Stopping Accidental Technical Debt

Buster Keaton looking at a poorly constructed house

In one of my earlier posts about technical debt, I differentiated between intentional debt (that taken on deliberately and purposefully) and accidental debt (that which just accrues over time without rhyme or reason or record). Dealing with (in the sense of evaluating, tracking, and resolving it) technical debt is obviously a consideration for someone in an application architect role. While someone in that role absolutely should be aware of the intentional debt, is there a way to be more attuned to the accidental debt as well?

Last summer, I published a post titled “Distance…is the one true enemy…”. The post started with a group of tweets from Gregory Brown talking about the corrosive effects of distance on software development (distance between compile and run, between failure and correction, between development and feedback, etc.). I then extended the concept to management, talking about how distance between sense-maker and decision-maker could negatively affect the quality of the decisions being made.

There’s also a distance that neither Greg nor I covered at the time, design distance. Design distance is the distance between the design and the outcome. Reducing design distance makes it easier to keep a handle on the accidental debt as well as the intentional.

Distance between the architectural decisions and the implementation can introduce technical debt. This distance can come from remote decision-makers, architecture pigeons who swoop in, deposit their “wisdom”, and then fly away home. It can come from failing to communicate the design considerations effectively across the entire team. It can also come from failing to monitor the system as it evolves. The design and the implementation need to be in alignment. Even more so, the design and the implementation need to align with particular problems to be solved/jobs to be done. Otherwise, the result may look like this:

Distance between development of the system and keeping the system running can introduce technical debt as well. The platform a system runs on is a vital part of the system, as critical as the code it supports. As with the code, the design, implementation, and context all need to be kept in alignment.

Alignment of design, implementation, and context can only be maintained by on-going architectural assessment. Stefan Dreverman’s “Using Philosophy in IT architecture” identified four questions to be asked as part of an assessment:

  1. “What is my purpose?”
  2. “What am I composed of?”
  3. “What’s in my environment?”
  4. “What do I communicate?”

These questions are applicable not only to the beginning of a system, but throughout its life-cycle. Failing to re-evaluate the architecture as a whole as the system evolves can lead to inconsistencies as design distance grows. We can get so busy dealing with the present that we create a future of pain:

At first glance, this approach might seem to be expensive, but rewriting legacy systems is expensive as well (assuming the rewrite would be successful, which is a tenuous assumption). Building applications with a one-and-done mindset is effectively building a legacy system.

The Seventy Million Dollar Question



Just when I thought I was done posting for the week, they suck me back in.

Juicero started lighting up my Twitter feed a little while ago. For those, like me, who have no earthly idea what Juicero is, it’s a startup that makes an “Internet-connected kitchen appliance”:

Juicero’s flagship product is a $699 countertop device that cold presses juice out of “packs” of already prepped fruit and veggies. The packs — reminiscent of the cups and pouches used in single-cup coffee brewers from Keurig, Flavia or Nespresso — cost $4 to $10 each and are available through a Juicero subscription, but not in groceries.

Juicero just picked up $70 million in Series B funding. ‘Cause digital.

There is just one hitch – you don’t actually need the high-dollar hardware to make the juice:

But after the product hit the market, some investors were surprised to discover a much cheaper alternative: You can squeeze the Juicero bags with your bare hands. Two backers said the final device was bulkier than what was originally pitched and that they were puzzled to find that customers could achieve similar results without it. Bloomberg performed its own press test, pitting a Juicero machine against a reporter’s grip. The experiment found that squeezing the bag yields nearly the same amount of juice just as quickly—and in some cases, faster—than using the device.

Fortunately (ahem), Juicero only sells the bags, at anywhere from $5 to $8 apiece, to owners of the hardware. Performance between the device and non-standard hardware (your bare hands) is mixed:

In Bloomberg’s squeeze tests, hands did the job quicker, but the device was slightly more thorough. Reporters were able to wring 7.5 ounces of juice in a minute and a half. The machine yielded 8 ounces in about two minutes.

Almost 5 years ago, I wrote a post titled “The Most Important Question”. In it I stated the following:

The most important question in architecture is “why”. When questioned about any aspect of the design, if you cannot justify the decision, you should revisit it. Being able to list outcomes, both good and bad, and explain the reasoning behind your choices will garner trust from both customers and colleagues. Most importantly, it should boost your own confidence in the design.

With a little editing:

The most important question in architecture innovation is “why”. When questioned about any aspect of the design product, if you cannot justify the decision, you should revisit it. Being able to list outcomes, both good and bad, and explain the reasoning behind your choices will garner trust from both customers and colleagues. Most importantly, it should boost your own confidence in the design product.

Asking “why” could have probably saved some people a lot of money, just sayin’.

Pride, Prejudice, and Professionalism in the Business of IT

interior of a 1958 Plymouth Savoy

Twenty-plus years in IT have led me to believe that there are very few absolutes when it comes to software systems. Two that do seem to hold true are these:

  1. Creating systems is esteemed far more highly than maintaining systems.
  2. Systems that are not maintained, will decay.

There are a variety of reasons for this situation, many of which are baked into the architecture of the enterprise. Regardless of the why, however, the two facts remain. Without a response to those issues, entropy is inevitable.

Over the past few days, I’ve seen several blog posts by two different authors dealing with this situation in two different ways:

Jason complains about the non-technical “leadership class” in his first post:

And hence we get someone making the big decisions about healthcare who knows nothing about medicine or about running hospitals or ambulance services. And we get someone in charge of all the schools who knows nothing about teaching or running a school. And we get someone in charge of a major software company whose last job was being in charge of a soft drinks company. And so on.

Again, this is fine, if they leave the technical decisions to the technical experts. And that’s where it all falls down, of course. They don’t.

The guy in charge of the NHS insists on telling doctors and nurses how they should do their jobs. The woman in charge of UK schools insists on overriding the expertise of teachers. The guy in charge of a major software company refuses to listen to the engineers about the need for automated testing. And so on.

This is the Dunning-Kruger effect writ large. CEOs and government ministers are brimming with the overconfidence of someone who doesn’t know that they don’t know.

In his second post he follows up with how to respond:

My pithy – but entirely serious – advice in that situation is Do It AnywayTM.

There are, of course, obligations implied when we Do It AnywayTM. What we’re doing must be in the best interests of stakeholders. Do It AnywayTM is not a Get Out Jail Free card for any decision you might want to justify. We are making informed decisions on their behalf. And then doing what needs to be done. Y’know. Like professionals.

I disagree. Strenuously.

If you go to the doctor and they tells you that you will need surgery at some point for some condition, would you expect to be forcibly admitted and operated on immediately?

If you were charged with a crime, would you expect your attorney to accept a plea bargain on your behalf without consultation or prior permission?

If neither of those professionals would usurp the right of their client/patient to make their own informed decision, why should we? Both of those examples would be considered malpractice and the first would be criminal assault in addition. Therefore, I disagree that acting on someone’s behalf without their knowledge or consent is a viable option.

John’s approach, rejecting helplessness and confronting the issues by communicating the costs (with justifications/evidence) is, in my opinion, the truly professional approach. We have a responsibility to make the problem visible and continue making it visible. We also have a responsibility to operate within the limits we’re given. We know far more about our area than someone higher up the management chain, but, that does not equate to knowing more in general than those higher up the management chain. Ignorance is relative. Micro-managing, getting deeper in the weeds than you need to is ineffective. If, however, you’re in the weeds, do you have the information necessary to say that the issue being “interfered with” is one without higher-level consequences? Dunning-Kruger can cut a wide swathe. Trust needs to cut both ways.

Imagine riding as a passenger in a car. You see the car drifting closer and closer to the shoulder. Do you point it out to the driver or do you just grab the wheel? You might prevent an accident or you just might cause one by steering into a vehicle coming up from behind that you didn’t see from your vantage point.

[Plymouth Savoy photo by Christopher Ziemnowicz via Wikimedia Commons]

Form Follows Function on SPaMCast 438


Once again, I’m making an appearance on Tom Cagley’s Software Process and Measurement (SPaMCast) podcast.

This week’s episode, number 438, features Tom’s essay on using sizing for software testing, Kim Pries with a Software Sensei column (canned solutions), and a Form Follows Function installment based on my post “Organizations as Systems and Innovation”.

In this episode, Tom and I discuss how systems must fit into their context and ecosystem, otherwise it can be like dropping a high-performance sports car engine into a VW Beetle. Disney-physics may work in the movies, but it’s unlikely to be successful in the real world. If all the parts don’t fit together, friction ensues.

You can find all my SPaMCast episodes using under the SPaMCast Appearances category on this blog. Enjoy!

Disruptive Decency

Well, this turned out to be very much a different post than what I’d first thought.

Last Thursday, CIO published an article titled “Your Pebble smartwatch will live on when Pebble’s servers shut down” that had good news for owners of the Pebble smartwatch:

But now that Pebble has been acquired by Fitbit and is presumably nearing the end of its life, Pebble users fretted that their watches would cease to work once Pebble dies. That’s not the case.

Pebble just rolled out an iOS and Android update that frees its watches from cloud-based online servers. That means when Pebble goes offline, your watch will still work.

Coincidentally, this was one year to the day since I posted “Google’s Parent Company is Stirring Up a Hornet’s Nest”, which talked about Nest’s decision to brick the Revolv home automation hub rather than continue to support them. Fitbit’s decision was a refreshing departure from the attitude demonstrated by Nest (and lampooned by xkcd above). The punchline was going to be: basic human decency seems to be a disruptive tactic these days.

And then I launched Twitter Monday morning:

By this point, I would assume Sunday night’s, incident needs no explanation on my part. Details are still coming out, but regardless of what develops, United Airlines is the loser in this scenario. There’s an old saying is that there’s no such thing as bad publicity.

The old saying is wrong:

The tweet above has plenty of company in the twitterverse, none of it flattering to United or beneficial to its share price. Tweets like this haven’t helped:

The perception that sticks is that an older man, a doctor, was violently removed from a plane in order to allow United to get four of its flight crew to Louisville and United’s CEO is upset about having to “…re-accommodate these customers” (not exactly what’s said, but certainly what will be taken away from that garbled message). Additionally, the poor job done on that earlier message completely undercuts the perceived sincerity of the latter one:

Given United’s past problems with customer service, one might expect more effort would have been spent to prevent incidents like this and they would have been better prepared for dealing with the aftermath of something that did go badly.

Wrong on both counts.

An excerpt from a recent interview of Oscar Munoz (United’s CEO) on Business Insider makes the situation all the more egregious:

Here, in Chicago, it’s miserable because if you don’t leave by a certain time, you are just dead. “I’m going to get there and there are going to be a billion people and the damn TSA line.” By the time you get to sit on one of our seats you are just pissed at the world.

So how do we make all of that a little bit easier? This is the thing. You’ve got that broad issue of anger and anti-industry noise. We’ve lost the trust and respect of the broader public, and so every action we take, they don’t particularly like, they see it negatively. We have to work on that broad communication. I am going to do it at this airline and allow myself to differentiate in the flight-friendly mode so that people don’t immediately have that visceral reaction.

Dragging people off a flight (literally) probably doesn’t fit into the mold of a “flight-friendly mode”.

So I will return to my original punch line: basic human decency seems to be a disruptive tactic these days.

You can’t always get what you want…


You can’t always get what you want
But if you try sometimes well you just might find
You get what you need

When it comes to systems, you can’t always get what you want, but you do get what you design (intentionally or not), whether it’s what you need or not.

In other words, the architecture of systems, both social and software, evolve through some combination of intentional design and accidental emergence. Regardless of which end of the continuum the system leans toward, the end result will reflect the decisions made (or not made) in relation to various stimuli. Regarding businesses (a social system), Ruth Malan, in her February 2012 “Trace in Sand” post, put it this way:

I have been talking about agility in terms of evolutionary ecology, but with the explicit recognition that companies, comprised after all of individuals, attempt to speed and alter and intervene and interject and intercede and (I’m looking for the right word here) shape evolutionary processes with intentional actions — concerted, but also emergent from more and less choreographed, actions and intentions. Being bumped along by the unpredictable interactions of others, some from within, but also from “invasive species” from other ecosystems looking for new applications for their adaptable, mutable capability set.

Organizations create and participate in business ecologies. They build up the relationships that stabilize parts of the broader ecosystem, and create conditions for organizational forms to thrive there. They create products, they create the seeds of the next generation of harvest. They produce variants on their family tree, to target and develop niches.

Ruth further notes that while business adapt and improve in some cases, in other case they have “…become too closely adapted to and integrated within an ecosystem that has been replaced or significantly restructured by some landscape reshaping change…”. People generally refer to this phenomenon as disruption and the way they refer to it would seem to imply that it’s something that happens to or is done to a company. The role of the organization in its own difficulties (or demise) isn’t, in my opinion, well understood.

Last Friday, I saw a tweet from Noah Sussman that provides a useful heuristic for predicting the behavior of any large social system:

the actual reason behind the behavior:

It’s not that people are actively working to harm the organization, but that when there is no leadership, where there is no design, where there is no learning, the system ossifies and breaks down. Being perfectly adapted to an ecosystem that no longer exists is indistinguishable from being poorly adapted to the present context. I’m reminded of what Tom Graves stated in “The game of enterprise-architecture”: “things work better when they work together, on purpose”.

Without direction, entropy emerges where coherence is needed.

This is not to say, however, that micro-management is the answer. Too much design/control is as toxic as too little. This is particularly the case when the management system isn’t intentionally designed. Management that is both ad hoc and rigid can cause new problems while trying to solve existing ones. This is illustrated by another tweet from Friday:

The desire to avoid the “…embarrassment of cancellation” led to the decision to risk the lives of a plane load of “…foreign TV and radio journalists and also other foreign notables…”. I suspect that the fiery deaths of those individuals would have been an even bigger embarrassment. The system, however, led to the person who had the decision-making power to take that gamble.

The system works the way you built it, even when you didn’t intend to build it to work that way.