↤ February 22th, 2011
The following is a rough translation of selected parts of Adobe CTO Kevin Lynch’s recent blog post on “The Multiscreen Revolution”.
Please note that I don’t hold any
substantial grudge against Adobe and wish them all the best finding a way to make the web a better place.
With apologies to John Gruber.
We are in the midst of a revolution across a variety of screens, with new input methods, new formats, and new distribution models. This revolution is being fueled by several fundamental drivers: processing power growth, powerful portable batteries, increasing bandwidth for wireless Internet connectivity, and a wide array of screen sizes and device form factors.
Average processing power has been on a continuous growth curve for PCs over the past 25 years, with Moore’s Law holding firm as processing doubles every 18 months. This growth is being further accelerated by the adoption of multicore processors.
I remember folklore and don’t shy away from restating the obvious.
In addition to PCs, we now see an increase in processing power on smartphones, Internet-connected televisions, and tablets. This is driving the mobile computing boom and enabling much richer experiences on these screens, with typical processor speeds over 1GHz. Even so, the processing power across these screens is comparable to what the personal computer experience was about seven years ago. While mobile computing power increases, we will continue to have an ongoing processing power gap between PCs on the one hand, and smartphones, TVs, and tablets on the other.
TVs have Internet now.
This creates a challenge for anyone building digital experiences, as they will need to deliver effective experiences across many non-PC devices, not only high performance personal computers.
I’m about to sell you a write-once-run-anywhere solution. But not just yet.
Our approach at Adobe is to take a mobile-first view on the new work that we are doing to design for the more constrained environments, then look to ways to enhance that experience for higher performance environments.
Others are eating our lunch and dinner on mobile. We’ve finally realised this and are running fast to catch up.
Advances in battery technology over the past 10-15 years have enabled the increase in processing power on mobile devices. If we look at this in terms of the amount of energy than can be stored per weight of the battery, the transition from the hefty lead acid batteries in 1945 to lightweight lithium polymer technology today has resulted in dramatic improvement. This is what is enabling us to carry around smartphones or tablets with processing power that delivers such a strong experience today.
I know more folklore, I’m mentioning 1945 so readers think I’m good at being CTO by knowing my history.
This shift follows the classic “s-curve” of innovation, where a technology slowly improves in the early period of its inception, experiences rapid improvement, and then goes into a mature period of slow to no improvement. We are now in the mature phase of current battery technology. It has basically hit a plateau over the past five years.
My editor suggested I pep up this post with graphics, thank god I have analysts as friends, they have all sorts of cool things to put in graphs, like s-curves. They are shaped like an “s”, get it?
By their very nature, mobile devices are much more reliant on battery power, and many web experiences are power intensive.
(Emphasis mine) I watched a Shakespeare play recently.
Since it’s likely that we’re going to be operating with the battery technology that we have right now for at least the next several years, content creators need to consider ways to optimize battery usage.
Technology advances, but not linear. Also, s-curves!
Since Flash is such an integral part of the web, we at Adobe have focused a lot of our energy on optimizing its mobile performance with battery limitations in mind.
By integral I mean skip-intros and the few places where open standards are not yet closing the gap.
In addition to working with our hardware partners to optimize Flash Player for their devices,…
It is not our fault that Flash is so slow on mobile devices!
…we’ve also focused on making Flash smarter in how it manages the CPU resources that it uses. For example, Flash will automatically pause the content that is running when the browser is hidden from view or the current browser tab is placed in the background.
…but we’re willing to look into it on our end, too. Here’s a straw man for our efforts that doesn’t really attack any of the issues that Flash does have on mobile devices.
The display itself is usually the primary power drain on devices, followed by using the radio for communication.
It is not our fault that Flash is so battery-draining on mobile devices!
Some of the simplest things such as turning the brightness down on the screen go a long way toward preserving battery life.
Make sure your tires are properly inflated.
Also, there is a difference in power usage based purely on the number of lit pixels vs black pixels – the more bright pixels on a page, the more power is used.
I don’t know how LED displays work.
Typically, with innovation like advances in battery technology, there will be another s-curve that will slowly ramp up and then give us a major leap ahead of the current state of the art.
There’s a lot of research and investment happening in the realm of battery technology. I expect we will see a major breakthrough over time in this area, which will enable even more radical performance for mobile computing, perhaps even bridging the gap with desktop computing.
We’re going to wait for others to fix any issues that make Flash slow and battery-draining on mobile devices.
The typical connected US household uses either cable or DSL right now, likely running between 10-20 megabits per second. In some countries, of course, it’s faster than this. Wireless data connectivity is starting to increase around the world, and there is a coming breakout where we will see a crossover: wireless bandwidth is going to exceed wireline bandwidth. People actually will have a stronger connection to data on the Internet with a wireless connection, which is being driven by 4G technologies, such as LTE.
I don’t care for anyone in nowheresville.
Wireless operators already are starting to roll this out, starting at speeds of 10-20 Mbps, and the technology has the ability to ramp up to 50-100 Mbps on a per user basis over the next several years. Of course, this speed will vary depending on which type of building the user is in and other factors, but generally we can expect to see wireless bandwidth over time that’s about five times faster than what we’re experiencing today.
I only paid half attention on Computer Networks 101 and didn’t hear the part where it says that latency, not throughput, is the problem with computer networks. I also don’t know how mobile networks work because I ignore the fact that latency is even more important there.
Overall it is going to be a plentiful bandwidth environment, and that’s going to be great for anyone building experiences such as streaming HD video, multiuser games, or rich, live collaboration on the web.
We are going to make the web even slower.
For many years, web designers and application developers looked at the average computer screen size and aimed at that in their work. Over time, this size gradually increased and now we are at a point where this has splintered into many screen sizes. One can no longer design to a single average size.
Remember I was going to sell you a write-once-run-anywhere solution. Get ready for it!
Some content providers have chosen to tackle this diversity through multiple implementations of their websites. But as more and more form factors are added, this approach becomes impractical. Ideally, to take into account all these screen sizes, content can be created once and made adaptable so that it will adjust to a diversity of screens. One emerging approach that is to use CSS to skin your site across displays. Another is to design multiple presentations of content while reusing common elements such as story flow, images, and video that may be dynamically adapted.
Open web technologies are not up to the task.
This transformation from desktop to mobile is happening now. Below is a chart from a recent Morgan Stanley report, showing that desktop connections to the Internet are continuing to increase. In the next three or four years mobile computing is going to exceed desktop computing on the Internet.
Mmmmh sweet charts. Pie charts. Mmmmmh pie.
All of these changes together represent a bigger shift in computing than the personal computer revolution.
I’m thinking CTO-thoughts.
At Adobe, this is an exciting time of change in how we look at building our tooling, runtimes, and cloud based services to help people best express themselves in the multiscreen world. To learn more about these trends and get some insight into how we see this is transforming web sites, digital publishing, video, enterprise applications, and gaming, you can watch my presentation at Adobe’s annual designer & developer conference, MAX (the keynote begins 15 minutes into the video stream).
Finally, the pitch! — Isn’t it great, the future is all scary, but Adobe’s got your back! Because we have s-curves.
Again, Good Luck!