Sunday, January 02, 2011

Resolution 2011: Managing complexity

I'm good with resolutions. Mostly because I know how to pick 'em. I make 'em doable.

Consider sleep. I like to exercise, but in my life sleep is more important. So I've resolved to sleep at least 52 hours a week [1]. I think I can do that if I track the numbers and identify where I fall short.

That's one for 2011. The other resolution is about managing technological complexity.

I've been on a complexity reduction kick for a few years , but this year my focus is on technological complexity. I'm starting with the plausible assumption that we all have a personal "complexity budget". Some of us can manage more complexity, others less, but we all have memory and processor limits -- even the AIs among us.

We can spend our complexity capacity figuring out how to adjust product development to available capacity, or we can spend it figuring out what parts of SharePoint are worth using [2]. Both tasks will be equally draining.

At some points in my life I had complexity capacity to spare - perhaps because I wasn't using it wisely. That's not true now.  Gains from improved productivity techniques [3] and growth of mind [4] are offset by entropic neurons. Most of all though, my life overflows. I'm not complaining -- it's an overflow of good stuff. It means though, that I need to use my complexity capacity wisely.  I can't be spending limited firepower figuring out which of my 15 Google identities is running a feedburner bot linked to a pseudonymous twitter account.

It's not easy to reduce technological complexity. It often means making do with less; giving up tools and solutions I really like. Often it means declining new incrementally better improvements -- because a 10% gain isn't worth the learning curve and failure risk. Sometimes it means giving up on old tools that still work but are increasingly unsupported. Yes, it's a lot like software portfolio management.

Looking at how technological complexity grows in my life I can identify four broad causes....

  1. Taking on too many simultaneous tools and solution sets.
  2. Failure to clean up. Ex: Abandoned user identities, google accounts, etc. Creates noise and clutter.
  3. Premature adoption of technologies and solutions. Ex: Any new OS X release, any new Apple hardware, trying to get Contact synchronization to work with both Google Contacts, OS X Address Book and iPhone, OS Spaces. Above all - Google Wave.
  4. Prolonged use of increasingly unsupported solutions in a world of forced software evolution [6]. Ex: document-centric web tools, wristwatches, printers.

I've gotten better at the first one, but the next three all need work. The 3rd and 4th are particularly tricky. My heavy use of Google's multi-calendar sync solutions is clearly premature [5], but it's been very valuable and relatively bug free. On the other hand, I think my jury-rigged Contact integration solution may be a bridge too far. On the other hand, I stuck with Outlook's Notes feature long after it was clearly dead.

Cleaning up is the least interesting measure, but one of most important. There are 1,575 entries in my credentials database, extending back to August 1995. Sure, most of those sites are long gone, but I still have too many active identities and credentials. I need to gradually cull a few hundred.

This project should keep me busy for a while. It will, of course, suck processors in the short term, but I expect near term returns and long term gains. Feels like a good resolution target.

-- fn --

[1] It helps that recent research suggests that amyloid clearance occurs primarily during sleep, and I'm speculating that a 10-20% decline in amyloid clearance translates to 10 extra years of dementia.
[2] The wiki and, in the absence of alternatives, the document store. Don't touch the rest, even the stuff that seems to work is poison. 
[3] At my stage "GTD" is child's play.  I use a mongrel of Agile development planning methodology, GTD/Franklin, and a pocketful of tricks including calendar integration across family and work.
[4] For quite some time mind can grow even as brain more or less sucks wind. Not forever, but for a time. 
[5] The UI for configuration multiple calendars has been bizarrely obscure for about two years. This is not mainstream.
[6] It's predator-prey stuff. Software evolution was much more leisurely before human-on-human predation took off with hacks, frauds, identity theft, malware and the like. Now old bucks have to keep moving, or we become wolf chow. Software cycles are faster, products die quickly, and we have to keep buying whizzier hardware. If not for malware, the curated world of iOS would still be years away.

See also:

No comments: