5 ms·

This makes some good point, except for (at least) one: > 3. Simplify your UI. > […] Taking away features can be so much harder than adding features, >

This makes some good point, except for (at least) one: > 3. Simplify your UI. > […] Taking away features can be so much harder than adding features, > but subtracting often creates the most value for your users. Really? By this reasoning, Gnome 3 must be far better than anything before, and Nautilus 3.6 must have greatly improved over previous versions. Heck, based on such a general statement, ‘mail’ is probably also better than any GUI email client and ‘cat’ must be the best text editor ever. ‘If you create a system that any idiot can use, then only idiots will find it useful.’ still holds, even though dumbed-down versions of everything become somewhat more popular from time to time.

johndavidback 14y ago
I think you're right in some circumstances and perhaps not so in others. Gmail in it's current amalgamation is slick and some features are hidden in menus that really don't get that much action. This is certainly a preferable user experience. At worke we track Omniture analytics on every click event on our websites, half of most buttons/links/actions are never interacted with.
- bwertz 14y ago
  Interesting data point that 50% of your products features are never been interacted with - I would assume that this is true for many products
- claudius 14y ago
  The difference is in simplifying the UI versus taking features away. The former is perfectly fine, the best UIs (in my opinion) don’t exist/are perfectly invisible. However, taking features away in order to ‘simplify the UI’ is the wrong approach – making these features accessible in such a way that they don’t clutter the UI, preferably sorted/ranked by frequency of usage certainly is superior to dropping them. That's why Emacs has M-x/M-:, menus were invented, every sane desktop environment has some way to bring up a programme by its name and human-readable configuration files were invented: Such that the most obvious parts of the UI can offer the most frequently used features and all other, less-frequently used but still necessary, features are accessible in some way.
- micahmcfarland 14y ago
  This roughly correlates with "the 80-20 rule": 80% of your users will use only 20% of your features.
yalooze 14y ago
Basecamp, by 37signals, is a great example of this in practice. They stripped back features so they could focus on the core. In the end they were left with a great (IMO) project management SaaS. No, it doesn't do everything, but what it does do it does really well.
hcarvalhoalves 14y ago
Concise != Poor I see you're a Physicist, so maybe an analogy can help: would you use the Standard Model to calculate how long your car trip will take? Probably not, Newton's formulas suffice for that, even though those only explain a portion of the world. You certainly understand the concept of elegance. It's the same with UI design. You want to achieve a balance between function and conciseness. When conciseness suffers, it's because you're working at the wrong abstraction level or have weak conceptual integrity, not that your UI is too useful.
- claudius 14y ago
  No. But I wouldn’t want some UI designer to remove the ‘SM feature’ entirely simply because he thinks that it’s barely used and only clutters the UI. The user has to choose the right tool for the right job and as a designer it is your job to make that decision as easy as possible – surely that includes sometimes not offering a feature, but that will simply force users to choose between your tool and another one if they have a specific job to do. Not to mention that you were probably aiming for something like (special) relativity rather than the SM if you want to compare it to Newton’s formulae. :)
  - hcarvalhoalves 14y ago
    > No. But I wouldn’t want some UI designer to remove the ‘SM feature’ entirely simply because he thinks that it’s barely used and only clutters the UI. If the designer is making assumptions based on what "he thinks" and defining function from that, this is simply a poor product and design process. This is designing backwards: the UI is a result of it's function, not the other way around. Given you mention Gnome 3 as an example, I can see where you coming from, and why the trauma about design. Gnome is not necessarily a badly designed product, they just don't cater to your demographics, or most of the Linux user base actually. You might like reading this short book called "Notes on the Synthesis of Form" [1], which is an attempt to formalize the process of design. I highly recommend it, it can be enlightening for people of all disciplines. [1] http://en.wikipedia.org/wiki/Notes_on_the_Synthesis_of_Form http://en.wikipedia.org/wiki/Notes_on_the_Synthesis_of_Form > Not to mention that you were probably aiming for something like (special) relativity rather than the SM if you want to compare it to Newton’s formulae. :) I'm no physicist obviously, but what I was trying to say is: you don't explain the motion of a car thinking about how particles interact, it's the wrong level of abstraction; meanwhile classical mechanics explains this elegantly, even though it doesn't explain everything. There's a balance between conciseness and usefulness, which is the same challenge faced by design, or any other discipline that deals with creating models really.
DanBC 14y ago
I agree. I really hate the trend of removing useful functionality in the aim of, what? I dunno. I'm happy to have that functionality hidden away in text mode config files if needed, to keep the gui nice and clean. That way you just need a check box saying "Turn on advanced features? [ ] Caution: these are not supported and you're on your own".

johndavidback 14y ago

I think you're right in some circumstances and perhaps not so in others. Gmail in it's current amalgamation is slick and some features are hidden in menus that really don't get that much action. This is certainly a preferable user experience. At worke we track Omniture analytics on every click event on our websites, half of most buttons/links/actions are never interacted with.

bwertz 14y ago
Interesting data point that 50% of your products features are never been interacted with - I would assume that this is true for many products
claudius 14y ago
The difference is in simplifying the UI versus taking features away. The former is perfectly fine, the best UIs (in my opinion) don’t exist/are perfectly invisible. However, taking features away in order to ‘simplify the UI’ is the wrong approach – making these features accessible in such a way that they don’t clutter the UI, preferably sorted/ranked by frequency of usage certainly is superior to dropping them. That's why Emacs has M-x/M-:, menus were invented, every sane desktop environment has some way to bring up a programme by its name and human-readable configuration files were invented: Such that the most obvious parts of the UI can offer the most frequently used features and all other, less-frequently used but still necessary, features are accessible in some way.
micahmcfarland 14y ago
This roughly correlates with "the 80-20 rule": 80% of your users will use only 20% of your features.

yalooze 14y ago

Basecamp, by 37signals, is a great example of this in practice. They stripped back features so they could focus on the core. In the end they were left with a great (IMO) project management SaaS. No, it doesn't do everything, but what it does do it does really well.

hcarvalhoalves 14y ago

Concise != Poor I see you're a Physicist, so maybe an analogy can help: would you use the Standard Model to calculate how long your car trip will take? Probably not, Newton's formulas suffice for that, even though those only explain a portion of the world. You certainly understand the concept of elegance. It's the same with UI design. You want to achieve a balance between function and conciseness. When conciseness suffers, it's because you're working at the wrong abstraction level or have weak conceptual integrity, not that your UI is too useful.

claudius 14y ago
No. But I wouldn’t want some UI designer to remove the ‘SM feature’ entirely simply because he thinks that it’s barely used and only clutters the UI. The user has to choose the right tool for the right job and as a designer it is your job to make that decision as easy as possible – surely that includes sometimes not offering a feature, but that will simply force users to choose between your tool and another one if they have a specific job to do. Not to mention that you were probably aiming for something like (special) relativity rather than the SM if you want to compare it to Newton’s formulae. :)
- hcarvalhoalves 14y ago
  > No. But I wouldn’t want some UI designer to remove the ‘SM feature’ entirely simply because he thinks that it’s barely used and only clutters the UI. If the designer is making assumptions based on what "he thinks" and defining function from that, this is simply a poor product and design process. This is designing backwards: the UI is a result of it's function, not the other way around. Given you mention Gnome 3 as an example, I can see where you coming from, and why the trauma about design. Gnome is not necessarily a badly designed product, they just don't cater to your demographics, or most of the Linux user base actually. You might like reading this short book called "Notes on the Synthesis of Form" [1], which is an attempt to formalize the process of design. I highly recommend it, it can be enlightening for people of all disciplines. [1] http://en.wikipedia.org/wiki/Notes_on_the_Synthesis_of_Form http://en.wikipedia.org/wiki/Notes_on_the_Synthesis_of_Form > Not to mention that you were probably aiming for something like (special) relativity rather than the SM if you want to compare it to Newton’s formulae. :) I'm no physicist obviously, but what I was trying to say is: you don't explain the motion of a car thinking about how particles interact, it's the wrong level of abstraction; meanwhile classical mechanics explains this elegantly, even though it doesn't explain everything. There's a balance between conciseness and usefulness, which is the same challenge faced by design, or any other discipline that deals with creating models really.

DanBC 14y ago

I agree. I really hate the trend of removing useful functionality in the aim of, what? I dunno. I'm happy to have that functionality hidden away in text mode config files if needed, to keep the gui nice and clean. That way you just need a check box saying "Turn on advanced features? [ ] Caution: these are not supported and you're on your own".