IT Does Not Love iPads, and that's a good sign

I'm not normally in the business of going full-Macalope on stupid articles about iPad. After all, there aren't enough hours in the day. Having said that, one particular article has surfaced recently that I hear is being used by at least one local authority IT team in Scotland to justify an LA-wide ban on iPads.

The article by Michelle Fredette in Campus Technology entitled "IT Does Not Love iPads" is so completely off-balance that it's almost funny - until you realise that this kind of weak thinking is actually being taken seriously.

In case you don't make it to the end, here's an executive summary of the complaints:

• iPads come in boxes
• App licensing works differently from Microsoft's
• Apple TV is finicky to deploy at scale
• You need a whole new network for iPads

The only one of these that actually makes sense is the complaint about a device (AppleTV) that is, in fact, not an iPad.

Anyway, let's get into it.

Is there a higher ed institution in the United States that has not fallen into a swoon over iPads? Some colleges hand them out by the thousands to their entire student body. Others stockpile hundreds for use by faculty, staff, and administrators, or to be checked out of the library by students. On many campuses, iPads have taken over the hearts and minds of everyone.

Let's start by using words like "swoon", "stockpile" and "hearts and minds" to get it clear from the start that all this enthusiasm for iPad is just a mass delusion. Got it?

Everyone, that is, except the IT department.

Given our decades of experience with computers that the IT department love and staff and students hate, I think this can only be considered a good sign.

These sexy tablets might be the apple of faculty and students' eyes, but for IT directors and their staffs, working with iPads in an enterprise network environment is not the stuff of a love affair.

"Sexy". As we know, Real Serious Business can only be done on thick, black computers that people don't like.

To state the problem simply: iPads are designed for consumer use, and as such, they're not set up for large-scale implementations.

I hope someone told LAUSD, who just voted to roll out a $50m, 47-school iPad program. Or McAllen School District, who already have 25,000 units in the field.

They're not even set up for two users to share the same device, much less for sharing over a network.

It's 2013 and the fact that iOS is not a multi-user operating system is still coming as a shock to some people? Also, it turns out the iPad is not a mainframe computer.

For schools making a major investment in iPads on campus, the solution is a combination of new policies and investment in third-party tools for managing the devices.

So there is a solution!

For many other institutions, though, the devices are acquired as needed, or in small batches for specific purposes. In such cases, schools don't necessarily anticipate the additional tools and administration the iOS devices can require--until IT starts bumping up against the limitations of a device that's not easily managed under the school's existing network and resource management infrastructure.

So organisations that don't bother to try and understand what they're doing tend to have a hard time?

The differences between iPad device administration and that of desktop machines or laptops are apparent at all stages of their use, beginning the moment the machines arrive on campus.

You don't even get a minute's reprieve from the awful otherness!

Take, for example, the case of Seton Hill University, a school that has distinguished itself as a forerunner in campus iPad implementations, including being named a second time as an Apple Distinguished Program. In the spring of 2010, the Greensburg, PA, school ordered 1,850 iPads in anticipation of providing them to students for the following fall term. What IT faced was a giant pallet of the devices, individually wrapped or in boxes of 10.

iPads come in boxes! Unlike the Windows PCs that come on tear-off rolls and the Android tablets that come in sheets you can cut to size!

Phil Komarny, Seton Hill's vice president for information technology and CIO, says that his staff had to take each iPad out of the box, update the operating system to the most recent version, image tag it, and put it back in the box to be ready for deployment. "There was nothing else we could do, because this device that they built is completely consumer," Komarny shrugs.

To be serious for a moment, there is something else you could do: get your reseller to do this for you. I'm only leasing 120 iPads and I've had resellers falling over themselves to help us do these chores.

If I was buying 1,850, I'd think this would be the very least a reseller could do for me.

Thomas Hoover, associate vice chancellor and CIO at the University of Tennessee at Chattanooga (UTC), had a similar experience with iPads in a previous role at Pepperdine University (CA), where the IT group handed out several hundred of the devices to students, who turned them back in at the end of the year. "We'd have to manually go through and redo all the iPads," Hoover explains. "It's not like a computer device that you can configure automatically."

So once a year you had to touch every iPad once? Poor dears.

Two years after the first Seton Hill deployment, Apple brought out Apple Configurator, a free download from the Mac App Store that can be used to configure 30 devices at once. But for many campus CIOs, that's too little too late. Those with big iPad implementations tend to rely on mobile-device management (MDM) applications like MobileIron that enable enterprise-level configuration, security, and app management.

And people with big Windows implementations certainly don't have to buy any additional software or services to manage those machines, am I right?

With the iPad, app management is a whole new ballgame for IT departments familiar with licensing and managing applications in bulk for desktop or laptop machines. Rather than selecting from software options hosted on the school's network, faculty and administrators download apps from the App Store, choosing from hundreds of thousands of options. Many schools simply haven't set up a good strategy for purchasing and tracking this new approach to apps. And in lieu of a policy, faculty often download an app using their personal Apple account, and then get the university to reimburse them. (Students are generally not reimbursed for their app downloads, which are considered a "books" expense.)

If you fail to plan, etc.

More than one CIO laments that the ad hoc nature of app downloads can lead to the school purchasing the same app repeatedly, especially if it's a popular one like Numbers. Hoover describes the problem like this: "You reimburse professor A for an app, and then they leave. IT wipes out the iPad because you don't want any sensitive information from the previous person, then yeah, you're going to have to buy that app again for professor B."

Layered model deployment. It exists for a reason.

Some schools have developed workarounds to avoid repeatedly purchasing the same apps. John Haverty, assistant director of user services at Washburn University (KS), says he started to recommend that departments create a generic account for their faculty members to use. "That way, if someone does leave, that software's not going to stay with them--it's going to move on to the next person," he points out.

Again, look at the layered model but you'd probably be cheaper just eating the cost of - what - $30-50 of apps when someone leaves. After all, how often do you turn over staff? The cost of processing their employment paperwork will probably outweigh the cost of the apps they need.

Haverty raises another app-related issue for tax-exempt universities, which is the time-consuming process for getting tax reimbursements on app purchases. "We had to go ahead and make the purchase, then contact someone at Apple to provide the tax ID to get the tax reimbursed." Haverty says that often departments don't bother.

Given that tax systems haven't really caught up with the internet yet, this is hardly a surprise.

Apple has a Volume Purchasing Program for education, released in 2011, which enables educational institutions to buy apps and books in volume at discounts and tax-free. And mobile-device management providers enable schools to manage their Apple volume licensing as part of the broader MDM. Indeed, MDM solutions solve just about all the major challenges presented by Apple devices, including network access, enterprise configuration, and device and app management, as well as security requirements like the ability to remote wipe a lost or stolen device.

OH WAIT - so all this complaining was actually for nothing because solutions to these problems already exist?

Apple, in fact, recommends that enterprises use them. But for schools like Washburn that have acquired devices like iPads and apps in a more informal manner, MDMs represent a time and resource sink that they haven't yet committed to.

If you fail to plan, etc.

But even if your school works with an MDM, there's still the problem of Apple TV working within the campus enterprise. Apple's media receiver is the ideal device for mirroring iPads onto a large screen because the two devices can connect wirelessly. Even better, Apple TV is configured to receive iPad input, so content looks like it should without tweaking.

Yup, Apple TV is great but calling it "the ideal device" and pretending it is essential for an iPad deployment is a great set-up for slamming it in the next paragraph.

Apple does provide a VGA adapter that can connect iPads to televisions and monitors, but, depending on the version of your iPad, you still often have to fiddle to get it to display apps.

The VGA adapter works great - especially with the Lightning port instead of the 30-pin - and I don't know what the "fiddle to get it to display apps" thing is about. I think that's just a lie.

So Apple TV seems like a natural choice for projecting iPad content. And faculty like Apple TV for features like the ability to stream Netflix. But as UTC's Hoover says, "making Apple TV work on the campus network is an abomination on a grand scale."

An abomination on a grand scale, eh?

Hoover may be a bit hyperbolic but he's captured the zeitgeist of IT's exasperation with Apple TV, which is the subject of a July 2012 petition posted on change.org by members of the Educause Wireless Local Area Networking Constituent Group. The petition notes that while Apple has created advertising that promotes the use of Apple TV in college conference rooms, auditoriums, and laboratories, "Apple TV, AirPlay, and Bonjour technologies make it very difficult to support these scenarios on our standards-based enterprise networks."

There's no doubt Apple TV and Bonjour can be finicky at large scale - this definitely needs some work - but I guess my question is what's so wrong with using the VGA adapter if AppleTV is simply impossible.

Issues with AppleTV are tangential to whether or not you should deploy iPad and are absolutely bogus as a justification for banning iPad from your school.

To set up Apple TV, says Jeff Kell, a member of UTC's network services, "you have to enable wireless multicast/broadcast traffic. Such traffic is sent out over every access point on campus, tying up airtime." This means that in an "enterprise setting, every access point will get a copy of every device's advertisements or discoveries in the entire enterprise." Not only is this a huge waste of airtime and bandwidth, Kell says, but it is "a disaster incident waiting to happen, such as some dorm kid streaming porn onto a classroom projector."

I might be old fashioned but I think of "disaster incidents" as being a bit more serious than this. Regardless, anyone who has actually deployed an AppleTV will know that you can turn on an on-screen code that any user must enter on their device to gain access to the Apple TV. For exactly this reason.

Also: don't these massive enterprises have VLANs?

In a blog post, Matthew Libera, performing arts technology consultant at the University of North Carolina at Greensboro, described the steps he took to use Apple TV in his classroom, which included creating his own network and sacrificing access to the internet on his iPad. While it was doable, Libera wrote, "there is no way in heck that I'll be able to convince any of my faculty here that this is a worthwhile undertaking."

Correct. If you have to do this, just get a VGA adapter and get on with your job. This is still nothing to do with iPad per se.

Indeed, many schools just forbid the use of Apple TV on campus. Or they turn to companies like Aruba Networks, which offers a solution for managing Apple TV and other Bonjour protocol-reliant devices in a university enterprise network. But again, this takes an investment in a third-party toolset that's generally attractive only for institutions that are all-in when it comes to iOS devices.

Banning Apple TV isn't actually that unreasonable in its current form. Still nothing to do with iPad.

All-in schools like Seton Hill say that that its iPad program would have been impossible without a major investment in a state-of-the-art network, which included replacing wiring for full campus coverage and upgrading both the wired and wireless networks. The latter, which provides campuswide 802.11n technology, was essential to support the demands of an iPad-oriented university population.

Dunno. I'd say campus-wide WiFi is essential to support the demands of any population of people born after about 1955, regardless of the computer they're using.

With the help of Enterasys network solutions, Seton Hill also revised its approach to managing the network. Now, the school handles network traffic on three virtual LANs: one for iOS devices, one for Mac OS X traffic, and a third for Windows and guest traffic. This approach streamlines network management, but also enables IT to manage network traffic at a more granular level.

OH WAIT - so there actually is a solution to these problems? I'm starting to notice a pattern in this article.

Lynn University (FL) is another iPad-committed campus that will roll out an iPad mini program to all freshman and transfer students this fall. It was able to create a robust campus network as a windfall from hosting one of the presidential debates last October. While the school had to pick up the tab for the new network, it got good deals on some of the technology from participating companies.

Note the subtle - and untrue - inference that you have to buy a new network to support an iPad program.

According to Lynn's CIO, Chris Boniforti, the school had to provide a completely new network environment for up to 6,000 media personnel attending the debate, and it had to be totally isolated from the university's network. Lynn essentially created a whole new network, roughly the size of its existing network, with the understanding that the school would bring it in to replace its aging network once the debate was over. The school estimates that process would otherwise have taken several years and pushed back its iPad mini initiative.

So Lynn got a new network faster because they hosted a presidential debate. Not clear how this is related to iPads.

For schools like Lynn and Seton Hill that have invested heavily in what Boniforti calls the Apple ecosystem, there seem to be fewer hiccups in using enterprisewide iPads. But IT directors who want to incorporate iOS products into their campus ecosystem without making such full-scale investments say they would like a little more support from Apple.

Let's be clear: what the author means here by investing heavily in the "Apple ecosystem" is actually "taking the time to understand how these devices work".

As one Educause wireless LAN constituent mused on the group's listserv: "This is where I daydream about the likes of several Apple engineers reading this list, thinking 'Gee, maybe we should consider how to make our toys work in the actual enterprise.'"

Aaaaannd there's the "toys" canard. It's amazing to me that the FAA approved American Airlines pilots to fly with toys in the cockpit.

So this article boils down to four actual complaints:

• iPads come in boxes
• You don't buy apps like you bought them from Microsoft
• Apple TV is tricky to manage at scale
• You need to buy a whole new network for your iPad program

The first complaint is just utterly bizarre. The second is true but, as the author actually points out, there are multiple existing solutions. The third complaint is the one that's actually true. Unfortunately, it's just nothing to do with iPad. The last is just false.

The entire article follows the pattern of building up small issues (and non-issues) to be insurmountable obstacles, then quietly admitting that a solution actually exists.

Using articles like this one to justify iPad bans is pathetic and embarrassing.

Teaching Programming with iOS and Amazon EC2

I just shut down the Amazon EC2 instance we've been using all school year, so I thought it was worth reflecting on. Last August, I wrote about my new approach to teaching Ruby programming on our iPads.

How did it work? In a word: perfectly.

The Server Side

Back in August, I set up an Amazon EC2 instance. I used the standard Amazon Linux AMI, and launched it in a Micro instance. For neatness, I allocated an Elastic IP Address to the instance and created a sensible DNS name for it.

The next step was to set up user accounts for each pupil. I did this manually because I could, but it's easily automatable if you needed to. Our iOS SSH client, Prompt, supports using SSH keypairs for authentication but distributing the private keys is a little fiddly so I reconfigured SSH on the server to allow password authentication.

Responsiveness of the remote server wasn't an issue at all. We used the EU West (Ireland) region for our Amazon EC2 instance so it wasn't far away from us and the interactive response was just fine. I'm not sure it would have been ideal to actually edit code in, but it was perfectly fine for running programs.

We ran into no problems with with the EC2 server at all. I set it up, it ran for over 5,000 hours of operation without a hitch and I turned it off today. All-in, it cost me £65 to run this server for the past academic year. I didn't do anything to limit the time the server ran - it was on 24x7 for the whole year. If I had automated shutting it down at, say, 10pm and relaunching it at 9am, I could have cut the cost to about £35 for the year but it's almost not worth my time to bother.

The Client Side

In my last piece, I was debating the exact combination of software to use. In the end, we settled on using Prompt for our SSH client and Textastic to edit the code.

Textastic has a really nice feature whereby, if a file was initially downloaded from a remote server, Textastic can send the file back where it came from with one tap. This really helped the overall workflow.

The idea of switching between two apps to edit and run code didn't seem to have much impact on pupils. With the multitasking gestures in use, the switch was easy enough. The biggest complaint was that, on our iPad 1s, resuming the switched-to app was slower than we would have liked. This problem will go away after our refresh.

Occasionally, Prompt wouldn't maintain the connection to the server while the pupils were editing code in Textastic. I haven't yet figured out if that's just a bug or if there's some server-side timeout happening. It wasn't a frequent enough occurrence to warrant a lot of debugging.

The Hardware

This class was the first time we had deployed Bluetooth keyboards for use with the iPad. When you're programming, you're always reaching for those weird symbols in the second and third keyboards on the iPad. Textastic does a good job of trying to help you by creating a row of five-way buttons across the top of the standard keyboard that allow quick access to a large number of symbols. The downside to this is that you lose even more screen space for seeing your code.

For these classes, I bought a number of these Bluetooth keyboards. They cost £9.16 on Amazon and they held up perfectly well for the year. They're not quite as delightful to type on as an Apple Bluetooth keyboard but they do the job just fine and none of them broke.

The Workflow

Once students complete an assignment, they have to turn in a report with some text, their source code and evidence of systematic testing of their code. We went through a few iterations of the exact workflow for this but, in the end, we settled on copying and pasting the code from Textastic and the testing runs from Prompt into Pages and presenting them in a monospaced font along with the rest of the report. The only downside to this workflow is that you lose the syntax colouring from Textastic.

This document was then emailed to me as a PDF where I could mark it digitally using PDF Expert on my iPad and return it to the students via email and archive my copy.

The Future

I'll definitely continue doing this. Given the success of this project, it's almost inconceivable that we could justify provisioning a whole computer lab for just one subject.

On a wider note, I think this milestone is emblematic of the increasing maturity of iOS. In 2010, computer programming was one of the cardinal subjects that iPad sceptics insisted it would "never" be possible to teach using an iPad. Well, guess what? Add another item to the pile of things that people said would "never" be possible in computing.

Never's a long time.

Beyond Consumption vs Creation

When the iPad first launched, many people reached for a quick analysis that it was a device "only for content consumption". Despite time and experience having proven those people quite obviously wrong, the debate seems to persist as to what the iPad is, precisely, for.

My own opinion is that the iPad is for about 80% of all tasks you can conceivably do on a computer. I have never thought of the iPad as a distinct entity requiring a total first-principles relearning of what it means to use a computation device.

As I've written before, the question what you want to do with your computer has never had more impact on exactly the device you should buy. Therefore, it's still relevant and worthwhile to ask the question of the iPad: what are you capable of, and what are you best at? Further, as the iOS ecosystem has developed, another question: if I add these accessories to you, what can you do now?

Still, I feel that the consumption/creation split is far too simplistic a curve to grade these devices on. It recognises almost nothing about the user's task beyond whether it's an input task or an output task. There's far more subtlety that we can reach for.

I'd like to propose a more useful pair of axes on which we can place these devices - smartphones, tablets and traditional PCs - than simply consumption/creation. I've been thinking about this for some time and I think it has some usefulness.

Task Complexity vs Task Duration

I'd like to propose that we can look at the 'sweet spot' for each type of device along two axes: task complexity and task duration. Task duration is the more obvious of the two: how long of a continuous period will you be using your device for the task.

Task complexity requires a little more unpacking. When I talk of "complexity", I'm looking at a combination of factors that make a task complex:

• The number of steps to completion
• The extent to which you're combining data from multiple sources
• The amount of data that is being manipulated
• The linearity or otherwise of those steps - the less linear, the more complex the task

There may be other types of complex task that I haven't thought about. The exact specifics don't matter too much but these give you the general idea.

Given that, here's a chart of how I think about the 'sweet spot' for each type of device.

So what does this chart really say?

I place smartphones near the origin. They're good for simple tasks done for a middling duration or tasks of moderate complexity for a short period. For example gaming, which is a fairly non-complex task, can be quite acceptable on a smartphone for a reasonable amount of time. On the other hand, editing a spreadsheet on an iPhone can be done but it's not something you'd want to do for a whole day of working. Many of the most effective phone apps that take you through a series of steps do so in a very linear and directed fashion.

The iPad section of the chart has a couple of notable features: the dog-leg area at the top-left and the area at the bottom-right of the chart. Let's dig into those.

Firstly, consider tasks of maximum complexity done over any duration: the iPad doesn't reach into that area of the graph at all. That's simply because there are some tasks of sufficient complexity that the iPad cannot currently be applied to them. The reasons are varied but fall into one of three areas:

• The hardware is not powerful enough yet. Examples here would include managing an entire high-resolution photographic library in a hypothetical "Aperture for iOS". The iPad simply doesn't have enough storage to make this possible, although the new 128GB iPad may well take a bite into some of these data-intensive tasks.
• The software has not been written yet. An example might be doing some CAD/CAM design. Perhaps iOS doesn't offer all the APIs required for some apps yet. We can hope that iOS 7 will start to eat into some of these tasks.
• App Store policy doesn't allow it. The classic example here is all the programming tools that we might wish to have on iOS which can't be brought wholly to iOS until policy changes.

Similarly, there are tasks of low-to-medium complexity that can be adequately performed on an iPad for long periods of time. Examples might include annotating PDF documents with some of the excellent PDF apps on iOS, or managing photos in iPhoto, composing music in GarageBand, reading iBooks and so on.

In the middle of the chart lies a broad area of tasks which are moderately complex, done for moderate amounts of time. This is where iPad excels and why it is such an excellent computer for schools. I've never argued that any current or past iOS device can "do everything" - patently, it cannot - but I do argue that it can handle 95-100% of everything a computer is typically called on to do in a school setting. The majority of our classes now use iOS exclusively, despite easy access to Mac laptops being available.

Finally, there remain several tasks for which computers are used which tablets remain unsuitable for the reasons listed above. Simply think of the apps that are missing from iPads: Final Cut Pro X, Aperture, Logic Pro, iBooks Author, Adobe Photoshop. These tasks - for now - remain the preserve of the traditional "desktop-class" PC (in which category I include laptops).

What would it take to push iPad into some of those areas? Well, the simple addition of a hardware keyboard can extend the duration that many people can use their iPad for.

Taking another step into the PC's territory may also call for something I've never really discussed before: a larger iPad. It's possible that a 13" or 15" iPad with 128GB of storage might open up entirely new categories of application to be built for iOS.

You can see that, with the iPad mini pushing down towards smartphone territory and the 128GB iPad enabling a certain number of data-intensive use cases, the reach of the iPad is growing. I hope that, in time to come, future versions of iOS will enable software supporting tasks of greater complexity to be built.

I think that task complexity vs duration is a much more useful framework in which to place smartphones, iPad and traditional PCs. There may be other areas of comparison - for example the physical context of use - but I strongly believe that we have to move beyond simplistic arguments about consumption and creation.

The Refresh: The Device

In the last piece, I explained my thinking on the platform we're choosing for our refresh. We're sticking with iOS, not just because we're already on iOS but because I don't see compelling reasons to switch to any other platform.

A few people got in touch to ask why I hadn't considered Chromebooks in my analysis. There are a few reasons. I know there are several "Chromebooks" available but the one most people are talking about is the Samsung Series 3. From a hardware point of view, I was surprised that the Series 3 only claims 6.3 hours of battery life. That's pretty poor for an ARM-based device. Heck, I can get better than that from my Core i7-based MacBook Air. Long battery life is not something to be sniffed at - it's genuinely transformational in the classroom.

I also think the laptop form factor is limiting compared to a tablet. I don't disagree that a laptop can be a very strong form factor for document production but the laptop still generally requires a surface to work on, is difficult to use standing and lacks functionality as an integrated media capture device (i.e. shoot video on the device then edit it).

I'm also not totally convinced about ChromeOS. We use Google Apps, so actually adopting Chromebooks wouldn't be particularly hard for us. I just struggle to conceive how we would do the range of things we want to do with computers using only web apps. If your main uses are office-type applications, the web and email, I'm sure the Chromebook does a pretty good job. Right now, I think I'd need to see the platform mature significantly and reach into areas like video editing, audio editing, rich art tools and so on.

So, with that out of the way, let's talk iPads. The one question I got after the last piece that beat out the Chromebook question was why I had not written about iPad vs iPad mini. The reason I didn't include that discussion in the last piece is actually interesting in itself: last time, I was writing about platforms. The iPad and the iPad mini are not two distinct platforms to be considered separately. They are two embodiments of one platform and should not be considered separately from one another.

Of course, they are different devices and they are not direct substitutes for each other. Each does a different job, and that's what I want to consider now.

We have learned over the past couple of years that, in school, an iPad can handle everything we've thrown at it. As a result, our school now looks significantly different to most other schools that have "a few iPads" scattered around. We don't have a computer suite any more. We don't have fleets of desktops or laptops to fall back on if the iPad can't handle the task - mainly because it's impossible to justify the cost of fixed infrastructure for such rare occasions.

So, where does the iPad mini fit in? I look at this from the point of view of the job we are hiring these devices to do. We are buying them to be a pupil's only computer for three years. Where does that lead us?

For some time now, I have used the following framework to think about 7"-class tablets:

A 7" tablet makes a great adjunct to a computer; a 10" tablet can replace it.

I first wrote that about the Google Nexus 7 and, having used an iPad mini exclusively since release day, I'm fairly happy to say the same applies to the iPad mini. My experience has been that I use the mini as much as I ever used my 3rd-generation iPad - and I take it with me to more places - but I've also noticed that my laptop has become more important to me.

Two years ago, the iPad mini wasn't practical. Today it is. Why? The cloud got good. Let me explain: in 2010, the options for fluidly moving between a laptop and an iPad on one task were pretty limited. In fact, it was initially near-impossible. Today, this is much, much easier. iCloud is working well and applications like Evernote are increasingly powerful on iOS. Two years ago, iCloud didn't exist and complex applications like Evernote and the iWork suite were not close to parity with their desktop counterparts.

If I had the budget to provide two computers to each pupil, those two devices would unquestionably be a MacBook Air and an iPad mini. Unfortunately I don't, and there's no way I could persuade people to give up their iPads, so we're going with the device we know can handle everything: the full-size iPad.

Another consideration is the internals of the iPad mini. Last year, the A5 architecture looked like it was history. It had a good run in the iPad 2, which is still on sale, but clearly the future looked A6-based. The iPad mini, being essentially an iPad 2 in a smaller case, changes that. The A5 architecture is going to be a major part of the iOS landscape for the foreseeable future.

That said, it wasn't the A6 processor or the 1GB RAM specification or even the retina display that led me to decide on the 4th-generation iPad. Basically, it's about buying the newest and most capable technology we can get. We're signing a three year lease on these devices and, given how fast the mobile world is moving, I feel we need to at least start our leases on the leading edge of technology. To start with older specifications and hardware - even if that device is brand new - is something I'm wary of. We have no roadmap for how things are going to develop so I intend to equip our kids with the best kit we can put our hands on today.

Refresh: The Platform

Recently, I started looking into our upcoming device refresh. We are already two and a half years into the iPad project at Cedars and the first three years will be over before we know it. It's crazy to think how quickly the time has gone.

Anyway, the first question I had to ask myself was: are we sticking with iPad?

The question that naturally arises from that is: what else could we use? Since 2010, we've seen a lot of "iPad killers" announced by the media, but precious few have actually made a serious dent in the iPad's position as the leading post-PC device. In 2010, choosing the iPad was a radical and bold step. In 2013, it seems like the obvious thing to do.

But is it? Where is the competition? What's changed since 2010?

Well, we're now in a world where there are a few more players than there were before. I've looked at all of them in some detail - some more than others, I admit, since my research budget comes out of my own pocket - but we're essentially looking at the following:

• iPad or iPad mini
• Amazon Kindle Fire
• Microsoft Surface (or some crappier hardware running the same OS)
• Google Nexus 7 (or some crappier hardware running a vendor-contaminated version of the OS)

Let's think about each of these competitors in turn:

Amazon Kindle Fire

The Kindle fire is getting better with each iteration. I don't particularly like the hardware but, as I keep writing, I'm not that interested in hardware. I'm interested in software. Amazon's fork of Android appears to be designed to do two things well: connect you to the Amazon content you've bought and show you more Amazon content to buy.

This is hardly a surprise. It's a conspiracy but quite an open one: Amazon wants to make money when you use the device, not when you buy it. The Kindle Fire is sold at below cost and Amazon makes it all back and more when you buy stuff from Amazon. As a consequence, the design of the system will necessarily be biased towards leading you down this path of seeing and buying things from Amazon. That's a product that a lot of people want - and that's fine - I'm just not sure it's the kind of thing I want to put in front of schoolchildren.

As educational technologists, we can't be naive about the business models behind these devices. The Kindle Fire is a vending machine for Amazon content.

Microsoft Surface

In recent Q&A sessions, I've been asked about the Surface far more than Android tablets. I guess this is because most schools are "Microsoft schools" - not that prior experience with Windows 7 will necessarily help you much when confronted with a Windows 8 tablet. There seems to be an idea abroad that if you get a Surface all your existing software will work with it. Almost nobody seems to have grasped the impact of the distinction between Windows RT and Windows 8.

My line on Windows 8/RT and Surface has been this: we know from a decade of trying that nobody wants the traditional Windows desktop on a tablet. Swivel-screen laptops have sucked for years and I'm willing to bet this trend continues. The only interesting thing about Windows 8 is the Metro UI and the only interesting thing about the Metro UI is whether developers will redesign their software for it. Given that Microsoft themselves haven't managed to properly redesign Office or Explorer for Metro yet, I'm not holding my breath.

Windows 8/RT may be a smash hit but the early indications are not promising. Microsoft Surface may be a smash hit but the early indications are not promising. I'm open minded about Windows 8 but I'm going to need to see some compelling evidence that Windows 8/RT is gaining serious traction in the marketplace. Microsoft is way behind on tablets and I don't have any need to root for the underdog.

Google Nexus 7

I've written before about the Nexus 7 specifically and I don't intend to rehash that discussion here except to note three major failings in the hardware for school use: the 16:10 aspect ratio of the screen; the lack of a rear camera on the device and the Nexus 7's inability to be connected to a projector. I just don't think I could sell my teachers on the idea that they'll go another three years without a camera or the ability to project devices on-screen.

There's also the software problem. Unless it's gotten dramatically better in the two months since I gave my Nexus 7 away, there simply does not exist the same high-quality productivity and creativity tools for Android tablets as exist for iOS. As before, I'm not saying they can never be made; I'm just saying they still don't exist. Where good Android tablet apps do exist they are, broadly speaking, near pixel-perfect clones of the iOS versions - so what's the advantage?

I'm also seriously concerned about the Android malware problem. There seems to be growing evidence that Android is the main target for mobile malware. This is hardly surprising, given the ability to side-load apps onto the device from anywhere on the web. I can't find any way on Android to lock out the ability to side-load, so we have to assume that this is something that some users will do. We know from decades of experience with Windows that such a model is problematic. I'm not prepared to deploy a platform where I have to routinely run anti-virus.

Whatever is working for Android in handsets simply isn't translating into traction for Android tablets. I have my theories but that's another post for another time. I can't say I have any great excitement about any Android device that isn't "pure Google". The track record on timely updates to non-Google Android devices doesn't seem to have gotten much better. Given the threat-rich environment that Android devices now face, being unable to get updates and security patches in a timely manner is unacceptable.

Most people's interest in Android in education is about price. Yes, the Nexus 7 is cheaper than an iPad but my retort is that it's a lesser device with fewer hardware features leading to greatly reduced usefulness in the classroom. Again, I come back to my core belief about tablet computers: the hardware is irrelevant except insofar as it allows you to have a great experience of software. You don't buy hardware in isolation; you buy it to run software.

The next gambit is usually that you can get £50 Android tablets for school. To people seriously arguing this I ask you: what computer are you using? I'm only interested in hearing about £50 Android tablets from people who are themselves using £50 Android tablets on a regular basis - which is, of course, 0% of the people arguing this. This is because £50 Android tablets suck.

iPad

And so to the iPad. It's worked well for us, so why change? Nothing's perfect and I promise you that nobody has a more comprehensive list of things that are annoying about iOS than I do. Yet, despite those niggles, I still believe the iPad is the first education computer worth criticising.

The hardware itself has held up well. Our failure rate has been in the low single-digit percentages over three years. There have been zero unfixable problems. Over three years, I have not once had to completely 'reformat' a device. The only time I get an iPad back in my hands is when it's physically broken. The track record on updates has been excellent; we have now had three major releases of iOS on the same hardware for free. The software ecosystem has grown into an incredibly rich and powerful set of tools. We routinely do things today that I was told would never be possible on iPad. We got a free cloud-based backup system that genuinely works. We got a powerful and free course management tool in iTunes U. iBooks Author is another powerful and free tool. iOS deployment techniques have advanced immeasurably since 2010.

So what's Apple's deal? There's no free lunch, right? Well, it's again simple and not a secret: Apple wants you to buy a new device on a regular basis. That's how Apple makes their money. Apple's cut of the App Store revenues don't even move the needle on their finances. Apple doesn't want to sell you apps or content; they'd much rather give all that away for free to make the hardware a more attractive proposition. Witness the number of free apps on the App Store. Commoditising your complements is a great strategy and it's working very well for Apple - although perhaps not so well for the app developers. Apple is a hardware company that makes great software.

Apple wants to sell you a computer. Amazon wants to sell you stuff while you use the computer. Google wants to sell your activity on the computer to advertisers. At the end of the day, it's all about what you can live with. We are now three years along the iOS road. It was the right decision in 2010 and I think it's going to be the right call in 2013 too.