Archives for September 2013

Convention over configuration > One True Way

“Everything should be made as simple as possible, but not simpler.” — Einstein

I have run across an interesting tension coming to a head with mobile product design and choice.

The constraints of mobile brought about this strong push towards simplicity and task centered design. Hell, the bar was raised in every dimension. Many apps that I use today are beautiful both visually as well as functionally.

As with everything, dogma can take things to the extreme, and I have seen this in spades with the iOS7 launch.

So many of the frustrations that I have could be solved with the odd setting or a bit of flexibility.

For example, the new springboard has some nice new features. I like the visual touch when you are downloading a new app, I appreciate being able to get to search from anywhere. However, not being able to use any area to the left of the main screen? Why not? It is so easy to get back to the first page that it would be nice to put something a flick away. It makes me feel like seeing a plot in a great neighborhood that I can’t purchase because a neighbor bought that plot to keep it clear. Except in this case I own the land!

The folders. I like being able to put more apps with them, but it has a harsh trade off:

• I can only see a few at a time since such little space is given to
  the area (on iPad too!)
• I used to place a folder in the bottom right and when opening it I could quickly tap in to open an app quickly (kept one row on purpose). Now that doesn’t work as I am launched into the new popup folder vs inline.
• The animations make me feel like I am leaping from pad to pad in a
  leap pad game.

If I could just change the animation timing (faster please!). If I could change the size of the main area. The size of the app icons. This flexibility would make me so happy…. Just as changing the size of text does (which is great on iOS7).

Let me change the app icons. Give me a central spot to set a default browser so every app doesn’t have to manually do it “Chrome or Safari?”.

People will sneer and talk about Android and how they go too far. The times have changed. Back then this smart phone thing was new and Android looked like crap no matter what. It’s time to move on and deliver a great default experience without locking out the user.

Mobile is personal. We are not all the same. Great we have some colored iPhones now (or you can get a case for your 5S 😉 but more customization options please!

Does your platform just show and tell?

Man, some iOS software development teams were giddy as they learned that old devices were able to install the last known working application.

This is an area of the AppStore experience that has been weak in the Apple eco-system. As a developer, I should be able to give the right version of an application to my user. For that decision, the type of device / version of the OS should be a factor.

This functionality (and others like it) change the way developers can think about new platforms. Without this ability, you can easily feel teased when WWDC comes around and shows you new sexy jewels, some of which have you gnashing your teeth. How many didn’t jump on the autoLayout bandwagon when it was first announced in iOS6 because they had to deal with iOS5 users? (How many didn’t think the tools were baked yet?).

Some are wary though:

“The likelihood of any complex app, especially anything API driven, working after several years of neglect are slim. Those that do work may be incredibly unreliable and buggy”

I don’t understand this. The point isn’t that apps will work forever. If the backend changes then the app may not work and THEN WHAT!

That being said, Apple has been very quiet about the details here. If there is a bug in an application, can the developer get a fix into the older version? How does that all work?

How about planning for that?

I always have a configuration service that my applications speak to and shake hands with when they come online. This allows a nice match of config, A/B testing, and analytics. The three legs of the stool.

Since the application passes up info such as the version of the app, OS, etc, the configuration service can then be smart enough to know if the app “should no longer work”. You can then put in a nag attack:

“Hey, you are really missing out. If you upgrade to v+1 you will get feature X, Y, and Z!”

And if you really get to a point where you can’t keep the application working, then you can use the nuclear option and force the user to upgrade to keep on going. You don’t want to do this to your users as a matter of course. In fact, I have never yet had to use it… but that doesn’t mean you shouldn’t plan for it. It is the kind of core functionality that you shouldn’t punt on, as you can’t go back into the past to fix it.

---

Native platforms such as iOS aren’t the only teasers. If you are beholden to any platforms out of your control (read: client side and more) then you are used to the tease.

The Web has the benefit of not being locked into an app model. You don’t have to wait for the user to update, you can basically force it. However, don’t get too smug. To be a Web developer is to be teased. How often do you see a great new feature in the platform that you can’t use because “I need to freaking support IE8 still :/” Hell.

We have seen amazing progress with the modern browsers which upgrade in short order, but I still wish that we could have gotten our cake and eaten it too by separating the app runtime from the browser shell.

I would give so much to be able to say “I require version ~5.4.x” and be able to map my application to capabilities.

But don’t confuse “simple” with “easy”.

“Wow, how did you lose all of that weight?”

I have heard that a fair amount, and I have already talked about the various issues behind the transformation.

I find something embarrassing about the question. In many ways the solution to the problem was simple. There was no magical insight. Simple, but not easy. I see a difference between those, although they are often thought of as one in the same.

Why is weight loss a simple problem?

I do not believe in the simplicity of “just move more!” or “just eat less!”. That simplicity has been proven to be wrong. I believe that we are a complex organism where the energy balance isn’t static and simple to manipulate. I think that Gary Taubes is onto something, and I can’t wait to see the results of great studies and research to back it all up.

That being said, if you put your mind to learning what works for your body, and you don’t have a crazy metabolic condition that impedes things more than the average bear, then you can see success. I detailed my regime and the changes over time (I am constantly trying things out). I started pretty strict and minimal, and have built up from there.

As a sugar / carb addict it was far from easy to change my diet to become ketogenic, but when that kicked in things fell into place. There is a constant battle with doing the right thing by food consumption, and I wasn’t perfect, but I kept going with a simple plan and went where the results were taking me. Hell, if it was all placebo I will take it. That’s the benefits with being results oriented.

---

Fitness isn’t an arithmetic problem. Neither is it a technology problem. Fitness is a human problem. — Dick Talens

---

I actually agree with much of what Dick Talens says in his post about quantified self “being bullshit”, and I think that the conclusion that this is a human problem is sound, but that doesn’t mean that gathering data is wrong. It is hard to improve if you don’t measure. The key is… what are you measuring. If you are looking at the amount of Nike Fuel then are measuring an input when you want to be watching output. You know… your actual body.

I hear some teeth grinding

But wait, now you lost some weight you want to come out and say “hey, guys, want to lose some weight? easy!”

Not easy, simple. A sugar addiction is as bad or worse as many other hard drugs. Food is also something you need to consume, and every day there is an easy exception.

To really cheese off the female population, giving birth is another example of simple but not easy. As I watch my wife reach month number 10 I can see that there is no stopping things. That baby is coming out one way or another, and as long as their aren’t complications, she has some hard (but simple) work ahead of her. Fostering and bringing up the child? Now that is some hard complex work. How am I going to measure if I am doing a good job as a father?

Finding the perfect mate is also complex. There are tactics for finding someone… meet as many people as you can, etc. However, how do you know if they are the one? The usual response: “you will know it when you see it”. Yowser, that is hard to measure.

If I can be the opposite of romantic for a second: In a world populated with billions of people it is hard to swallow the idea that you have a chance of meeting the one most ideal match. Our DNA is built more for finding someone to proliferate genes. That is what we are wired to do. If we were wired to find some magical ideal our species would be long gone.

There are so many truly complicated and complex problems out there to solve, so maybe it makes sense to start the engines by going after the simple ones first.

This ties into the Checklist Manifesto. Sure we could debate quality vs. quantity but if you just do and complete tasks, you will make progress. This is now my starting point across a number of focus areas. Once I have the habit going I can then work on optimizations.

I have found that completing things and getting the flywheel going, has a profound effect on my happiness and growth. One of the best side effects of becoming hopefully healthier and living a fuller life through weight loss is that I now see that I can complete things.

Now, what to tackle next.