Thoughts on interface
Posted on January 17, 2004
Filed Under /dev/null/ | 42 views |
There are certain conversations that happen cyclically: “cold enough for you?”, “hot enough for you?”, “Apple is going to die”, “Microsoft will be killed by X”, and “A good interface is…”.
It’s the interface one I’m once most interested in at the moment (hell yes it’s cold enough for me, it’s cold enough for all of us right now so stop asking). A lot of vertical space gets spent writing about how this application’s UI sucks and that application’s UI rocks; how this appliance is easy to use, how that appliance is brutally hard. Almost as much space is spent exploring the reasons why this is so (not quite parity since it’s easier to complain than to analyze and hypothesize).
After years of creating software, all of which has had a GUI of some sort, I’ve come to a few minor conclusions about UI design, a list of general rules if you will (some have been borrowed from other people - if I cribbed your idea, well, that follows one of the rules now, doesn’t it?) that direct my efforts when building any sort of software:
- The UI matters. A lot.
- Bad UI is easy to make, good UI is hard.
- Simple is always better. The simpler the better.
- Keep your abstractions as simple as possible, but no simpler.
- If you’re creating a brand-new UI paradigm, you’re probably wrong.
- You are not an artist, you are a developer. Find an artist to do the art.
- The UI will occupy 90% of the development time and provide 10% of the development fun.
- Your UI will occupy 90% of the user’s time and provide 90% of the their fun.
There are lots of good excuses for developing a bad UI. Not coincidentally they’re also the very same excuses that should be used for not yet shipping the software.
(These are the posts that got me thinking about codifying these rules: Direct Manipulation and Drag & Drop and User Interface Hierarchies Rant).
Comments
Leave a Reply