Why Free Software has poor usability, and how to improve it

This article is somewhat misleading, in that of the 15 things listed, only the bottom three are identified as unique to free software.<p>Indeed, some of <i>the</i> most disgustingly unusable stuff I have ever seen has come from companies, not only Microsoft but "enterprise" people, groups that should have a vested interest in software quality.<p>A problem I didn't see listed, is that the people <i>acquiring</i> software may not care. Many of them absolutely don't. I've been in organizations where the people in charge of getting stuff don't have a clue what "features" or misfeatures are actually impacting the usability of the people they are supposed to serve. They don't ask, they don't factor it into the buy/download decision, and no one seems to trace the resulting productivity problems back to them.

&#62;&#62; Coding before design. Software tends to be much more usable if it &#62;&#62; is, at least roughly, designed before the code is written.<p>In my experience, software and web development tends to fail when early work is made with specific UI assumptions in mind. The successful projects I have been a part of have concentrated on putting basic functionality first, and then iterating around the usability design and other features.

I disagree with the premise of the article. On average, Free software seems to have better usability proprietary applications.<p>Furthermore, usability is rapidly improving, since developers seem to be becoming more aware of usability issues in general. Sure, there are examples of horribly unusable apps, but I find the ones that suck the most are the proprietary windows apps, even the ones from big companies, and some of the most usable and consistent apps are open source.<p>Even so, there is always room for improvement, and the biggest issue is simply the lack of developer time to implement all the great usability ideas that pop up.

Rather than creating a large reply, I'd rather put it down to what is the single biggest problem is that, in my eyes, trumps every single thing in his essay: Ugly widgets, icons, fonts and badly thought out standardised user interfaces in both the KDE and, to a lesser extent, GNOME projects.<p>I think it's obvious why - every application inherits from them and their bad design decisions result in most applications having the same usability problems.<p>NB: The author apparently thinks that making me scroll a hell of a lot is good for usability - instead I ended up turning off CSS.

Designers are used to a top-down approach. i.e. They create a design (a full-on design spec with waterfall or a UI improvement/iteration). They like to have design authority. And, honestly, democratic design is generally pretty awful.<p>Devs generally get involved with OSS to make a different and to build something that THEY want to have (not to serve the the average user). Devs tend to be the "managers" of OSS projects.<p>So the big question is: When there is a disagreement about user experience in an open-source project, who wins? Who has authority? Obviously, you shoot for consensus, but it often is hard to achieve. Whem you can't achieve consensus in normal software development, a manager will usually pull rank and say, "Well, it feels like a design decision. Let's err on the side of going with our UX designer and the other people who agree with him/her".<p>I've never done design for an open-source project, but I'd wager that it's pretty democratic with a strong bias towards what the programmers want (which often don't jibe with what the designer wants).