In Open Source, Learn to Decide | Hacker News Reader

In Open Source, Learn to Decide

nacin.com

36 pointsptbello15 comments

That screen is exactly the reason why I personally prefer Android and open source. I know what those options mean, and I appreciate the opportunity to choose.

This really is a situation where you can "have your cake and eat it too". For a good user experience you need:

1) sane defaults 2) a minimal, apple-like settings screen 3) an "advanced" settings screen

The important thing about #3 is that you must make it clear that most users do not need to access that screen. A button on the main settings screen call "advanced options" does not necessarily make that clear -- there are still people that will fiddle with it even though they don't need to. My suggestion would be to make it minimally hidden, such as Firefox's about:config.

The funny thing is that the screenshot (http://farm8.staticflickr.com/7160/6532001299_e08f052149_b.j...) on which this post is based is actually a request of making and unavoidable decision (how the phone should present itself to the computer plugged via USB), and the "clutter" is a clear description what both options mean.

There's several problems with that screen. I'm not a professional, and there should definitely be user testing, but these are my observations.

- There's way too much text. Users don't like reading. http://www.codinghorror.com/blog/2009/10/treating-user-myopi...

- Nobody who isn't a nerd will know what MTP and PTP stand for.

- This screen could be broken down, roughly, into three large buttons for Windows, Mac, and Linux with no extra text. Windows needs no explanatory text or further configuration. The screen displayed after the Mac button is touched could instruct the user to download File Transfer for Mac. The screen after the Linux button could contain instructions for whatever it is Linux users do to connect via PTP. Heck, you could even show Linux users the current screen; they probably won't care.

> This screen could be broken down, roughly, into three large buttons for Windows, Mac, and Linux with no extra text.

Exactly. This turns it from an option that the user has to choose to an option where they already know the answer.

But those Win/Mac/Linux buttons would be simply a wrong solution -- the idea is to peak a standard that will work with a certain application, not OS. You just can't make it simpler without sacrificing functionality.

Clueless user will of course be left only with an option just to try both, but IMO this glitch is way better than macky "I won't work, but this is their fault".

But surely you don't need ~40 words to communicate the difference between two options.

Your three choices (Windows, Mac, Linux) won't work. If you want to go that route, the question would be "what application do you use to transfer your files"? You can see what a rat's nest that can turn into. Maybe you can split it into 3 categories: file manager, photo manager, music/media manager and then have further instructions on a second screen.

That sounds like a reasonable approach.

I WANT THAT FUCKING CHOICE. Don't you DARE take away my fucking choice. I WILL CUT YOU! rages

[I exaggerate here, but I think my point is clear. Not every user thinks choices are a bug. I will switch away from any platform with less choice to any platform with more choice, every single time. And I am not alone.]

So how is Linux From Scratch working for you?

Good choices are accompanied by sensible defaults. LFS lacks any defaults at all...

(I am in fact currently moving down the "user-friendliness scale" of linux distros, but I haven't hit the bottom yet. Right now I'm ditching Ubuntu for Debian, to escape the ongoing friendlification.)

The thing that always brings this home to me is how much harder it is (for a developer) to rethink a design to remove the need for an option than to not bother and defer the decision to the user.

It's sometimes instinctual in these situations to be lazy, but a small amount of reflection (try having a conversation with an imaginary Steve Jobs) makes it obvious that the time and energy cost for you will be easily justified by the time saving for a non trivial amount of end users.

I think the decision only has to be made at the top level, the graphical UI.

One of the things I like about Mac OS X is how solid the command-line "defaults" mechanism is. It has allowed me to expose everything that might be useful to configure, while still trimming the graphical preferences UI and setting a series of factory defaults. So an app can be both simple and flexible at the same time.

Relevant reading is Barry Schwartz's The Paradox of Choice, which makes this insight as obvious as it should be.

Open source segments need the design considerations - Even with its size and limits with scope.

An important read but not very relevant to open source whitneyhess.com/blog/2011/04/23/youre-not-a-user-experience-designer-if/

The user perspective is the most important gain for any designer/team. Build a team to challenge each other in user perception and details. Settings and preferences are the most valid arguements for such choices. Clutter is unacceptable.

Iterate on much focused segment of your problem.

Reply on news.ycombinator.com