: but in the end there is just a lot of bad design, that at best makes things pretty, but often not more useful.
gcb0 10y ago
nope. it's not. everyone who studied industrial design have it down to a science. what happens is that webdesigners and silicon valley ued/UX people came from graphic design. so it's new to them. and the industrial design people are shunned away just like system analysts are on the coding side.

by state 10y ago

knieveltech 10y ago

I would agree that interaction design for the web is in it's infancy inasmuch as it still displays a tendency to stick it's fork in the power outlet when left unsupervised. Otherwise I disagree strongly. All of the design lessons learned leading up to the .com implosion have had to be collectively re-learned at least twice (remember jello mold layouts? eyeroll) in the last 15 years. The root problem: designers either refuse to make peace with the fact that layout and control interfaces have been largely solved problems for close to 20 years now, or they fail miserably at conveying these limitations to potential clients.

state 10y ago
I'm not saying progress isn't happening — I'm saying that there's a big difference between a discipline and what interaction design is in 2016. I think these incremental steps are good, but it's going to take the equivalent of the Bauhaus to solve this problem. You need a strong foundational movement to anchor people's work. You're right that the problems have been solved, but they aren't being upheld very well.
flukus 10y ago
A third problem, they have to accept that they don't have complete control, only suggestions.

Silhouette 10y ago

It's hard to disagree, but I think the problem is more fundamental: way, way too many people working in tech assume that because Google and Facebook and Apple and Microsoft are huge and financially successful overall, that means they know what they're doing and should be emulated. This has been true in recent years with the trendy but awful design. It's not just the typography. The obsession with flat design and giving up all the distinctiveness and affordances of other styles might even be worse. It's also been going on for a long time with assuming if Google says it in its guidelines for webmasters then it must be good, as opposed to being good for Google's search engine. It's even going on with software, specifically the various libraries and tools these organisations produce. Some of those libraries and tools are competently executed, but few are as remarkable or transformative as the hero worship might suggest. This culture of emulation naturally leads to the kind of bland, homogeneous, poorly conceived work that is all too prevalent in our industry today. How could it not? Ironically, I was just reading another HN post about a jobs board for older geeks, and the related discussion about the endemic problem with ageism in our industry. What we are talking about in this discussion is literally what happens when you hire a bunch of very young, very inexperienced, very untrained people, given them very snazzy hardware for their very good eyesight and their very dexterous hands to work with, give them effectively unlimited resources but very limited accountability and management, and leave them to run the show. Anyone who wants to do better certainly can do it, just by studying relevant fields like graphic design and usability and by acquiring the necessary skill and experience (or, for businesses, by hiring people like that). But unsurprisingly, the people who are willing to go against the grain tend to be those who have been around for long enough to have other ideas and to have seen the same sorts of mistakes made before, and who wants to listen to them when you can go watch someone from the Apple team wax lyrical about their new barely legible font, someone from the React team misunderstanding basic MVC, or someone from Google who has never run a small business web site in their life telling everyone what makes good quality content?

freelancezombie 10y ago
Amen.
nercht12 10y ago
Eh... I think that's true in part, but I've also met people who really love that flat Metro design. It's hideous in my book, but it's trendy, yes, but that's because some people like it. More so than that, looking at website design from a monetary perspective: It's fast, cheap, and easy. It takes alot less time to put together a sleek modern website that utilizes the flat but nifty new CSS3 features than sitting down at the drawing board and creating something truly beautiful... which, when you're done, may or may not turn out how you dreamed. With websites in high demand, it shouldn't be any wonder than a typical but FAST-and-CHEAP-to-produce site will win out for the majority of cases.

adrianratnapala 10y ago

Everything you say might be right, but there is something extra to jonstokes' comment: ametures and nerds often produce more functional sites than real designers. The mechanism here is that whatever esthetic and psychological principles are in play, an interface must also present the needed info and functionality. Craigslist and HN do that in a more-or-less bullshit free way. UX designers can get caught up in other concerns. In fact there is a systematic bias: since we know that users don't like too much information, UX pros will want to hide info and functionality. But since no one likes boring designs, they also want to add cooling-looking (i.e. distracting) stuff.

flukus 10y ago
I find that once you're trying to hide info or functionality it's because you don't know the users workflow. Understanding workflow is the key to usability, at least with LOB apps, trying to cram everyone's needs into a single table or building UI's that are essentially skins over database tables make for some awful UI's.
- tluyben2 10y ago
  > cram everyone's needs into a single table or building UI's that are essentially skins over database tables make for some awful UI's ... which, after you learn how to use them, often work much faster than the alternatives. There are anecdotes a-plenty here like our local (tiny) bank who held out replacing their terminal based with their Windows version because in tests the keyboard based, archaic, every-on-one-screen application was far more productive than the new one and this particular application is only to be used by seniors in this bank who, because of the size, are not replaced / added very often. The Windows version, although not even that well designed, has been created to be faster to learn for new employees and the screens have been designed to be aesthetically pleasing and therefor sparsely layed out over many more pages. The old staff hates it, the new staff (not very many) is just used to this kind of easy to learn, nothing to master kind of thing. They don't have to be exclusive though; what happened to the 'expert mode' in apps/applications? That button you ram if you are an expert and suddenly 100 screens are packed into one, keyboard navigation is switched on and I can do my work much faster than clicking HUGE easy to learn, text heavy buttons? In a throw-away society, all this interaction design makes sense; no-one is at a job long enough to really get into software and you don't get through the trial period if you don't learn using it fast enough. In that regard you need big empty pages with videos explaining what everything does and a boned out interaction process that has been carefully thought out. I think we need that if the budget is there, but I think we need the expert mode as well and that, as far as I have seen in my life time, (interaction) designers are not very good at packing dense nor is it needed; if you are an 'expert', you'll cope (or don't care in some cases) about the (interaction) design.

freelancezombie 10y ago

Designer here, how can we solve this? I'd love to have my work peer reviewed because design is so subjective and I feel like I'm probably making elementary mistakes despite learning for years.

nercht12 10y ago
In case you haven't read this yet: http://www.joelonsoftware.com/uibook/fog0000000249.html http://www.joelonsoftware.com/uibook/fog0000000249.html It's old, but still useful.
tracker1 10y ago
Serious suggestion... Open up a text editor, and lay out your functionality in 80 columns wide, and 25 lines high... This is how it was in the really old days with text interfaces... everything comes up fast, you can create boxes to isolate information, but literally can only fit so much text on the screen... Now, take this, and turn it into a relatively scaling gui... for mobile, go 40x50. There is something to be said from experience as an ansi artist and a sysop in the bad old days before the web.
- anotheryou 10y ago
  One big problem I see is, that power users need something quite different than first time users. You can make some compromise by tucking away advanced features without removing them, but in general it's still an issue. First time users are probably what you focus on, the rest just has to be a tad better than the competitors. First time users need to be convinced by pretty design and a super simple linear workflow that solves just one problem. Power users couldn't care less about design (SAP is still alive). They love it feature packed and want to access as many options as possible with few clicks. The extremes are icons without labels, color coded shortcuts on every keyboard key¹ or command line interfaces. ¹Avid-Keyboard for Video editing: http://www.drted.com/IMG_1225.JPG http://www.drted.com/IMG_1225.JPG color coding in general is mostly ugly, but helpful. edit: but in the end there is just a lot of bad design, that at best makes things pretty, but often not more useful.

gcb0 10y ago

nope. it's not. everyone who studied industrial design have it down to a science. what happens is that webdesigners and silicon valley ued/UX people came from graphic design. so it's new to them. and the industrial design people are shunned away just like system analysts are on the coding side.