In the last year, there’s been an uptick in discussions by designers and developers on the wisdom of continuing to support older browsers. This discussion has been fueled by the rapid rise of HTML5, and in particular, the scramble of major browser creators to support most of its features. This change has been the result of shorter times between browser updates, coupled with auto-update strategies by everyone except Microsoft. This has opened a much older issue dating back to the early 2000s about the need to support older and non-standard web browsers in design. In the old days, it was IE6 (yes) versus all the 1990s browser experiments. At that time, the competition was fierce, and designers and developers practically attacked their audience for their use of a particular browser. Remember these?
Today, after the Great IE Stagnation of the 2000s, it is HTML5 browsers versus IE6 and other oldies. The browser wars have come back, in all their glory.
The New Browser Wars
Several posts have responded to the rise of ultra-capable HTML5 browsers by suggesting we finally dump old WAP cellphones, IE6, and the other “usual suspects” that contribute to difficult design and development.
Here’s an oldie from Quirksmode:
This post by Louis Lazaris suggests we drop support entirely for these old hounds, and educate consumers not to use them – aggressively, if necessary.
This article has a companion designer manifesto which could be distributed to stakeholders and ordinary users to explain to they why they need to upgrade to a non-IE browser:
Louis Lazaris runs a very impressive and thoughtful website at Impressive webs, so we can’t just ignore this shift. But it is clear that the article is a strike against an aggressive application of the principles of Progressive Enhancement, developed about a decade ago by Todd Parker et. al. This strategy seeks to support the “browser-challenged” with a graduated experience and constant information at all levels of presentation. In contrast, the Manifesto above tries to convert the browser-challenged to a cool new experience – exactly like the late 1990s “browser wars.”
Some, like technologizer.com creator Harry McCraken in his post on the unwelcome return of “best viewed in” language, possibly fueling a new 1990s-style browser war. He cites sites like Apple’s HTML5 Showcase, which displays this message to IE users:
Unspoken, but also implicit in this discussion, is a strike against marketing guys working with (or being) your client that demand a website look the same on all web browsers. Kowtowing to clients who demand a common experience. The thought, most obvious in the discussion of this and related articles, is that dumb clients with marketers who don’t understand the web, and see it as a sort of television, are responsible for unreasonable support of the “browser challenged.”
In its original formulation, Progressive Enhancement implied universal support for everyone at the information level (meaning all browsers got the same content), with presentation of the content graded by browser capability. This may result in a very complex design, and there have been many proposals to simplify browsers by classifying them into “old” and “new” categories. Along this line, a recent post by industry guru Paul Irish suggests that we treat old browsers like black and white televisions, and develop a simplified scheme for dividing browsers into three broad groups: high-res, medium-res, and low-res. A little like Progressive Enhancement, this is Tiered Adaptive Front-End Experiences (TAFL).
The take-home is that we don’t try to create color on a black and white system, and we don’t try to create a modern browser interface with an old browser. Instead (presumably), we design a black and white experience which matches what an all browsers displayed when said old browser was young. Possibly, Paul Irish was also thinking of those crazy marketers again – if they only understand TV, let’s describe the web to them with a TV metaphor.
There’s a lot of excellent follow-up discussion on this post by Paul. Instead of hedging, the argument for supporting old browsers comes out, naked, and in the open there. Opinions range through, as far as I can tell, the following categories:
Designer-centric arguments in favor of dropping old browsers
- We need to educate our clients in graded browser support
- We need to educate our users to upgrade their browsers
- Let’s throw down the gauntlet and tell those lame users on old browsers to shape up
- Kill those who use Internet Explorer, and squirt their blood
- Provide TV-based browser groupings like HD, SD, LR groupings. We have clients and stakeholders who need this language
- Let them know that IE is doomed, since it doesn’t auto-update like Chrome and Firefox do
Here are some technical points
- TVs adapt content, so they are NOT like computers and browsers (this leaves out “content adaption” on old mobiles, which is actually like TVs)
- We can take a tip from 3D games and WebGL apps, which “watch” their environment, and dynamically adjust to what the browser is capable of (FPS)
- You can’t predict the future – browsers trends may change in the future, so we can’t assume that Chrome is the future
FInally, contrary arguments from a client/stakeholder perspective
- If I show a design in Photoshop, then the site doesn’t look the same in HTML/CSS/JS code, my clients freak out
- My clients overlay a transparent wireframe on the final design and freak out if it isn’t pixel-perfect on any browser
- Browsers are always going to be different, now and in the future, and clients will always require that they look the same
- Clients won’t Chrome frame over polyfills because polyfills are invisible, but Chrome Frame because it requires the user do an additional installation
- Customers will compare your site’s appearance to other websites using gobs of polyfills to look the same on all browsers
- If you don’t make everything look the same, clients assume you are LAZY
- IT departments will always resist browser upgrades, and see browsers as something to customize
My own school (a design school) won’t install Chrome by default. The systems (like most corporate intranets) are locked down doubly-hard, so Chrome Frame can’t be installed. Even if you get it installed, a program called Deep Freeze by Faronics that erases all changes, so you have to re-install everything the next morning. You also have to re-install any changes if the computer is re-booted. Similar methods preventing users from upgrading are commonly found in corporate Intranets, which have millions of firewalled computers not counted by browser share websites like StatCounter.
How does a Sustainable Virtual Design perspective work into this new browser war? In the following, I consider browser support as a part of a sustainability framework.
Sustainable Browser Support
What does a sustainability perspective tell us about supporting these old dogs? Let’s review the core principles of sustainability, found in other areas of (non-virtual) design (taken from my “Info” page)
- The 3Rs (Reduce, Reuse, Recycle)
- Sustainability Helix
- The Living Principles
Sustainable Virtual Design (Sustainable Web Design or Sustainable Game Design), adaptation, taken from the list developed by Nathan Shredoff in his book, Design is the Problem: The Future of Design Must Be Sustainable:
|General Sustainability Principle||Sustainable Web Design Goals|
|Make meaningful products||Make websites that are have real value, not fashion or tech-tricks|
|Easy design rollback||Iterative or Agile design workflow|
|Source Renewable Materials||Switch to a “Green” webhost|
|Design products to work in the future||Implement classic design strategies|
|Design with the user in mind||Create effective User Experience (UX)|
|Ensure democratic access||Build accessible, responsive websites|
|Interchangable Parts||Apply standards-based design|
|Minimize energy and resource consumption||Web Performance Optimization (WPO)|
|Don’t corrupt the virtual system||Search Engine Optimization (SEO)|
One of the most important parts of standard sustainability frameworks is “democratic access.” In the virtual world, this means we have to consider the people supported as a counterweight to issues like energy efficiency and efficient design of websites without lots of polyfill cruft.
A sustainable framework means that we have to start with a customer focus. Until web browsers develop artificial intelligence, we’re supporting people, not browsers or web pages. We have to maximize access to our site so it can be used by anyone. The only exception (also seen in industrial and other design areas) is where we are trying to get users to use a “greener” product. In other words, in the case of sustainability, we follow these “push-pull” rules:
Support design by everyone
Encourage/pressure/force consumers to upgrade ONLY if the new product is “greener”
In other words, we should force people to use new browsers only if we can demonstrate that they lower the carbon footprint of the user, the site, and the general Internet. Anything else conflicts with democratic access. It is the application of paternalism, or a “nanny state” theory driving a practice of forcing end-users to do things. In this case, IE may actually have an edge. Recent studies sponsored by Msft support the idea that IE uses less energy than other browsers. Here’s a sample chart:
As we often see in political discussions, the anti-IE commentary following the article on Mashable doesn’t even address the issue – instead, it is a place for “professionals” to spew their stock venom. There’s almost no discussion of energy use and its impact on web sustainability. Instead, we hear (from so-called professionals) “I hate IE” instead of actually considering the problem. In these flames we see “designer-centric” thinking once again rearing its head – users (in case you didn’t know) could care less about IE. But they might be interested if IE really uses less power and lets laptops run longer.
I suspect these cohorts would celebrate Google Chrome’s faster performance, which may imply reduced power consumption:
Designer-Centric versus Sustainable Perspectives
The very idea of browser classification as “good” or “bad” implies a designer-centric design focus, since few people know or understand browsers. Even tech-geeks focus on their hardware vendor, rather than specifics of operating system software. They don’t know, or care about Firefox versus IE6. Looking at Paul Irish’s discussion, we see some posters are user advocates, while some argue in favor of design saving the people from themselves. This is, in other words, the “designer-centric” versus “people-centric” discussion that has gone on ever since Web 2.0 got started. Those supporting massive polyfills for old browsers essentially argue in favor of end-users and stakeholders, and those that support “the future” argue in favor of designers and developers.
If we are thinking sustainability, there is no rule that says “be developer friendly” – it’s all about the people.
In other words, we can’t leave out the ‘browser-challenged’ to satisfy designers.
Designers also must remember that their own work is hardly green. Graphic and Web design uses lots of big, heavy workstations, and the software tools are generally energy and resource hogs, in fact the only programs on a computer comparable to the resource grab of a web browser. Designers stick with things in their own profession. I suspect that if we get energy-efficient HTML5-based “cloud” design tools, most designers will stick with theri CSX tools, despite repeated suggestions that they upgrade into the future.
To wit, another blast from the past:
How much memory do your tools take up today? How many watts? You’re not so great yerself at upgrading.
Using Green Arguments to Promote Browser Upgrades
The above discussion reminds us that sustainability also considers energy minimization. Websites and web pages consume power, and the Internet consumes around 2% of world energy. The percent in post-industrial societies relying heavily on web products and services is higher, and rising. In time, the carbon footprint of the Internet will be bigger than dirty old tech like the airline industry.
While demanding that users upgrade to satisfy the so-called “needs” of designers is not sustainable practice, asking that they “go green” is sustainable. And Old browsers require extensive polyfills to work, thereby consuming more energy. If, as is frequently the case, these old browsers run on old hardware, it is likely to be less energy-efficient as well.
Considering the new browser wars in terms of “green” principles might just break the logjam between supporting cross-browser “perfect” design versus moving the web forward. As I’ve said before in this blog, clients who demand a common browser experience at all costs may listen to “green” arguments. The reason is, that, unlike technical browser issues, marketers, clients, stakeholders, and end-users are all familiar with the ideas or Reduce, Reuse, Recycle. One of the benefits of the HTML5 world is that, in theory, we can produce more energy-efficient design at:
- The production/workflow level
- The page downloading and rendering level
If we play the energy card, clients make just realize that designers and developers aren’t just being “lazy.” Instead, we are shifting responsiblity to them and their vision – they are being informed they are “not green”. It has worked before, in other industries – see how so many individuals and corporations have changed their behavior based on green principles. To put it bluntly, a CEO or marketing beast who demands pixel-perfection can be cut down to size by being told they are not “sustainable”. Designers and developers can estimate carbon footprints, and compare different design strategies. Using a “push-pull” metaphor, design choices can be shown as “democratic access” versus “energy hog websites.”
Now, there will be some clients who want to have their cake and eat it too. These clients will seek out someone who promises to develop pixel-perfect cross-browser design without the energy penalty. But their claims will be refuted, more easily than those by fly-by-night SEO operators. You have a hard time arguing with the Second Law of Thermodynamics.
Replacing “Best Viewed” with Nudge
Does democratic access doom us to maximum polyfill overload, except when we can convince people they are sucking too much electricity? There is an alternate way, which is more nuanced than another round of “best viewed” icons on web pages.
In recent years, what web designers call User Experience, or UX, has been considered in Behavioral Economics under the name of “Choice Architecture.” Like UX, Choice Architecture analyzes how users access and work with a product and service. Analysis of choices often shows that users make decisions that are actually bad for them (on average) when given greater choice. In the extreme, one can use choice architecture as an argument for a “nanny state,” where the government makes the best choices for use when we aren’t likely to. The designers demanding that we drop poor old IE6, IE7, IE8 fall in what I call the “virtual nanny state” category.
Fortunately, there has been lots of work on Choice Architecture, and there are good discussions of how to implement it without taking over and de-democratizing the process. A good example is Cass Sunstein and Richard Thaler’s Nudge theory, which describes how to structure user choices that subtly nudge them to better choices, which protecting, or even expanding the freedom to choose. They call their theory “libertarian paternalism” which means people still have easy choices, but the likely best choice for the most people is baked into the design. This, if you can do it, satisfies Sustainable Virtual Design’s goal of “democratic access.”
While the original book mostly considered political programs, their blog does a lot of implicit UX. Here is their version of a “best viewed in” occasioned by moving off WordPress.com to their own site:
We’ve moved to http://www.nudges.org. Come with us.
The book is well worth a read by designers and devlopers, especially those who are considering issues of support for old browsers. There are lots of ways to encourage upgrades without slapping clients and users in the face. In the process, you remain a good guy and ensure democratic access to your site. And most “nudge” principles are design ideas.
Here’s a great example of what “nudge” means in design, taken from a recent blog post:
If you look at the Nudge blog, you’ll see that the web is of rising interest. In fact, people who are familar with “Choice Architecture” are literally re-discovering design patterns and UX principles on a daily basis, like this one for an action button enabling a “freemium” model for a software service. New to them, but part of their overall theory:
The reason to pay attention to “nudge” is that there is an underlying theory from Behavioral Economics, instead of the rantings of designers against Microsoft and lame users. Nudge theory has developed strategies to set the level of persuasion versus forcing. In Sustainable Virtual Design, we want to use good theory to inform our own practices – there’s little to be gained by holding an unsupported opinion.
A good example of a “nudge” is having a default choice. Each year, we have to renew our decisions regarding healthcare coverage. In a world of absolute freedom we would require that everyone choose, or pay the price. In practice, a failure to choose is counted as a choice – for the current, default platform. Does this sound like a browser upgrade? I think it does. It’s pretty much what Microsoft did in the 2000s with their non-upgrade policy for Internet Explorer. Other vendors opted for a default of automatic upgrades. Most users are hardly aware that their Chrome or Firefox browsers are being updated on a regular basis – it takes a conscious effort to stop the upgrades.
I suspect that Sunstein and Thaler would interpret the call for aggressive upgrades of browsers for those not used to it as a departure from their “libertarian paternalism” approach.
Sunstein and Thaler spend a long time discussing how the attitudes of politicians, social engineers, and others (read designers) require a hyper-rationality on the part of users. This rational behavior simply isn’t part of our brains. We might want it to be so, but in practice, people do all sorts of things a hyper-rational player would not do. A hyper-rational player would get the best browser for their system (and upgrade their hardware), but the typical user will not. You won’t change anything by saying that everyone should be rational, or get zapped.
In the case of browsers, designers are being just plain nasty. Designers have long applied subtle methods to enhance user experience, which in turn implies user choice. Aggressive demands by a website are out, since they violate UX. So how come all this learning flies out the window on browser upgrades.
Nudge may also help solve the “IT problem.” In corporate intranets, it is IT, not end-users, that are the culprits. An attack on them by designers feels like a hack – with similar results. They’ll just firewall IE6 forever and ignore the new web. But developing “nudge” aimed at IT might actually help the upgrade process along. This is a market segment for web design which in the past has been ignored. It is time to consider IT as well beyond irritation to rabid attacks.
To conclude, I suggest that those participating in the browser upgrade controversy consider what they would do if they treated it like a regular user choice, applying UX principles. To understand the back-end social context of people having old browsers, a read of “Nudge” would help. This could result in a better strategy for moving the web forward than pushing clients and users to upgrade or die. Comeon, designers can do a better job here! Behavioral economics, replacing the new browser wars, should form the foundation of democratic access, giving us true Sustainable Virtual Design.