4 ms·
Hmm. There is a (minor) flaw in your argument. But before I get to that, there are a few places where non-developers have user interfaces that reward expertise
by webmaven 10y ago
Hmm. There is a (minor) flaw in your argument.
But before I get to that, there are a few places where non-developers have user interfaces that reward expertise. One prominent example are Bloomberg terminals: http://graphics8.nytimes.com/images/2013/05/13/business/sub-jp-bloomberg/sub-jp-bloomberg-superJumbo.jpg http://graphics8.nytimes.com/images/2013/05/13/business/sub-...
Notice that the interface (which is extremely customizable, feel free to look up other images, each will be rather different) is more-or-less a tiling window manager with terminals in each window that have rich media, nice fonts, non-ASCII UI elements (albeit ones that seem somewhat stuck in the '90s), etc. Quite bit to learn from there.
So here is the flaw: I am afraid the reason most interfaces that non-technologists have to use cater to intuition and a pleasant appearance rather than rewarding expertise is all to simple: No one wants to spend any time becoming an expert at using a gajillion specialized, but infrequently used, software interfaces, each of which would, according to your ideal, be optimally designed to allow an expert to perform the associated task efficiently and well.
The average person pays off their credit card once a month, and pays their taxes once a year. The incentive for them to learn to do these things more efficiently isn't very compelling, and the number of people who have 20 active credit cards for whom it would be compelling isn't large enough to be worth creating an expert UI for (that may change as another couple of billion people get online, if software markets don't fragment further)
Now, all that said, there is certainly a lot of room for improvement in web-based user interfaces: animations can be faster and more subtle, the use of whitespace can be reduced, typography can be more restrained, decoration and color can be used only when it conveys information (basically, everything Tufte has been telling us for a couple of decades).
Windows' Metro and Modern, and Google's Material, are both nice steps in that direction (with the exception of animation), and each represents a lot of difficult design work by large design and development teams. Less certainly is more, both in the sense of the return it offers but also in the investment required. The simpler and less cluttered a user interface is, the more you have to sweat the tiniest of details. This post on redesigning bits of the Chrome browser's, uh, chrome is a good case in point: https://medium.com/google-design/redesigning-chrome-desktop-769aeb5ab987 https://medium.com/google-design/redesigning-chrome-desktop-...
You can expect user interface redesign churn to slow down only once display resolutions stop climbing (because they have exceeded what can be distinguished by the human eye) and form factors stop changing (because the only remaining meaningful constraints are ergonomic).
- brandur 10y agoThanks for reading! > First, there are a few places where non-developers have user interfaces that reward expertise. One prominent example are Bloomberg terminals: I didn't mean to say that it was just developers that have access to these sorts of power tools, but it is the most common case. I would have actually used Bloomberg terminals as an example to support my arguments, and in general am hugely in favor of this sort of app that rewards the time invested in learning it all the way up to advanced levels. > The average person pays off their credit card once a month. The incentive for them to learn to do it more efficiently isn't very compelling, and the number of people who have 20 active credit cards for whom it would be compelling isn't large enough to be worth creating an expert UI for. Yes totally, but what if you had just one common UI that was pretty standard and which your credit card company could easily plug into while building interfaces for their users? Modern native apps for smartphones are probably the best example there because even though they're not perfectly consistent, at least they have standardized toolbars, navigation, and controls (far beyond what you get on the web). I think this idea could be take even further.
- webmaven 10y ago> what if you had just one common UI that was pretty standard Yes, well, the reasons that is unlikely is because banks need to differentiate their offerings and... you know what, scratch that. Just observe that this hasn't happened on the desktop in the way you describe, though every so often someone does try to reintermediate an industry in that way, which usually only works if you (a) control the distribution platform, so can bundle in the standard UI to the platform, and (b) manage to fool the industry into going along, eg. iTunes, Kindle). Banks, specifically, have yet to fall for any of the attempts along those lines, and there have been quite a few. The situation is unlikely to change until/unless newly formed banks embrace splitting out the "dumb money pipe" as shared infrastructure and modern standardized interfaces so their value added services are separate (but integrated). To some extent you see glimmers of this in medicine with (almost, but not quite) portable electronic medical records. What progress has come about is solely due to the government ratcheting up the carrots and sticks to get "meaningful use" to happen, and those may get rolled back now.
- 10y ago