Kill the settings, build opinionated software

<p><pre><code> Some people argue software should be agnostic. They say it's arrogant for developers to limit features or ignore feature requests. They say software should always be as flexible as possible. We think that's bullshit. The best software has a vision. The best software takes sides. When someone uses software, they're not just looking for features, they're looking for an approach. They're looking for a vision. Decide what your vision is and run with it. </code></pre> Well, I, in turn, think that's bullshit. IMO, the best software conforms to the way the user works, rather than trying to force the user to work in a way that's unnatural to them.<p>The arguments against a confusing preferences pane and the arguments for sensible defaults in no way contradict the notion of making software than is flexible enough for power users. Poor UI design (3 different places to set prefs?) is orthogonal to the issue of whether or not something should be configurable.<p>Power Users don't matter? Maybe for certain classes of consumer applications that don't face a steep barrier to adoption in the first place. But for some apps, it's the power users who are the early adopters, who first begin using and evangelizing for the software and without whom it would never catch on in the first place. Do you really want to piss those folks off?<p>Make the damn "advanced preferences" pane more or less hidden (ala, about:config) but document the shit out of it and make sure it's at least there for the people who <i>do</i> care.

Here's a little anecdote: Years and years ago, in a moment of free time, I added a bunch of silly settings to my software (EasyBCD, now with 7-10MM+ users). It had almost no users then, and I just added these options to "fill up" the empty preferences dialog. A year or two later, I removed a couple, and the complaints could be heard around the world.<p>The features, you ask? They really are ridiculous: Save the window size on exit. Set the default size of the UI text. Set the default font of the UI text. Open in "advanced mode" by default. Open to a particular page.<p>None of these features really saves a user more than a click. Some of them are purely aesthetic. But their popularity - which I greatly underestimated - was incredible. Till now, people file bugs when their favorite "silly preference" ends up resulting in odd behavior. And they <i>always</i> catch on immediately when I change something there.<p>Sure, it's hard work (read: boring, time-consuming, donkey work) to maintain these options. But I think they're part of the reason EasyBCD has succeeded.<p>(Note: EasyBCD is a software most people install once and never use again. It's definitely not something you would use on a daily basis. But people LOVE power and control, even if it's just an illusion.)

This article is little more than a generic rehash of popular memes. Let's break it down:<p>- 37Signals is great, they do simple! - anecdote - Apple is great, they do simple!<p>I hope whatever software this guy does has less opinion and more substance.

I don't think it's black and white; you can have an opinionated, simplified front end, and a back end that is wide open. The back-end makes some things easier, such as testing, and helping admins deploy to 100+ users.<p>A flexible back-end can even be used to create an opinionated front-end. Consider a "factory defaults" file; preset any behavior, whether or not it has a checkbox. Try every combination, and convince yourself that one is better. Even with these decisions made up front, the interface is there to let other people disagree with you.<p>Mac OS X has a nice balance, because the built-in "defaults" command line program can tweak any setting that would also be available in the application's code. Just because a program is a GUI, doesn't mean that it should rely exclusively on GUI elements for configuration.

At the farthest reaches of powerful software, you will always end up with a programming language and libraries.<p>At the other extreme, you have one thing that does exactly one task and as such can have a nearly "invisible" interface.<p>My suspicion is that it's actually kind of rare to need the areas in between - that apps like spreadsheets and photo editing are unusual, and if you _suspect_ that you have an app that could be developed in such a broad, deep, configurable way, your best bet is to start building in a library form and construct prototype one-shots - possibly even making them your MVP - attacking various different problems to see if you're right, before you try to make a unified monster UI for the perceived need.

What if my opinion is that software should have <i>more</i> settings? I am very opinionated about that.

Probably explains why Basecamp didn't click for us.

To put that in a less confrontational way: instead of designing features, and then providing options to maximize your audience, first pick an audience, then make software that will work perfectly for them without the need for options. Result: a smaller, more dedicated audience, versus an audience of "everyone" and a product that is as well-suited to an individual task as Microsoft Office.

I always cringe when I hear "don't listen to your customers." That is absolutely the wrong behavior. If you don't listen to the people who are going to buy your product, you won't build a product that solves a real problem.<p>I listen to my stakeholders (potential users, potential buyers, current users, etc), but I don't necessarily do what they say. Most people try to give you a solution to their problem, you need to dig through their "solution" to find their actual "problem", then design the best solution (and like all design, it is best not when nothing else can be added to it, but rather when nothing else can be taken away).<p>Oh, and I completely agree that software should be opinionated. People are paying you to solve their problem, don't get lazy and push responsibility back on them!

Minor annoyance, but what's the bullshit sideshot against Android multitasking? Are we to believe now that only Apple get's multitasking right? "Apple invented multitasking"? It works fine on my N1. I think Steve didn't even say what makes multitasking on the iPhone so special.<p>That's how Apple marketing works, and frankly, it makes me dislike them despite of the neat design of their products. If their products are so great, why do they even need to work that kind of marketing?

Apple doesn't listen to users and they're the most successful technology company in the world. They have a fearless leader who's not afraid to piss people off... Just like Zuckerberg.... Except we all complain about him. I suppose he needed the cult like following before you start to piss people off. That is the key.