There’s a lot of predictions for 2010 going around at the moment, which are usually fun and insightful, esp. the ones from Stephen and James ;> If you want some more ribald prognostication, check out the annual Whacky Predictions episode of the IT Management & Cloud podcast.
Rather than list a whole bunch of things, I thought I’d narrow down on one area I’ve been thinking about a lot recently, all the new platforms – both web and device driven – I’ve been seeing emerge and how they effect the IT and technology world we we know it. Much of the below is speculative, but it’s laced in with plenty of here-and-now.
New platforms and devices change what’s required from the IT industry.
This is the first time in human history that we have truly a ubiquitous device… What you can do with a transaction across a mobile platform is very different than what you can do with it a point of sale.
–Matt Quinlan, CTO for Visa, Inc. at #STGEvent09
There’s a host of new platforms and networks that seem ready to (finally) pan out for technology companies and their user-bases. These new platforms are mostly new devices, but they’re also new software development & delivery platforms as well such, commonly put into the category of Platform-as-a-Service, or PaaS, which lead to the applications delivered over those platforms. The devices here are essentially computers in different form-factors. Whether its delivered over the web or through a device, the applications operate in highly networked scenarios, favor consumers over business users, and are geared around using internet native (or “cloud,” if you wish) applications & services.
A host of highly networked and dependent platforms are emerging and available, lead by the ongoing adoption of Internet-based applications in mainstream life and a renewed interest in the smart phone market thanks to the iPhone. While cellphones (or “mobile”) are the current focal point for much of the conversation here, other devices are just as capable and look to be picking up interest and speed. The end-users for these platforms tend to be more consumer-centric. Business has found itself trailing consumer technology of late because IT has been forced over recent years to concern itself with security, compliance, and the culture of careful change instead of focusing on using technological innovation to make more money for The Business.
With the rise of websites like Facebook (with a reported 350 million users) and Salesforce (with 55,400 corporate customers, and 1.5 million individual subscribers), usage of the web as an application platform has gone beyond the document-centric, email, retail, and even desktop-to-web applications that marked the first large part of the web’s life. These larger web sites have started to open up their platforms for third party extensions and even application development, creating the cloud category of Platform as a Service.
Newer applications are not always web-based, but still layer on-top of the Internet. They may still rely on web technologies, but many of these new applications are building their own user interfaces, sometimes require other devices (like iPhones or tablets) and are distinctly not “web native.”
The majority of technology companies that should be servicing these new devices and networks are still geared around contemporary web servicing and on-premise offerings, both in software and systems. This is little wonder, as these companies are solidly pinned down by he shackles of success: servicing traditional IT departments. Many are in the process changing, but are not ready to deliver right now.
Cloud computing is looking to become a de facto standard for infrastructure in this space, while the concepts that fall under the title of “agile infrastructure” seem to be the best hope to meet the demands of this space: frequently delivering new applications and updates, while being ready for spikes in resources like network, process, and storage demand.
Smart phones, tablets, toys, TVs, and other devices are now on the Internet. Software goes here.
When it comes to the non-traditional (not web or desktop apps) delivery platforms, there’s a host of previously non IP-addressable devices that are getting on the Internet. Some of these devices tends are lead by a single brand, having a more general market, while others lack a stand-out leader at the moment:
- iPhone – the renewed smart phone market with the ever present Blackberry, Android-driven phones on the scene, and others like Palm with the Pre. What defines this space currently is having the internet (and the applications and services it brings) in your pocket, at all times, along with your cellphone and a camera. Without the huge industrial design leaps Apple made with the iPhone (and the accompanying, consumer-centric advertising and retail focus), this would still be the “Crackbery” market, which is very different than what we mean now-a-days when we say “smart phone.”
- Kindle – the “tablet 2.0″ market. While the Amazon Kindle is eBook centric, its the current brand to watch as we wait for the rumored Apple tablet to come out and watch other people figure tablets out. Clearly, there was something wrong with the generalized tablets and flipable laptops from several years ago: perhaps narrowing down the focus to eBook reading and (once again) focusing on consumers as end-users (vs. generic “business productivity) helped.
- Netbooks – after overselling the poor performance of early netbooks, the signs are that sales and desire for netbooks are still strong. Hardware vendors are certainly delivering more of them, and I have to think that the low price tag has much to do with it. Spotting a netbook endcap at Target is screws with your traditional computer channel sales notions. The light form factor with longer battery life and built in webcams, memory card slots, and networking do make for something that looks like “everything you need” on paper. While netbooks are easy to dismiss as “just small laptops,” that’s the whole point: netbooks make laptops more accessible to more people, both in low price and with more portability that allows people to take netbooks with them more frequently, therefor, using them more often.
- TV – while we’ve been relishing the decline of Mainstream Media for many years now, some TV and set-top box manufactures have taken another whack at the WebTV idea – using televisions as an end point for the web and internet. LG is embedded Netflix access into their TVs, Roku and AppleTV are streaming video, and gaming consoles (see below) are making the TV into a more general computing device.
- Games – with the introduction of the Wii, gaming consoles widened their audience beyond the shoot ‘em up crowd, to virtually everyone in markets who can afford them and has leisure time. Highly network enabled and dependent systems like XBox Live hook up this widened user-base to the Internet; with XBox, to services like Facebook, Last.fm, and Twitter with millions of users already (though, probably less after the initial burst of interest). Netflix and other streaming services pop up here as well, making gaming consoles general purpose Internet devices.
- Toys – just as with these other devices, toys are becoming both computers in their own right and network enabled. There are many examples here, Freaky Creatures, Webkinz, the new Dora teenager doll to name a few as gadget master Omar Gallaga points out, and several things from Leapfrog. The network enablement may be “dumb,” such as interfacing with the toy over a USB plugin with a desktop and web application (with some brilliantly simple hacks here, such as the Poken), but toys are nonetheless becoming another new devices that interacts with the Internet.
- Cars – slower to adopt to the emergence of these other new devices, cars have finally caught up to interfacing with other devices and, more weakly, the Internet. In non-US markets, there’s integration beyond GPS. In Korea, at night driving in from the airport to Seoul you can spot people downloading and watching(!) real-time to TV to devices in their car. Cars increasingly come with Bluetooth and USB drives, like the newer Ford Focus which can be used both for “dumb” connectivity, but also more sophisticated functionality like making calls and playing music.
There are more examples out there to be sure, and much of the above is in the emerging rather than “buy it now!” category. And, again, there’s still the same strong-hold in web applications among these new platforms, the above are just new devices – or “screens,” if you’re Adobe or Microsoft – that fit into this ecosystem.
Users no longer tolerate slow and dumb computers.
Whether for business or personal use, the applications and services delivered over these new platforms and networks are expected to have several characteristics:
- Fast Performance – users of these new devices have not been trained to wait for “the computer to boot” and at best expect to have to reboot their internet router every few days. For any given application or service option, there are typically so many different options that users will quickly switch if they encounter poor performance. Devices can be another story where switching costs for the expensive devices can be prohibitive. That said, users can be easily disillusioned by a slow device and give up on the category entirely, reverting to desktop use.
- Always Available – if there’s a network connection, users expect full functionality. More over, users are increasingly becoming aware of artificial closed systems (like Sony’s or Apple’s) where devices are not universally compatible with each other, across brands. If and when USB power adopters become ubiquitous for cellphone and other devices, users will have a daily reminder of how useful standard interfaces can be to them.
- Rich Experience – users of these new devices expect a rich user interface and, equally importantly, user experience that drives workflows that make sense, like matching mobile form factors. Remember that much of the renewed excitement in this ecosystem comes from the beautiful industrial design Apple does. While business-centric IT has gotten away with poor user experience in the past by claiming cost savings by delivering poor UX, users of these new devices have yet to have their expectations
- Frequent Functionality – be it installing new applications and services over the air or easily updating existing applications, users expect the applications on their new devices to be in constant development. Users expect little bundles of fun and excitement routinely through the life of the application. These bundles can be the tiniest of feature upgrades, but their rapid frequency is key. In addition to delighting users, frequent functionality keeps users from getting bored and moving on to alternatives.
Uses & Scenarios
Technology every where and at all times changes how people go about their daily work and lives.
Users of these new platforms are typically interacting with web sites and services, and at the moment, very often are geared towards the social networking. More generally, many of the users are simply communicating with each other for personal or business reasons, often sharing simple messages, but also photos, videos, and documents. The goals range for simply keeping up with people to collaborating around a document or other artifact.
Media and entertainment seem to top the list too, both with streaming and downloadable formats: music, movies, and TV. Services like Salesforce have made good business out of servicing traditional business IT needs with these technologies, and other large companies have been releasing small, piecemeal offerings of their existing services.
For many white-collar, information workers, very little is actually “manufactured” beyond a presentation or a document – the activities of planning, reporting, and managing trump any quant notions of creating something physical. On the other hand, existing web technologies and these emerging technologies are enabling a revival in something akin to cottage industries, though the globalized market for their output can have both good and bad effect on smaller outfit’s pricing.
Though something of a cliché story now, farmers in developing countries can use mobile technologies to lookup more favorable market rates, avoiding selling at a poorer price demanded from middle-men using information arbitrage. But, with eBay, Craigslist, and niché retail like etsy allowing everyone in a given market to put their surplus and handmade crap up for sale, “rare items” are can no longer drive a higher price from geographic scarcity.
Conveniently, the underlying structure and processes for personal and business use when it comes to “activities” are very similar. Even in the area of security, both business and personal users tend to want the same high degree of control and accountability for who has access to what piece of data, if only after they discover the lack of those protections.
What’s different this time is that the new technology actually seems to work; but it’s not as open as we’re used to.
These new platforms cover the range from general purpose, standardized web offerings, components and units (like netbooks) to highly specialized and proprietary (like iPhones). All of the them take advantage of the Moorian march in increased processing power and ever lower chip prices, along with cheap storage and improved (though not yet ubiquitous) broadband.
Front-ends and Application Stacks
Front-ends for these new applications vary, and the user interface frameworks are highly fragmented to a degree that’s alien to the web application world. Mobile devices typically have their own, native frameworks, while tablets, TVs, game consoles, toys, and cars certainly favor their own toolkits over something “standard.” On the web, Ajax dominates with some encroachment from RIAs like Adobe’s Flash Platform and, Microsoft hopes, Silverlight.
Aside from web standards, there’s very little when it comes to standards here and the have been moribund for sometime (though that’s starting to change, hopefully). With the exception of Google’s ongoing attempt to insert HTTP and HTML 5 as the standards in this space, most vendors are opting for their own technologies:
- Native SDKs – such as Apple’s Objective-C toolkit for the iPhone and, we’d think, their rumored tablet.
- Flash from Adobe and Silverlight from Microsoft – both are gunning to be the Apple and Google alternative for user interface technologies across Adobe’s “open screens” and Microsoft’s “three screens,” the two phrases each uses to describe their ventures in this new devices space.
- “Open web” – led by Google and supported by Mozilla and the hoards of developers used to web-centric and primarily open source toolkits, this camp is trying madly to extend the success of HTTP and HTML (via HTML 5 hopes) as the primarily development platform for these new devices.
- Protocols – these seem more standard or at least de facto, TCP/IP, email, IM, XML, JSON, and so on.
All your logins are belong to us
Identity management standards are just now shaking out here with traditional identity frameworks struggling against their enterprise complexity. While consumer-centric, “global identity” services like Microsoft Hailstorm had dramatic failures in adoption in the past, identity standards developed outside of the traditional standards process like OpenID and OAuth are getting shocking traction. Services like Facebook Connect are getting a huge amount of use. This go around, users seem much more willing to allow targeted advertising centric companies like Facebook manage their identity than desktop vendors like Microsoft, which I personally find weird.
Servers & Infrastructure
On the backend, the servers and other infrastructure are largely the same at the hardware-level – just faster and more resource-rich as dropping prices allow for. What’s different are the ways in which this infrastructure is deployed, the software used to deliver services to applications, and the management software used to control and manage all of it:
- Cloud Computing – applications delivered for these new platforms require web-strength scalability, and both the cheap boot-strapping and ongoing costs of running the applications and services. Traditional on-premise data centers that power application back-ends rarely have all of these characteristics at once: being both affordable and “enterprise grade,” as we used to call it. More than simply being cheaper, cloud computing technology focuses on rapid software delivery versus perfect and careful, water-fall IT provisioning and change management.
- Agile Development – agile software development’s emphasis on rapid software delivery and the highly disciplined project management that goes along with it fits well with the requirements for frequent functionality and the ever emerging technological landscape of these new devices. Agile development narrows a development organization’s focus to (a.) narrowing down software requirements to something doable in the given (often short) time frame, (b.) development practices that deliver small, but high quality chunks of code that are operational, and, (c.) a tight feedback loop and desire to continuously approve through self-study and project management analysis. Instead of focusing on accountability and reporting (as traditional, water-fall methods do), Agile cares primarily about delivering software on-time. (One of my former BMC colleagues has been writing about this take on Agile development over at The Agile Executive a lot recently, along with podcasts I do over there as well.)
- High performance stacks – rather than buying into “enterprise grade,” do everything platforms and frameworks, developers for these new devices will have to use the smaller, higher performance, and sometimes highly customized stacks. What I call “stackless stacks.” Standardized and general stacks are sometimes available, but even if options are open source, they’re often in a “standards” world of their own with respect to easy compatibility with other stacks. More importantly, the dynamic deployment and production requirements requires underlying applications stacks that are near-effortless for sys admins to manage, something traditional stacks have often done poorly.
- Agile Infrastructure – the missing piece to realize agile development‘s full potential, Agile Infrastructure allows those managing the applications and services in production to rapidly deploy new changes, meet spikes and dips in performance demands, and go against their trained (and rewarded) practice of fearing change. More important than deploying the initial versions of an application or service, operations staff must adopt practices and tools that let them rapidly respond to customer demands around existing applications and services, and also keep those existing and older applications in good, running condition.
These four areas are really just different faces of the same thing, delivering software more frequently. They appropriately share the same structural need to unify the roles in IT more closely. Smaller shops that have had large success in this new device area (and high traffic web sites) tend to collapse the development and operations role as much as possible, and organizations oriented around traditional IT roles and divisions will likely need to do the same.
Servicing New Platforms
How are IT vendors well positioned or not for all of the above?
Much of this vision made up what I call Sun’s Pink Dot Theory, but Sun’s execution and timing seemed off. Really, the Pink Dot Theory was just an early version of parts of IBM’s Smart Planet.
IBM’s focus there is instead on tooling current industries transformation and roll-out rather than making most of their bets on developing countries (or “growth markets” to use kinder economic speak) and pure open source. Along with a lot IT modernization in previously technologically moribund sectors like waste water management, for our concerns here IBM’s Smart Planet starts from the answer to the question, what does an IT company sell to the world when everything is IP addressable, when we have The Internet of Things? Much of IBM’s Smart Planet is taking these kinds of new technologies and trying to implement a cyberpunk utopia, an Internet of Things with the hope of avoiding the dystopia that comes with that cyberpunk future. What I like to call dry-cleaned cyberpunk.
The Adobe Open Screens project plays well here, and Microsoft is racing towards a similarly good position. Redmond still has a very desktop-driven focus – which will probably do well no matter what. Google’s goal here – their big bet – is for HTTP and HTML 5 to win.
Apple is, of course, the one most focused on the consumers of these new platforms – above developer, IT, etc. Their M.O. is to only care about the end-user experience and traditional consumer marketing here which is working well for them.
Cisco is in a weird position where they talk well to this vision (that’s John Chambers for you), have the networking equipment to service it, but act like its software and platforms they’re selling instead of packet handling. There’s a classic cash-cow vs. green field problem there, but it could use a bit more finessing.
Chip folks like Intel, AMD, Nvidia, and co. are also well positioned as long as they stick to their strengths: innovating new features and higher speeds in their chips, lining up and encouraging “partners” who use those chips, and driving demand for the chip’s functionality among consumers.
And then there’s the hardware and cloud people. In this world, you can’t use the FUD excuses of security and compliance to slow down change to new models of IT (as you can, currently, in the cloud computing world). Elasticity and auto-scaling do become important successful applications and people like Sonoa are doing some interesting work here in the cloud performance management space.
Servers and storage become important, but the question is owning vs. renting: on-premise vs. cloud. The (initial) cost of hardware vs. metered cloud services could become negligible soon, making the choice one of functionality instead of cost savings. The real costs look to be coming in managing the infrastructure and time spent provisioning systems for the new applications and services, and then doing ongoing change management. Speed is what tends to matter here, over completeness and sometimes quality – that type of thinking is anathema to contemporary IT management theory.
Systems people like IBM, Sun, HP, Oracle/Sun, Dell, and so on need to focus on the management of agile infrastructure. What that means is a discussion I’ve had and seen going on many times over the past year with the people from places like Puppet (Reductive Labs), ControlTier, Chef, VMWare/SpringSource, and increasingly independent software developers (thankfully) operations types.
Most IT Management software vendors are focused on managing the morass of 20+ years of legacy IT, enterprise applications and workloads. All important, not a foundational part of these new platforms. As such, the incumbent management software is not as well poised to help systems outfits – or themselves – deliver on agile infrastructure. Re-tooling and re-focusing is needed as the basic desires are drastically different – frequent delivery vs. careful change. The resulting processes and software on both sides mirrors those concerns.
Servicing the Civilians
Call it the “consumerization of IT,” or whatever you like: the core difference with these new platforms is that end-users expect more out of their “computers” and the related software. The tech world has to take cures from Apple, Google, and XBox: keep innovating internally like super-dorks, but learn to deliver more on the benefits and functionality that “civilians” want and less on what IT-heads crave.
(I haven’t had a chance to read through the 600+ pages, but this detailed, numbers oriented report from Morgan looks highly related to the above.)
Disclosure: several folks mentioned or implied are clients. See the RedMonk clients list.