Month: May 2017

The sum of all fears

Update 5/25/2017: This is a post I started over a year ago.  In the interim Ubuntu has officially dropped the plan on a convergent desktop.  Mark Shuttleworth might argue that convergence will eventually happen but ultimately that doesn’t matter.

“In business being early, or being late, is the same thing as being wrong.”

Outstanding article over at TechRepublic discussing the lack of momentum that Ubuntu has had as of the last couple years.  The basic rundown is that the author believes that the long term goal of “the convergent desktop” is causing other less important goals to slip.

For those who haven’t heard of the convergent desktop (or simply convergence) it is the idea currently being chased by both Microsoft and Canonical (the company behind Ubuntu) whereby your phone/tablet can also be your desktop/workstation.  Sometimes this is associated with a seamless user experience that “transcends” both use cases (i.e. is the same environment on both platforms) but more often is based on some kind of modal shift when device size changes.  So for Windows 8, it become more Windows 8’y when on a phone, but feels a little more like Windows 7 when on a 22″ monitor with mouse.

Google is, of course, more concerned with turning everything into an extension of the web via Chrome and/or Android.  This means that they ultimately don’t care if it is a desktop or a phone running applications; as long as the data is stored in their cloud or provided by one of their services.  So what is Apple’s strategy concerning convergence?  Ahhh, now you get to the meat of the problem.

Apple, always laser focused on user experience, figured out a while ago that convergence SUCKS.  It really really does and here is a brief explanation why.

A great desktop experience is going to be focused on use cases where people are going to use desktop applications.  I call these users “creators” because they primarily use their computers for creative endeavors.  Think software development, editing photos, writing books, mixing music,  making spreadsheets, etc.

In this vein, the tools for creating are centered around the ability to produce new material.  Keyboards are spectacular input devices for creators.  I can type faster than I can write. Even though my ultra-book has a touchscreen, I never use it because my ten fingers are faster for creating things that a single pointing finger is.  When a fine grain control inside a two dimensional canvas is needed, a mouse is significantly better than either a touch screen or a touch pad.

A great tablet experience is going to be focused on use cases where people are not going to be creating.  I call these users “consumers”.  When using my tablet I am almost solely relegated to the role of consuming information.  Reading emails, watching Netflix, looking up receipts on Google, etc.  Consuming requires less functionality than production and added interface utilities for these edge usage cases would just take away from the user experience.

Now obviously most users spend some time during the day being both a consumer and a creator.  This is not a statement of the value of how a user uses their technology but an implicit realization that different use cases should be centered around how best to actually use their system.

It is hard to make a really functional sports car that can also be a useful pickup truck.  Trying to make one into the other generally causes you to have a tool that is good at neither.

The Measure of Who We Are

The older I get the more value I place on having timely and frequent feedback. It sounds like an insignificant thing but if you want what your doing to actually be useful it is paramount.

Never underestimate the power of an evolutionary process with a tight feedback mechanism.  –Linus Torvalds

Frequent feedback is so important because of the abundance of bad ideas that actually mask themselves as initially useful. Anyone who has done software development or design has had 100 people say “I have this million dollar idea I need you to implement.”  The reality is that most ideas are bad and seldom deal with the practicalities of reality.  Often their shortcomings are not obvious.

Good ideas (aka theories) are actually pretty uncommon.  We build these perfect structures in our heads and then begin to wonder why our dreams cannot also be truth.  Once we latch onto a theory we will remain “loyal” to it and will seldom surrender it easily until external facts force us re-evaluate them.  In actuality these ideas are not even theories but beliefs, which is what a theory really is until it has been tested.  A feedback loop is really a way to force us to test something with an external reference we cannot ignore.

The single biggest problem in communication is the illusion that it has taken place. –George Bernard Shaw

One of the major changes in the structure of technology companies and their succeed has been the dissociation between a business BEING an idea, and a business trying an idea.  The new mantra in places like Silicon Valley is get actionable feedback, make a change, fail fast, and pivot.  Failing fast provides as much feedback as possible in the shortest period of time; so fewer resources are spent on bad ideas and thus increasing the likelihood of isolating good ideas.

On the macro level, the key to the effectiveness of capitalism has been the success of the market as a feedback mechanism.  Local government’s superior representation is a direct reflection of the ease at which local officials can be replaced.  The feedback loop for a City Mayor is quite a bit tighter than for the President of the United States.

Notice how, in the last example, by focusing on the feedback mechanism we can evaluate two systems that are superficially identical but function dramatically different.  Why does progressive democracy work so well for Denmark and not for the US?  Because Denmark’s feedback loop is closer to Missouri’s than it is to the whole of the United States.

With genetic engineering, we will be able to increase the complexity of our DNA, and improve the human race. But it will be a slow process, because one will have to wait about 18 years to see the effect of changes to the genetic code. –Stephen Hawking

Feedback is so fundamental that systems that have weak or non-existent evaluation mechanisms are losing their authority overall.  We value academic disciplines by how consistent their feedback loops are.  Things like math, science, and engineering all have well defined and extensively tested methods to evaluating themselves and make corrections when shortcomings are identified.   Non-empirical disciplines like art, philosophy, and religion are suffering from the lack of reference available to the natural sciences, and their overall effectiveness is thus reflective.  Remember, at one time science and mysticism were one and the same until science formalized the scientific method, liberating itself (and humanity) from the confines of dogma.

“Science gives man ever greater powers but less significance. It gives him better tools and with less purposes. It is silent on origins, values, and ultimate aims.  It gives life and history no meaning or worth that is not canceled by time and death.”. –Will Durant

This is not to say that ideologies and theories are likely to disappear. On the contrary, as religion had demonstrated, systems that have no effective feedback loop are nearly impossible to remove entirely because there is no way to “prove” the shortcomings of their beliefs.  You cannot fail a test you’ve never taken.

Nor do I mean to suggest that systems lacking structured feedback methodologies are bad.  I strongly believe in the value of philosophy, art, and religion as part of making a full life; and I passionately love my liberal arts education.  The need for improvement isn’t an absolute and, contrary to popular belief, neither is the need for truth.  But, when the desire is to approach understanding, even if only asymptotically, there is simply no better system we know of then a quality feedback loop.