Tuesday, February 08, 2005
Ever noticed how the most random things spark rants?
I was browsing my news feeds today and I saw a comment in one friend's LJ regarding open-source and usability. Essentially the argument was that on average, open-source usability is rather poor. Because of the nature of the development, OS software tends to be feature-rich*, but for the average user, fewer features is better and easier to understand. Here's why I think that statement is bunk.
Feature-rich is always better. If given the choice between A and B, where A can do X, Y and Z but B can only do X, and everything else is equal, who in their right mind would choose B? The problems arise when A does X, Y and Z, B does only X, but B does X much better than A does. Now we really have a choice, but which has nothing to do with features. The biggest factor limiting OS development as a software movement is documentation, interfaces and attitude.
One, let's face it, developers can't write documentation worth crap, and that's just fine. They're not supposed to. Let the guys with English majors who can explain things worry about documentation. The developer's priority is to make solid code that can be understood and propagated; a totally different skillset. The problem is that the majority of OS work is done mostly by developers, and having nobody else to document the things, end up doing it themselves.
Two, when you're a proprietary software R&D group, with a budget to spend, you can do tons of research into interface design. Microsoft's UI is copied all over the place not only just because it's Microsoft, but because to a degree, what they've done is right. Those millions spent into user studies to see how people think when they sit in front of Excel haven't all just been pissed away. That said, it's possible to get an enormous amount of interface design feedback without spending any money at all. You just need to listen, and that brings me to
Three, the notion that developers know best. They wrote the software, and therefore know everything about how it works, especially things that should not be done. Because of that, when users complain about something, it's easy to dismiss them because they didn't write it and just don't know how to go about things the right way. I've done this myself: "Well, it works fine for me. You just don't know how to do it properly." As long as this attitude persists, OSS will remain less intuitive than proprietary software.
The upside to all of this however, is twofold. One, OSS has nowhere to go but up. For the same reasons that it's feature-rich, it can only get better, and therefore any concerns brought up will eventually disappear given enough time. Second, the statement that OSS is only used by geeks is more important than is immediately apparant. The majority of OSS is used behind the scenes in contexts such as server apps or operating systems, where the user does not even have to be aware of it. This means that the only ones who have to worry about usability in those settings are other geeks who know how it works and all the above concerns matter less.
So to conclude: yes, open-source software has usability concerns, but no, they are not feature-related. What's more, they are rapidly dwindling to gone at best, and temporarily resigned to less-prominent projects at worst.
--------------------------
* The reasons for this basically boil down to the fact that developers can't compete for salary, since most of the software is free. They therefore compete for prestige, or who can put the coolest, bestest stuff in the program. ¶ 11:08 a.m.
I was browsing my news feeds today and I saw a comment in one friend's LJ regarding open-source and usability. Essentially the argument was that on average, open-source usability is rather poor. Because of the nature of the development, OS software tends to be feature-rich*, but for the average user, fewer features is better and easier to understand. Here's why I think that statement is bunk.
Feature-rich is always better. If given the choice between A and B, where A can do X, Y and Z but B can only do X, and everything else is equal, who in their right mind would choose B? The problems arise when A does X, Y and Z, B does only X, but B does X much better than A does. Now we really have a choice, but which has nothing to do with features. The biggest factor limiting OS development as a software movement is documentation, interfaces and attitude.
One, let's face it, developers can't write documentation worth crap, and that's just fine. They're not supposed to. Let the guys with English majors who can explain things worry about documentation. The developer's priority is to make solid code that can be understood and propagated; a totally different skillset. The problem is that the majority of OS work is done mostly by developers, and having nobody else to document the things, end up doing it themselves.
Two, when you're a proprietary software R&D group, with a budget to spend, you can do tons of research into interface design. Microsoft's UI is copied all over the place not only just because it's Microsoft, but because to a degree, what they've done is right. Those millions spent into user studies to see how people think when they sit in front of Excel haven't all just been pissed away. That said, it's possible to get an enormous amount of interface design feedback without spending any money at all. You just need to listen, and that brings me to
Three, the notion that developers know best. They wrote the software, and therefore know everything about how it works, especially things that should not be done. Because of that, when users complain about something, it's easy to dismiss them because they didn't write it and just don't know how to go about things the right way. I've done this myself: "Well, it works fine for me. You just don't know how to do it properly." As long as this attitude persists, OSS will remain less intuitive than proprietary software.
The upside to all of this however, is twofold. One, OSS has nowhere to go but up. For the same reasons that it's feature-rich, it can only get better, and therefore any concerns brought up will eventually disappear given enough time. Second, the statement that OSS is only used by geeks is more important than is immediately apparant. The majority of OSS is used behind the scenes in contexts such as server apps or operating systems, where the user does not even have to be aware of it. This means that the only ones who have to worry about usability in those settings are other geeks who know how it works and all the above concerns matter less.
So to conclude: yes, open-source software has usability concerns, but no, they are not feature-related. What's more, they are rapidly dwindling to gone at best, and temporarily resigned to less-prominent projects at worst.
--------------------------
* The reasons for this basically boil down to the fact that developers can't compete for salary, since most of the software is free. They therefore compete for prestige, or who can put the coolest, bestest stuff in the program. ¶ 11:08 a.m.
