You can spot the less useful design principles after a while. They tend to be wishy-washy; more like empty aspirational exhortations than genuinely useful guidelines for alignment. I’ve written about what makes for good design principles before. Matthew Ström also asked—and answered—What makes a good design principle?
- Good design principles are memorable.
- Good design principles help you say no.
- Good design principles aren’t truisms.
- Good design principles are applicable.
I like those. They’re like design principles for design principles.
One set of design principles that I’ve included in my collection is from gov.uk: government design principles. I think they’re very well thought-through (although I’m always suspicious when I see a nice even number like 10 for the amount of items in the list). There’s a great line in design principle number two—Do less:
Government should only do what only government can do.
This wasn’t a theoretical issue. The multiple departmental websites that preceded gov.uk were notorious for having too much irrelevant content—content that was readily available elsewhere. It was downright wasteful to duplicate that content on a government site. It wasn’t appropriate.
I think that the design principle from GDS could be abstracted into a general technology principle:
Any particular technology should only do what only that particular technology can do.
Choose the least powerful language suitable for a given purpose.
Or, as Derek put it:
In the web front-end stack — HTML, CSS, JS, and ARIA — if you can solve a problem with a simpler solution lower in the stack, you should. It’s less fragile, more foolproof, and just works.
ARIA should only do what only ARIA can do.
CSS should only do what only CSS can do.
HTML should only do what only HTML can do.
This was originally posted on my own site.