3 ms·
I come to coding from a Graphic Design background and I wish I had computer science under my belt as I test the waters of programming, but what I'm learning rea
by err4nt 12y ago
I come to coding from a Graphic Design background and I wish I had computer science under my belt as I test the waters of programming, but what I'm learning reading HN is that many programmers could benefit from the thinking they taught us in design school.
What's the difference between an 'artist' and a 'designer'? One is employed! When you are designing something (even with code) your measure of success is how well you deliver what needs to be built. With 'art' there is no such requirement—art can become whatever your whims desire.
One thing they warned us very strongly about in Design school was never to become 'married' to our designs. If a client pays you to make something and doesn't like it it's NOT good design because it didn't meet their needs. Don't be personally offended, it's not art, it's design. You can do art when you're paying yourself for your time.
I write code and add things like subtle text-shadows and gradients to enhance legibility (scientific reason) to make my designs more accessible, but if I look at the commit logs I often find that my supervisor goes back and removes all text-shadows, box-shadows, and gradients just because he believes gradients and drop shadows are a passé design trend from Web 2.0 that he probably read in an article. I could get offended that he's essentially neutering the accessibility of the designs I'm making and nobody will ever see the full design as I intended it - but I got paid for writing it just the same so what do I care?! He has the website he wants and I have employment. Bottom line is I'll keep delivering value whether he wants it or not and once that handoff is made it's his to do with as he pleases.
- _dps 12y ago> subtle text-shadows and gradients to enhance legibility (scientific reason) Honest question: can you provide citations for this claim? I have been trying to produce a compendium of all the empirically-supported claims about visual design affecting usability. I have yet to find citable claims for these cases, and would greatly appreciate a leg up in the effort :)