Haven’t heard a lot from Russell Beattie in a while, but then saw this piece on Hacker News and had to post a link to it.
Of course, there’s more to this issue than meets the eye; are all your users going to pay attention to fluff and not to substance? Are all your users not going to go straight to under the hood to make sure your data isn’t corrupting on the back-end? Are all your users going to be satisfied by a surface-level treatment of the project?
Take an app like OpenClinica, for example, where data drives the application from beginning to end, and moreover, the data is all clinical research information where looking at a patient’s information has to be protected by user ID, user role, and so on.
When we started development back in 2002 and showed it to the first client, I’ll be honest; it looked like crap. But, our client was used to collecting information on paper and entering it into MS Access, which, let’s face it, also looked like crap in 2001-2002. The process before OpenClinica was as design-y and flashy as an Excel ’98 spreadsheet, because basically, that’s all that it was.
Since then we’ve come a long way. It looks better, and works better too :). It didn’t in the beginning, but design would have gotten in the way of the data import, user requirements, and weekly reviews. Design would have eaten away at our budget, regrettably, but after we started engaging the OSS community at large around the time of our first release, the design (finally) caught up to the application.