Matthew Bruce Team : Web Production Tags : Web Design Web Development Online Trends Featured
23Jul2014

Browser backwards compatibility testing – is the end in sight?

Matthew Bruce Team : Web Production Tags : Web Design Web Development Online Trends Featured

Web professionals really do have it great. The skilled and experienced are in great demand.

We command some of the highest average (and rising) salaries, also enjoy great job security.

We turn up to fantastic work environments (have you seen Wiliam’s hip and trendy office?) and generally are very happy with what we do for a living.

Web developers even came up trumps in a recent survey by Adzuna (the UK’s answer to Seek), being voted the best job of 2013.

But... ask any web developer what the worst part of their job is; up there with scope creep, impossible deadlines and any shortage of bacon will be the pain of ensuring backwards compatibility.

As much as web development is awesome, catering to the whims of older browsers has been a bugbear of the industry ever since the early days when strict conformance to web standards was deemed altogether unnecessary.

Maybe this was simply due to the youth of the Internet, or a more deliberate attempt by players such as Microsoft and Netscape to create market-dominating browsers at the expense of conformity (??). It doesn't matter - cue a multitude of first and second generation web browsers that interpreted code differently, or even came up with their own proprietary tags (WTF were they thinking?).

It's ironic to think that every time we have that annoying chat with the client about supporting older browsers, it was the (IT) industry that created this mess in the first place.

 

Backwards compatibility - that old chestnut

For years we’ve heard the same old gripes, primarily a lambasting of the shrinking – yet still annoyingly significant – percentage of web users who fail to upgrade their extremely old versions of Internet Explorer. In the red corner is an angry developer who is of the opinion that all users should upgrade or be damned, and in the blue corner is the producer / client who wishes it were that easy, but still as to think about how alienating even a small percentage of users will realistically translate into a larger than desirable number of users being disregarded.

Even as of June 2014, even Internet Explorer 6.0 – the worst browser in the history of browsers, period – retains a global market share of around 1% (once you take out China, who are a bit behind the times).

Doesn’t sound like much but on a heavily trafficked website, even just 1% of one million unique visitors equals around 10,000 users who aren’t getting the experience that they should be.

Take that a step further and consider for the typical e-Commerce website, you’re likely to be only converting between 2-6% of traffic into actual sales (disclaimer - this is very dependent on the industry and the type of conversion, but generally it's a pretty low average), so we’re now talking about writing off in order of 200-600 legitimate, tangible orders or enquiries, simply because we didn’t bother with graceful degradation of an old browser.

Maybe some companies can afford to lose that many customers through simple carelessness, but I know a lot that can’t. Then again – when one considers the additional development required to support IE6 (or 7 and 8 for that matter) the cost of this compatibility work may outweigh any additional profits.

At the end of the day, even if you could justify the expense based on traffic and revenue from projected sales, I’m sure there would be reluctance by many business owners to invest in supporting older browsers that are so certainly approaching their end-life.

 

The end is nigh

The solution has been with us for a few years now, but it’s only been recently adopted by Microsoft since IE version 9 and above. All modern browsers now update themselves automatically – so for users with newer machines, mobile devices and in particular the younger, wealthier and the hip(sters)... it’s nearly impossible to surf the net without having downloaded the latest release.

And since all modern browsers take web standards compliance seriously, we’re finally winning the war. Hooray for common sense!

Now that current and future versions of browsers are behaving as they should - with automated updates as a standard feature - we are seeing the light at the end of the tunnel. It's only a matter of time before we can finally flick support for any browser that isn't part of the new world order. 

That time is sooner than you think - if not already upon us. Right across the IT industry, things are looking up and web developers are right to be excited. It begs the thought again on every new project we commence from this day forward... is it finally time to waving goodbye to the stress, cost and annoyances of backwards compatibility testing?

This all changed a couple of months ago when Microsoft finally announced that they would no longer be supporting Windows XP – and hence support for IE8

IE8 finally kicks the bucket

As noted above, the biggest problem with older browsers was always that they didn’t automatically update. This was true of all versions of Internet Explorer pre version 9 (or IE9). When you don’t have this functionality, users who are not tech-savvy will not be prompted to upgrade and the net surfing experience gradually degrades over time, even though the user probably doesn’t realise it all that much.

While IE6 and IE7 have reached the stage long ago that most web developers stopped supporting them, IE8 was still hanging around like a fart in a phone booth (exactly, who uses payphones these days). Continually hovering at around 10% market share for desktop users. This all changed a couple of months ago when Microsoft finally announced that they would no longer be supporting Windows XP – and hence support for IE8 which was the native browser version released with that OS.

This is fantastic news, and it also hits another demographic which has been hard to shift – those system admins in large corporate firms who have chosen not to upgrade office-wide systems and PC’s while XP was still officially supported.

These firms no longer have the luxury of staying on an obsolete platform which means the next logical upgrade path is Windows 7 at worst, or Windows 8 with all its bells and whistles and next-gen features such as a touch-friendly interface. Both Windows 7 and 8 will support versions of IE that automatically update, and workers sitting at their desks who were formerly unable to update their web software due to group policy restrictions, will catch-up with the rest of the first world.

 

The future is mobile

Early in 2013, Wiliam as a business decided that all future projects would be responsive over adaptive. The rise and rise of the smartphone, phablet and tablet – as well as touch screens in general – forced the hand of designers the world over, to sit up and pay attention to the types of web interfaces they were pumping out.

Prophetically though we all kind of realised this was going to happen sooner or later, less people than probably should have been were paying attention to some of the more interesting usage statistics available these days. Older statistical reports about web browser market share need to be taken with a grain of salt – or at least analysed more carefully, as they can easily be misleading and lead to the wrong ‘informed’ decisions being made.

Browser usage statistics aren’t as simple as they used to be, because you now have to look at a new dynamic that you never had to look at before smartphones really existed (back in 2010) or became mainstream (roughly 2011). Critical research by Morgan Stanley shows that we have just reached the point globally where there are more mobile and tablet devices connected to the internet, than there are desktops, laptops and consoles.

This gap is only going to widen – even faster than it already has, as new smartphones and tablets become more affordable and reach the hands of not just the affluent but of the mainstream, right across the globe.

The implications for design here are incredible – all of a sudden you have an industry which really needs to think hard about their approach to design.

Do you now design for mobile first, when traditionally you would start with the larger format? This turns previous design methodology on its head and is one of the reasons why Wiliam made the choice to go responsive with all new client projects.

Next time you read a report that says Internet Explorer 10 has a market share of 10% - remember that’s only referring to half the devices out there, so in actual fact it’s really only likely to be around 5% of connected users. The new trend in the simplest terms, is that Windows 8 is gone; desktops are on the decline in the majority of industries, and Android and iOS are the new flavour of the month.

 

It's not all rosy...

We may have finally reached the tipping point, where the majority of projects (and an increasing number forever more) can do away with backwards compatibility testing.

This does not 100% rule out cross-browser testing. During the laziest of tests on some of the sites we build at Wiliam, I still notice the odd peculiarity when you've been developing a site mainly with Chrome, but then load it up in Firefox to see how it holds up. Not always perfect, but hey, at least we don't need to worry about previous versions of Firefox!

Mobile interfaces stand in their own corner as well. When a client asks, we normally follow a few basic guidelines when it comes to mobile device compatibility:

iOS

  • Recommended platforms - iPhone 4S, 5S, iPad or 4th Generation iPod Touch running iOS 6.0, and above;
  • Compatible platforms - iPhone 3GS or above running iOS 5.0 and above.

Android

  • Recommended/Compatible platforms –Handsets that are running Android version 4.0.3 or higher, using default internal browser (Browser.apk or Internet.apk) or Chrome.

 

Cross-device testing is its own kettle of fish. Plenty of clients will insist that there are bugs with the application, when they could simply be showing on their personal device due to some software they have installed or a setting that is not default or standard. You aren't going to be able to help those people all the time, and the best practice at Wiliam is to only formally test against virtual mobile platforms, such as using an application like Browser Stack.

This way you eliminate the problem of dealing with specific devices and guarantee only that your application works to particular standards and environments. It's also impractical to own one of every physical device that is available on the market - the financial cost and simply the need to keep up with the Jones' by owning every new mobile, tablet or similar device... it would be enough to send you mad, let alone broke.

None-the-less, I think we are in for exciting times and a lot less stress. It's probable that you may have noticed this already. I for one (as a web producer / project manager) have the occasion where I need to be heavily involved in the quality assurance (QA) of a web application.

And it's much more fun dealing with a bug once, than having to worry about testing it in a multitude of browser and obsolete browser settings.

 

Further good reading on this subject

Here are some good links I found when researching this article.

Smashing Magazine

Stats Counter

Reddit

Sitepoint