There’s a well-known concept in programming that refers to the negative effects poorly-made decisions can have on the quality of software over time: Technical Debt. The Wikipedia article gives some examples of the causes and the consequences of technical debt.
This financial analogy is a useful one, as it nicely describes the long-term impact of debt – the longer you have it, the worse the problem becomes. Conversely you can have credit (e.g. savings) in your account for a long time, waiting for the proverbial “rainy day” to take advantage of your good planning. Want to splash out on a new pair of sparkly galoshes? No problem!
At the An Event Apart conference in Orlando in October 2016, Jeremy Keith spoke about "Resilience: Building a Robust Web That Lasts" which was a talk about progressive enhancement cleverly disguised as it didn’t use the phrase ‘progressive enhancement’. In that talk Jeremy dropped a knowledge bomb, calling building sofware using the principles of progressive enhancement like building ‘technical credit’.
This, in my opinion, is genius. It’s a gloriously positive spin on technical debt, which is too often seen as the product of bad developers. It’s saying “you, developer, can make a better future”. I love that.
It appears there is little online which talks about this “technical credit” concept. In fact, the only decent resource I could find is a 2014 paper from the Conference on Systems Engineering Research entitled ‘On Technical Credit’. The author, Brian Berenbach, gives a brief but eloquent introduction to the idea that we should concentrate on what should be done, rather than what shouldn’t be done to make a system better.
From the abstract:
"Technical Debt" … refers to the accruing debt or downstream cost that happens when short term priorities trump long term lifecycle costs… technical debt is discussed mostly in the context of bad practices; the author contends that the focus should be on system principles that preclude the introduction, either anticipated or unanticipated, of negative lifecycle impacts.
Sounds great; let’s stop bad things happening. How? The abstract continues:
A set of heuristics is presented that describes what should be done rather than what should not be done. From these heuristics, some emergent trends will be identified. Such trends may be leveraged to design systems with reduced long term lifecycle costs and, on occasion, unexpected benefits.
Emphasis mine. I’ll wait here while you to read the rest of the document.
At this point hopefully you can see the clear link to the principles of progressive enhancement. Let’s look at a few examples emergent trends – which I’ll call ‘properties’ as the paper uses this term – and the (un)expected benefits that progressive enhancement may give. But first, a quick refresher on what progressive enhancement is.
The principles of progressive enhancement
I can’t put progressive enhancement in a neater nutshell than Jeremy does in his talk ‘Enhance!’:
- Identify the core functionality
- Implement it using the simplest technology possible
- Enhance!
For websites this boils down to practical principles like these:
- Use standard HTML elements; let buttons be buttons and links be links
- Don’t rely on JavaScript (or even CSS) to deliver the core content and functionality of the page
- Make the raw HTML usable even if everything else fails; which means you may well need to develop server-side handling of form submission
- Check for support of features before attempting to use them
But there’s no hard-and-fast set of rules for progressive enhancement, because every site has different functionality. That’s why it’s considered a philosophy rather than a checklist. As Christian Heilmann said, progressive enhancement is about asking "if" a lot.
Emergent properties
Someone once said words to the effect of "the only constant is change", meaning that the only thing you can rely on is that things will not stay the same. That’s good! Progress is positive and brings with it new opportunities.
These opportunities can be seen as emergent properties – new or existing attributes of things which emerge as time goes on. For example, the increasing uses of mobile computing devices and fast home connection speeds are emerging properties leading to opportunities for new types of business. Likewise, the prevalent use of social media and its unprecedented bulk collection of data about its users is allowing new models for advertising – and, unfortunately, more nefarious uses – to emerge.
These emerging properties are often very difficult if not impossible to predict. Progress can lead to unexpected outcomes. Technology in particular is often put to unanticipated uses and exhibits unexpected behaviour when used at scale.
Who, for example, could have predicted the explosion of new devices and form factors just a few years ago. Devices once the domain of science fiction are now commonplace, and the range of new input types – notably touch and voice – is revolutionising how people interact with technology.
While fixed line download speeds are increasing many in developing nations, who arguably are the ones who could benefit the most from widespread Internet access, are stuck with very slow speeds, if any at all. Clearly we have a long way to go to achieve parity in global access to the Internet.
(Un)expected benefits
With such a wide array of both expected and unexpected properties of the current technological revolution, building our systems in such a way to both be resilient to potential failures and benefit from unanticipated events surely is a no-brainer. The ‘On Technical Credit’ paper defines this approach as Technical Credit:
Technical Credit is the investment in the engineering, designing and constructing of software or systems over and above the minimum necessary effort, in anticipation of emergent properties paying dividends at a later date.
This is Progressive Enhancement. It’s about putting some thought in up-front to ask those tricky "what if" questions. Questions such as:
- What if images don’t load?
- What if web fonts don’t render?
- What if JavaScript fails?
- What if a 3rd party resource blocks rendering?
Thinking about these, and many other, potential problems leads you to follow the recipe given by Jeremy which I quoted above:
- Identify the core functionality
- Implement it using the simplest technology possible
- Enhance!
Implementing core functionality using the simplest technology possible – in the case of a website by using well-structured semantic HTML markup generated on the server – gives some expected benefits:
- Resilience against failure
- Better (or at least easier) accessibility
- Raw (i.e. up-front) performance increases
Plus it provides a strong foundation to take advantage of unexpected occurrences; those emerging properties mentioned earlier.
From brand new browsers to old browsers working in a new way, your well-structured HTML will deliver your content even if everything else fails. Support for new input types on a myriad of unimagined devices will be taken care of by the browser – rather than you having to find Yet Another JavaScript Component™ that adds the support you need. And as new APIs are added you can layer on support for these knowing the foundation of your site is rock solid.
Spending Technical Credit
So you’ve built your system carefully, thinking about the many ways in which it could fail. You’ve done ‘over and above the minimum necessary effort’ and can now sit back, confident in the hope that should a rainy day come you’ve accrued enough technical credit to weather the storm.