Compared to, say, a cameraman in a war zone, tech writers have it pretty good. The worst injury we can suffer is a stiff back from an especially long product keynote. The biggest job hazard is having our bad predictions thrown back at us, either by readers, or, in my case, by myself.

Not all of my predictions for 2011 were awful. I suggested that Microsoft would unify its PC-tablet roadmap with Windows 8, though it would take several years.

I was talking up Bring Your Own Device 20 months ago, when BYOD was not yet a hot thing, much less a widely-used acronym. And I argued that with mobile devices, it's Management, not (Protection from) Malware, that matters.

Still, this column is me donning a virtual hairshirt, as I own up to all of my wrong trendspotting for the soon-to-end year.

Bad Prediction #1: Superphones would finally become a thing in 2011. As the name of a new category of phones succeeding 'smartphone', SuperPhone has everything going for it. It sounds awesome, heroic even, like it was created by some writer at DC or Marvel Comics.

Going by the rule of threes, I figured that the smartphone market had gotten mature enough that it would split into 3 tiers: SuperPhones on top (read my definition of their specs here), smartphones in the middle, and featurephones for the mass consumer at the bottom.

I think there were several reasons why I was wrong:

- Generally, the tech industry creates new jargon faster than a teenage girl changes her clothes. But it  also hates to interrupt a hot streak by confusing consumers. And smartphones were definitely on a hot streak in 2011, growing by leaps and bounds. Why risk confusing customers by pushing a new, vague category?
- SuperPhones were hard to categorize because they didn't really look all that different from smartphones. They weren't particularly thicker or larger-screened, nor did they all come with keyboards. Peoples' hands and pocket sizes aren't increasing; wearing a phone clipped to your belt was no more fashionable in 2011 than it was in past years.
So there was no way SuperPhones were going to become a category, unless we all became as tall as basketball players or if fanny packs become fashionable again. Wait a tic...
- SuperPhone was a term only embraced by the Android camp, and only a few manufacturers at that. Apple was careful never to use the term. That hurt, considering most consumers, if asked what phone models they can name, unprompted, can probably only name the iPhone.

According to Google News, there were just under 400 articles using the term SuperPhone, compared to 110,000 using smartphone. Will SuperPhones finally catch in 2012? I'm not betting on it.

Bad Prediction #2: Adobe Flash would become relevant on mobile. As Apple zigged away from Flash, I figured that the rest of the industry, especially the Android camp, would zag towards Flash. That happened initially - both Samsung and RIM made a big deal out of the fact that the Galaxy Tab and PlayBook were Flash-compatible.

But users never really seemed to care all that much. There was a plethora of non-Flash content for them, including, crucially, YouTube videos. Developers moved to other platforms. By November, Adobe decided to cut its losses by effectively ending Flash for mobile.

Bad Prediction #3: The market would naturally come around on the BlackBerry PlayBook. After the negative reviews of the PlayBook at its release in April and the predictions of doom for RIM, I railed against the "premature prognostication for the sake of being first."

I took the contrarian stance that what was once overrated had now become underrated, and implied basically that I believed that PlayBook sales would surprise everyone over the next six months.  After all, RIM ships 15-million+ BlackBerry phones overseas every quarter without any U.S.-based pundit seeming to realize it.

Well, I was surprised - just by how much RIM apparently over-ordered and how many PlayBooks sat on retailers' shelves. And when Amazon shipped the Kindle Fire that was oh, 90-95% identical to the Playbook, but only $199, RIM had to react by cutting the Playbook's price 60%.

While that may look like a desperation move, I think it will have the desired result of getting 1-2 million Playbooks into customers' hands by early 2012. That will bring the developers and then start the whole virtuous cycle of apps-attracting-customers-attracting-developers-attracting-customers, etc.

Already, there are already 4,200 apps for the Playbook. When Android compatibility comes in the spring, that number effectively grows by a factor of 90 or so. In other words, I don't think I'm outright wrong on the PlayBook as I was with Flash; I just haven't been proven right, yet.

Bad Prediction #4: The rise of low-cost 'FeatureTablets.' Just as inexpensive, low-powered featurephones vastly outsell smartphones worldwide, I figured that featuretablets from no-name brands like Coby running Android would become huge with consumers. These would be sub-$200 tablets that would sport single-core ARM chips, poor battery life, few apps and bad touchscreens.

These would be the cheapie devices that Best Buy or Frys would advertise as doorbusters to get people in the door before trying to upsell them to iPads or Galaxy Tabs.  Then came Amazon with its $199 Kindle Fire, This is a state-of-the-art dual-core device with a large selection of content.

By bringing a mid-range tablet at a low-end, below-cost price, Amazon has guaranteed that it will sell boatloads of Kindle Fires and make it almost impossible for anyone else to undercut them. I mean, there will be tablets at $100 price points, but I doubt if they will ever form the bulk of the market as cellphones/featurephones still do. At least not in the developed world.

Bad Prediction #5: Android app downloads would grow exponentially. In July, I created a chart taking historical app download data from Apple and Google and tried to extrapolate growth trends based on that.

Based mostly on exponential growth for the last two quarters, I boldly predicted that Android app downloads would blow past iOS by the end of this year.  In retrospect, it's obvious to see that I was too optimistic.

Google hasn't given a count in the last several weeks, but according to AndroLib's real-time counter, 7.2 billion Android apps had been downloaded by December 29 - far lower (and more reasonable) than the 40 billion I thought was possible.

However, my App Store projection, based wholly from Cupertino-announced data, appears nearly spot-on. Apple's App Store is likely to get 10 billion downloads in 2011, leading to a cumulative end of 2011 total of 20 billion downloaded apps. That's pretty close to the 23.7 billion downloads I predicted for the end of January 2012.

I still think it's likely that Android app downloads will overtake App Store downloads. There are just too many devices out there now. Will it happen in 2012? Perhaps.

The Sybase Unwired Platform has generated a fair amount of interest this past year from enterprise developers. But to be honest, there's been frustration, too.

The frustration comes not from the SUP technology itself, which made a huge leap with version 2.0 this May with the launch of two major new features: 1) the ability for tens of millions of Web developers to code for mobile in languages familiar to them such as HTML5, Javascript and CSS; 2) the ability to create 'hybrid' apps that run inside a Web browser (for easy porting between, for example, between Android and iOS) while still offering the rich UI of a native app.

Nor was it the frustration of developers who felt trapped in the uneasy situation of feeling reliant upon and competitive with SAP in enterprise mobility. Any experienced enterprise developer has probably dealt with co-optition before. Microsoft, for instance, is one of the biggest sellers of applications on its own Window platform. And Google this summer bought Motorola, putting it in competition with its many Android handset and tablet partners.

While SAP has brought out 30+ mobile apps of its own, it has also taken pains to reassure and woo developers. For instance, the new SAP Store for Mobile Apps will only reap 15% of app revenue - just half of Apple's take from its App Store. 

SAP's public goal is to build an ecosystem such that partner apps outnumber its own 4 to 1. It pretty much hit that at SAPPHIRE Madrid, showcasing more than a hundred apps in development or nearing completion from partners.

Rather, the biggest complaints centered around these things:  1) difficulty for developers, especially those working at smaller or one-person shops, to get a trial copy of SUP 2.0;

2) the difficulty at getting trained at SUP right away;

3) the cost of SUP compared to other development platforms that have adopted freemium or other similar low upfront-cost models.

Developers can be a noisy bunch. So suffice to say that the frustration was palpable.

During deep-dive sessions at the SAP Influencer Summit last week, SAP and Sybase executives revealed plans to address these concerns. Thanks to SAP Mentor Dagfinn Parnas for being the first to tweet out that in 2012, SAP is planning to introduce:

1) a hands-on trial version of SUP for developers;

2) an SAP center to directly support SUP developers with learning resources, developer forums, access to experts, and more;

(This is not to diminish the ongoing efforts of people at Sybase like Stanley Stadelman, Loren Corbridge and others who have been working hard to produce materials such as the most recent "MBO Best Practices" white paper as well as the SUP Dev Blog.) a new, less-expensive runtime license of SUP.

I don't have any more details at this point. But the reaction from developers in the SAP ecosystem has been positive.

The developer center is "excellent news," tweeted Australian developer, John Moy. "It will certainly improve developer engagement...Whatever gets the technology in the hands of developers easily and freely is what matters."

Moy also applauded the SUP run-time license. "If true, this improves [the] value prop" for companies who are only interested in running apps built upon SUP (which requires an SUP run-time license), but are not planning to develop their own custom apps on SUP. The latter still requires a full SUP license.

Another developer, Kevin Grove, tweeted to me that this "sounds like win-win for all. And it reaffirms that SAP listen to the #SAPCommnet dialogue."

At the SAP Influencer Summit on Tuesday, the company touted its  mobile milestones and achievements for 2011, while promising major  upgrades and improvements to its mobile software.

Below are a bunch of slides from the keynote presentation by Sybase  Executive Vice-President and head of SAP Mobile Applications, Dr. Raj  Nathan, along with some of his quotes, my commentary, and reaction from  the 200 or so analysts and journalists tweeting from the event (#SAPSummit) in Boston.

Slide decks for other presentations by co-CEO Jim Hagemann Snabe, SAP Executive Board member in charge of technology and innovation, Vishal Sikka and President of Solutions Go-to-Market, Sanjay Poonen, and other videos from breakout sessions later today and Wednesday are available at the SAP Virtual Summit page.


(Click on the slides to open larger in a new window, and then right-click to save to your computer.)

As I hinted in my preview blog to the Summit, SAP had some ambitious internal goals to meet this year.  There was some momentum at my employer Sybase, but there was no  guarantee that the integration wouldn't make things grind to an ugly  halt.

Well, SAP has met its goal of bringing out 30+ line-of-business mobile apps.

And SAP has performed on the sales side, too: 350+ new customers, 17.5 million total end user seats sold.

Visit the latest Enterprise Mobility newsletter from Sybase to view links to the customer videos below - Tommy  Hilfiger, Boston Scientific, Charite Berlin, CSC, Simba Dickie, Tasnee  and more.

(I realize that seats sold does not equal active workers deployed and  using today, as my headline above implies. Which is why I clarify it  now.)

SAP/Sybase are, according to Nathan, responding to the mobility  trends above, including the need to manage the entire lifecycle of both  consumer and enterprise apps, improve the user interface of apps, and  accomodate and woo developers who want to use HTML5 technology, and  more.

Basically, enterprises want their mobile device management (MDM) software to do more. Strong security is already becoming commoditized  in the iOS arena. The only forward for non-platform vendors is to  augment with the features above, says Nathan.

For SAP and Sybase, this means bringing Afaria and Sybase Unwired  Platform ever closer together, until they become a truly unified  platform.

It also means adding the ability to manage non-enterprise apps  and letting enterprise developers build lightweight Web apps ever  faster. At that point, what you have is a greater platform that does  what some are calling Enterprise Mobility Management.

Raj also talked up the new SAP Store for Mobile Apps and the roadmap above. Read this for more slides and an interview with Usman Sheikh, the vice-president at SAP Ecohub in charge of the new SAP Store for Mobile Apps.

SuccessFactors will no doubt be the star of tomorrow's SAP Influencer Summit. The future is always more fun to talk about. But if you want to hear about victories in the last 12 months, may I suggest that enterprise mobility is your topic?

When SAP bought Sybase in May 2010 for $5.8 billion, there were plenty of doubters. Some of that was based on misinformation - that Sybase was still only a database company, not, as in reality, a leader in multiple emerging markets including enterprise mobility, analytics and mobile commerce - held by those who really should've known better.

But there was also legitimate doubt about the "force multiplier" effect promised by co-CEO Bill McDermott from taking Sybase's software to SAP's massive customer and partner ecosystem.


There was also uncertainty whether SAP and Sybase together could deliver the mobile apps it promised to bring to market.
Indeed, what SAP focused on at 2010's Summit was the opportunity - how mobility was a key ingredient of "the real-time enterprise."
[To see what else was discussed at last year's Summit, read this.)
My prediction is that tomorrow we'll hear all about the tangible proof of progress.
Some of these details are out, some will be entirely new, such as:
- hundreds of new corporate customers, including Tommy Hilfiger, toy maker Simba Dickie, Computer Sciences Corp., hospital Charite Berlin,  Boston Scientific and more.
- many millions of customer seats;
- the hundreds of SAP-related mobile apps now or soon available, including 30+ built by SAP;
- continuing improvements in the Sybase Afaria mobile device management software and the Sybase Unwired Platform for developing enterprise apps;

[To keep up with what's happening inside the SAP mobile ecosystem, check out the December edition of the Sybase Enterprise Mobility newsletter.]

Keynote speakers at this year's event include co-CEO Jim Hagemann Snabe, CTO Vishal Sikka, Sybase executive vice-president and head of SAP mobile app development, Dr. Raj Nathan and Global Solutions president, Sanjay Poonen.

Held in Boston, the event begins Tuesday Dec. 13 at 8 am Eastern Time and runs til the end of business on Wednesday, Dec. 14.

You can keep up with all of the news in real-time by following me on Twitter @ericylai tomorrow morning, or tracking the Twitter hashtag #SAPsummit, or watching the tweets and blogs as they roll in at the SAP Newsroom.

So Google Android framework engineer Dianne  Hackborn responded Thursday evening to the accusations leveled by  ex-Android intern Andrew Munn.

Here is Hackborn's rebuttal. Thanks to readers of my earlier blog on this topic, "Why I still have faith in Ice Cream Sandwich overcoming Android's Fundamental Lagginess," Derek Morr and Tom Van Doorslaer, for pointing it out to me, and also sharing their thoughts.

Morr's comment was pithy: "In short, the intern got it wrong."

I agree that is the basic gist of Hackborn's post,  though of course her technical explanation makes it clear that reality  is in shades of grey rather than black and white.

I'll try to boil her post down into ten statements without hopefully losing all of the nuance.

1) On accusations that Android doesn't prioritize key threads such as UI rendering, "this is outright wrong."
2) Android lets app and screen rendering be prioritized as default or background. User interface threads normally run at the main default. Application processes running in the background are forced to run in the background.
3) Background threads are bunched together using a "Linux facility called cgroups." Collectively, these threads are only allowed to take up 10% CPU utilization maximum.
4) There was a 'foreground' priority thread in the original Android but it was abandoned b/c it turned out not to smooth UI rendering all that much.
5) Android uses the two sets of priority threads as part of its goal of creating a 'sandbox' architecture that separates all apps, incl. 3rd-party ones, for security reasons. This, says Hackborn, differs from iOS's design which didn't originally accommodate 3rd-party apps.
6) It is true that there is not a separate real-time thread just for screen  rendering, as in iOS. However, this is no magic bullet.
7) Rather, setting up a separate thread just for drawing the UI in real-time would not have been worth it,  due to a bunch of complex reasons I don't quite understand, though  Hackborn seems to be hinting it is related to Android's overall open app  architecture.
8) Android only "recently" began to use hardware acceleration for drawing inside the UI. That's because hardware acceleration isn't as simple as making your graphics chip handle the UI.  It takes a lot of memory and multiple processes to manage the graphics  chips as "most mobile GPUs still have fairly expensive GL context  switching."
9) "There are of course many things that can be improved in Android today, just as there are many things that have been improved since 1.0. As  other more pressing issues are addressed, and hardware capabilities  improve and change, we continue to push the platform forward and make it  better."
10) But it's no technical piece of cake to  enable iOS to touch-scroll at a smooth 60 frames per second, either.  Hackborn quotes a comment to that effect from an outside developer.  "Based on this statement I don't see any indication that there is  something intrinsically flawed about Android in making lists scroll at  60fps, any more than there is in iOS."

So what's your takeaway? The key of course is how  you weigh Hackborn's admission that no, there is no real-time thread  dedicated specifically to UI rendering, but yes, there are clear ways to  strongly prioritize UI rendering over potentially-interfering  background tasks.

Also, as I point out in my post, "Why I still have faith in Ice Cream Sandwich overcoming Android's Fundamental Lagginess,"  this may all be simply the equivalent of an arcane philosophical  argument in an advanced college seminar, except between two software  engineers.

The real-world evidence shows that the new  Ice Cream Sandwich update DOES appear to cut the herky-jerky behavior of  Android, vis-a-vis the glowing reviews for the Samsung Galaxy Nexus  smartphone running Ice Cream Sandwich, and the less-glowing reviews for  the ostensibly more-powerful (4 cores!) Asus eee Transformer Prime  running the existing Honeycomb version.

The Blue Screen of Death still casts a negative halo around Windows despite basically disappearing from PCs a decade ago, after Windows XP arrived. Similarly, I wonder if Android will be unfairly dogged by a reputation for a sluggish user interface for years even if version 4.0 'Ice Cream Sandwich' effectively solves this nagging rendering problem.

Before I go on, here's the backstory for those of you who didn't see the dueling Google+ blog posts that appeared earlier this week between Android engineer Dianne Hackborn and ex-Android-future-Windows-Phone intern, Andrew Munn on the subject of Android's laggy rendering compared to iOS, Windows Phone 7, and other modern smartphone OSes

(Also! After I posted this blog, Hackborn responded with a strong rebuttal of Munn's assertions. See my summary here.)

Now, I had always assumed that sluggishness in Android was symptomatic of quality issues that are the inevitable result of Google's open ecosystem strategy: 1) until Google's acquisition of Motorola, its inability to optimize hardware with OS since they are built by different companies (unlike the iPhone or iPad); 2) the lower bar for entry for apps into the Android market that results in more potentially buggy apps.

Then I heard about Ice Cream Sandwich (hereafter known as ICS), and how features such as hardware acceleration (the ability to let the GPU more efficiently do the rendering work) would supposedly smooth out Android's stuttering UI.

In that context, Hackborn's post was curious. Rather than confirming that ICS' hardware acceleration will improve Android's touchscreen response, she focused on correcting "technical misinformation" and explain that earlier versions of Android, including the version 2.2 Gingerbread installed on most phones today, actually already use some hardware acceleration.

With its latest update, has Android fixed its nagging screen lag problem? Early evidence strongly suggests 'Yes.'

Being an engineer, it's not in Hackborn's DNA to be market-y or even humblebrag. Which is admirable, but it left the uneasy impression that rendering lag would continue to be a huge problem.

That's exactly what Munn argues. Citing several online sources plus ex-colleagues at Android, the 3rd-year engineering student explains that Android's lagginess is fundamental to its architecture. Basically, Android was originally designed to be used on regular laptops with a keyboard and a mouse. It doesn't prioritize real-time tasks like screen rendering with a separate, protected thread. As a result, the UI can't handle constant input and screen redraws, as fingers on a touchscreen tend to create.

This is the main reason why Android isn't and can never be as smooth as iOS or Windows Phone 7, which do prioritize real-time rendering, says Munn. Only a fundamental rewrite, the kind that would break hundreds of thousand of existing Android apps, will fix this nagging issue, he warns. Otherwise, "Android UI will never be completely smooth."

Munn is a great writer, and combined with his ostensible insider knowledge and his no-BS truthtelling stance, he makes a compelling argument.  The fact that neither Hackborn nor anyone else at Google has responded to Munn after two days is also telling.

But let's put this into perspective:

- Munn worked on Android full-time as a college intern for all of 3 months.

- He admits that he was neither on the Android framework team responsible for UI rendering, nor has he even "read any Android rendering source code."

I am going to go out on a limb and say that Munn probably — hear me out — actually maybe, may in fact, not be the supreme authority on Android rendering that we'd all wish him to be.

Also, here's the thing. After reading Munn's post, I decided to see what the real-world evidence, not just the theory, says.

I found all of the reviews I could for the only phone running Ice Cream Sandwich today - that would be the Samsung Galaxy Nexus - to see if the UI was indeed smoother than before. Below are excerpts from reviews that explicitly commented on the UI:

Engadget: "The Galaxy Nexus is definitely one of the fastest Android handsets we've  ever played with. Everything feels snappy, everything looks fluid --  Ice Cream Sandwich isn't just a new version of Google's mobile OS, it's  what happens when Android hits the gym and becomes lean and mean."
Boy Genius Report: "Android 4.0 is fast. Extremely fast. Scrolling  between the five home  screens (please, please let me customize the  number of home screens) was  silky smooth with practically no lag  whatsoever."
GigaOM: "Android 4.0.1 on the Galaxy Nexus feels like it actually borrows more  from Windows Phone 7.5 than iOS, at least in terms of aesthetics, and everything in general seems to work better and smoother."
The Verge: "As far as phone performance is concerned, however, the Galaxy Nexus  feels blazingly, stupidly fast to me. Touch response is excellent on the  phone — everything reacts quickly to your movements. Homescreen  scrolling was snappy, moving into and out of apps was instantaneous,  swiping through long lists was stutter free, and web browsing (even on  heavy pages like ours) was super speedy. ..It's obviously a combination of great hardware  and great software, but the Nexus is probably the tightest feeling,  snappiest Android phone I've ever used. It's awesome."
Gizmodo: " The user experience is extremely fast and fluid. Scrolling  around webpages is quicker and smoother than any other mobile browser  I've used (and with all of its new enhanced features, I would call ICS'  version of Chrome the best mobile browser out there)."
PhoneArena: "Using an Android phone has never looked or felt better...It  just flies no matter if you're swiping  through homescreen, scrolling  long lists, webpages, or opening and  closing heavy applications. The  occasions where we noticed a slight  slowdown (like when switching to  landscape QWERTY) were so few, that  we'd go as far as to say that the  Samsung Galaxy Nexus is one of the  fastest smartphones out there  today."
SlashGear: "Although the TI dual-core is capable of 1.2GHz in  the Galaxy Nexus (the chip itself is offered at up to 1.5GHz in other  devices) during our testing it spent less than 5-percent of its time at  that speed. In fact, over 85-percent of the time the CPU was running at  just 350MHz, with the remainder somewhere around the 700-850MHz point.  That aggressive throttling – and the fact that the phone never felt slow – is testament to Ice Cream Sandwich’s frugality and refinement."
Laptop magazine: "Very fast performance: On our tests, the Galaxy Nexus was super  smooth and responsive. Featuring a 1.2-GHz dual-core processor, the  phone instantly launched the App menu and let us jump between apps (such  as the browser and "Fruit Ninja") in a second. The Galaxy Nexus also  delivered excellent graphics performance when we played "Riptide GP." The action never stuttered, and we could make out detailed reflections in the rippling water." "All that hardware is brought to bear on the centerpiece of this new   phone, the first smartphone shipping with Google’s Android 4.0 “Ice   Cream Sandwich” (ICS) operating system. It represents a giant leap in   usability for this popular operating system. The first thing I  noticed  was its ability to smoothly scroll down long lists, the  complaint I’ve  had against Android-packing phones from the beginning. Finally, you can  scroll up and down a Google+ stream and slide down  lists of emails with  smoother response, although not all apps I tried  enjoy that  butter-smooth scrolling yet."

That's nine reviews saying ICS on Nexus is butter-smooth. Meanwhile, I found only 3 reviews that mentioned a hiccup-y UI:

CNET: "Though we were hoping that it would be different, the Galaxy Nexus  still  has that slight laggy effect that we've seen on other Android  phones.  Indeed, you'll notice it here when scrolling through lists. It  is better  than we've seen on previous models, so it doesn't ruin the  touch  interface, but you do notice the difference when switching from  an iOS  or Windows Phone 7 device"
TechCrunch: "Usually the Galaxy Nexus hums along, but I’ve had one or two moments  where where it reminded me of my Nexus One, pausing at odd moments and  apparently ignoring finger taps. This hasn’t happened often — and Google  says at least one of the issues I saw is a known bug that has already  been fixed on devices customers will receive. So we’ll see how it  performs over the coming weeks."
Time magazine: "For the most part, the Galaxy Nexus’s 1.2-GHz dual-core processor makes  for a fluid experience, but I did encounter some instances when the  phone briefly ignored my taps and swipes–a usability glitch that’s  pretty much unknown in the Apple world."
So, a 3:1 margin in favor of ICS!
To make this as rigorous as possible, I decided to look at reviews of the new Asus eee Transformer Prime, too.
The Transformer Prime is a 10-inch tablet that uses a quad-core Nvidia Tegra 3 processor but runs on Android 3.0 Honeycomb.
While the Prime's screen is much larger than the Nexus, their resolutions are basically the same: (1280x800 for the Prime, 1280x720 for the Nexus).
Both use chipsets based on the Cortex-A9. But the Nexus's TI OMAP chipset relies on the 3-year-old PowerVR SGX540 graphics chip with 4 cores, while the Asus tablet's Tegra 3 chipset sports a brand-new, 12-core GPU.
In other words, will the Transformer Prime's much more powerful rendering engine overcome the handicap of an older, stutter-y version of Android?

The evidence would say no. Six reviews mentioned the laggy UI, one review was neutral, and one said the problem had gone away.

CNET:  "While the Prime's IPS screen was immediately clear and sharp when   first we powered it on, it was the screen transitions that really   impressed us. The first time we tapped the Apps button, we were treated   to a noticeably higher frame rate transition than on any previous   Android tablet. We hoped this fluidity would carry over to apps  like Marvel Comics, but that was not the case.  Reading a comic through  the app on the iPad 2 is still a considerably  smoother experience, but  this may have something to do with specific  optimizations of the iPad  app. We can't be sure, however."
Laptop: "As with every other Android tablet we've used, we noticed occasional  moments of slowness in opening a menu or launching an app, but there  were fewer of those moments than on other devices."
The Verge: "While Nvidia says the general UI and OS should feel faster, I have only  found that to be the case some of the time. While you can see in the  video that swipes across homescreens are swifter, the waiting that  frequently occurs when opening menus or toggling between apps on  Honeycomb tablets isn’t completely gone."
Engadget:  "That said, we were sorry to still see some occasional stutters and   hiccups from time to time, instances where the device would hesitate for   just a half-second or so before responding. There are three  performance  modes that are easily selected between in the pop-up  settings menu, but  even on its highest we couldn't get it to be a  consistently smooth  operator. They're the kind of stops and starts  we've seen on just about  every Android device to date and it's a bit of  a shame that even four  whopping cores running at 1.3GHz can't do away  with them."
PC Magazine: "You won't see the blinding speed  when you're poking around the main UI or  some of Google's apps, as  they're occasionally nonresponsive, although  screen transitions are a  bit more fluid than on other Android tablets."
SlashGear: "What  you’ll find here is that the two tests we’ve  got running here, the  Prime outdoes the original Transformer by quite a  bit, but the iPad 2 still leaves both of the Android tablets [Transformer and Transformer Prime] in the dust."
PCWorld was neutral:
PCWorld: "Beyond the obvious boost in CPU performance and gaming, the benefits of the quad-core Tegra 3 wasn't apparent in all activities. Touchscreen swiping was smoother, for example, but in general, navigation, multitasking, and in-app experiences didn't feel dramatically faster."
Only GigaOm gave a hesitating thumbs up:
GigaOM:   "The ASUS Transformer Prime tablet, powered by Nvidia’s quad-core  Tegra  3 chip, seems to be running silky-smooth in this first look."

So to recap, Doctor of Science style:

Hypothesis: Even if handicapped by a lack of real-time rendering thread, Android's screen lag can be reduced to the point that, for all intents and purposes, it matches iOS or Windows Phone 7 in perceived smoothness.

Experiment 1: Does ICS on Galaxy Nexus run smoothly?

  Result: Yes, by an overwhelming 9:3 margin, according to professional reviewers.

Experiment 2: Does the more-powerful hardware of the Asus Transformer Prime overcome the stuttering Honeycomb UI?

Result: No, by an even more overwhelming 6.5:1.5 margin.

Conclusion: Throwing faster processors at Android isn't what solves its stuttering problem; its improvements in ICS. Improvements that appear to have banished the lag problem, even without having apparently rewritten its rendering engine, as Munn argued must be done.

With more than 85% of Android users on either Froyo or Gingerbread, this would be great news and potentially the end of their most nagging, annoying issue.  Here's the list of devices and when they'll be eligible for official installs of ICS.

They include virtually all dual-core devices released by HTC, Acer, Asus, LG, Motorola, Samsung and Sony this year.  I'm eager to see if ICS does indeed improve screen lag on those devices, too.

That would be definitive proof of my theory that even if it's not the most elegant solution, Google's engineers are doing enough workarounds to the real-time threading issue so that it won't matter anymore.


A quick plug: I'll be covering the Mobility portion of the next Tuesday's SAP Influencer Summit. Basically, my parent company will trot out top executives to speak to top market analysts about next year's strategy and roadmap. If you're interested in what we're doing in enterprise mobile apps, development and management, follow me on Twitter at @ericylai on Dec. 13 starting 6 am PST or watch the hashtag #SAPsummit.

Mobile Years are like Dog Years: highly accelerated. Case in point: when Ford Motor Company started thinking about Bring Your Own Device back in May 2007, it figured that demand for laptops would outstrip that for smartphones or tablets. And the mobile devices that workers picked would run Windows Mobile, Nokia's Symbian operating system and Palm, predicted Ford experts.  

You may chuckle. Indeed, when Randy Nunez, a senior network engineer for Ford, shared this anecdote during his excellent presentation last week at IDC's mobileNext Forum in San Francisco, it drew a laugh from the audience.

But when Nunez received that fateful e-mail from his CIO asking to investigate BYOD, Microsoft was the chief alternative to BlackBerries. iPhones were only several months old at the time, while Android was not yet announced.

The moral? Deploying mobile devices in a large company is "a little like painting the Golden Gate Bridge," he said. By the time you think you're done, the environment has changed and "you have to start all over again."

There were many great presentations by enterprises at mobileNext, many of them touching on BYOD.

(To read about how 3/4 of enterprises today have BYOD policies, click here. Or what to do after you've instituted BYOD, read here.)

But Nunez's talk about Ford's experience stood out for its frankness as well as its wise insights. Ford's motivation to look into BYOD as a reaction to two trends: Consumerization of IT, as well as the new habits of 20-something Millennial employees.

To re-emphasize how forward-looking Ford really was in 2007, a quick search of Google News shows zero stories about BYOD, at least related to mobile devices (there were stories about Bring Your Own Dog and Borrow Your Own Dancer, but I digress).



Ford's Randy Nunez: Deploying BYOD in a huge company like Ford is a bit like "painting the Golden Gate Bridge."

To investigate BYOD, Ford created a cross-functional team consisting of managers from the IT, legal, HR, accounting and other departments.  "We tried to quantify the risk versus the reward," Nunez said. Naturally, each member of the group brought their own concerns about security, laws and user support. Combine that with the fact that there was no hard and set deadline, and the BYOD task force couldn't agree on the way forward.

So it created a report outlining the benefits, risks and costs, and sent it to senior management, to make a ruling.  That not only broke the deadlock, says Nunez, but also got strong buy-in from the executives. With that, Ford moved forward in the second quarter of 2009 - exactly two years later - on a program it calls eMail on Personally Owned Devices, or ePOD.

ePOD is not the most cutting-edge BYOD program around. Workers who are accepted into ePOD are only allowed to check their Ford e-mail, as well as their personal calendars, contacts and task lists from their personally-owned devices, on or off Ford campus.

"This is for the casual user, not for someone for whom mobility is critical for their job," said Nunez. Also, ePOD users need to be tech-savvy, as Ford doesn't provide any technical support. Calls to Ford's help desk about ePOD are referred to a Web site where employees offer each other technical support.

ePOD is more about employee convenience than about saving them money. Employees pay for the cost of the device and all subscription fees. Ford only bears back-end costs such as servers or software licenses.

For Ford, these limitations were the result of the financial reality it and other American automakers were facing several years ago.

"It was a cost-effective way of getting into the game," Nunez said.

ePOD has still proven popular with Ford employees. Ford has 2,700 employees in the BYOD program. 800 use BlackBerries (there are another 3,000 employees who use BlackBerries that are owned and paid for by Ford).  While BlackBerry support debuted in 2009, Apple support began a year later, in September 2010.

Today, 1,900 employees use iPads, iPhones and iPod Touches. These devices are secured using Mobile Device Management (MDM) software.  Nunez says the BYOD program has been considered a success at Ford. Executive support has been key, he says, as has the narrow scope of the BYOD program, which has resulted in a low TCO.

"We didn't give the keys to the kingdom away," he said.

What is in the future for Ford's BYOD program? Ford plans to begin supporting Android devices this month. That rollout was slowed due to concerns over Android security, he acknowledged. Deployments of apps remain in limbo, he said, partly because the company wants to make sure it can secure the apps as well as have sufficient network connectivity on campus to support that heavier load.

Ford's recent financial success also has the company weighing whether to offer employee stipends as reimbursement for their personal devices, Nunez said.

I've never worked in IT, but I imagine that the relationship between system administrators and their programming counterparts is often tense, with each camp jockeying for resources and alpha-dog status.  

So I hope I'm not opening a can of worms by suggesting that IT could learn something from their developer pals.

In the past decade, many developers have embraced the Agile style of development. Done right, teams of Agile (also known as Scrum, Extreme, Lean, and a myriad of other names) developers work together better, write simpler, better code, and release new versions faster than through traditional styles of development such as Waterfall.


Agile's 'Release Early, Release Often' mantra harmonizes perfectly with Web 2.0 and mobile apps, though it can also accelerate and ease development of large applications, too.

Agile's other tenets - flexibility, bottom-up collaboration, people over process - can also help IT managers wrestling with how to manage mobility. Why? And, how? Let me explain.

1) Hardware lifecycles are shorter. Old-school IT departments are used to decade-old Unix servers or laptops that you hold onto for 4-5 years. They'll be shocked to find that with smartphones and tablets, a 2-year lifecycle is typical, and 3 years probably the utter maximum.

That's not because mobile devices necessarily break faster. Devices that are equipped with consumer-grade protective cases can last just as long as ruggedized ones if employees are motivated.

Rather, innovation is faster. It took several years for PCs to go from single-core to quad-core. But just within this year, we went from a single-core iPad to the quad-core Asus EEE Pad Transformer Prime.  

2) Platforms rise, fall and change more quickly, too. As Sybase CTO Irfan Khan pointed out recently, Android wasn't even around 4 years ago. Now it's tops in smartphones. Meanwhile, it took almost a decade before Windows XP was finally, recently displaced by something else.

Or take this comment by a speaker at the IDC MobileNext Forum. His organization, a large car manufacturer, started planning a Bring Your Own Device policy back in the middle of 2007. The platforms they figured they would need to support included Windows Mobile, Symbian and Palm. By the middle of 2009 when they actually launched the BYOD program, those 3 platforms had been whittled down to just one: the iPhone.

The speaker's comments drew muffled laughter from the audience. They weren't laughing AT the speaker, though. They were laughing at themselves, perhaps remembering once-beloved WinMo and Treo devices that they hadn't thought of in what seems like forever, despite it being just several years.

Now, it's possible that the market will eventually shake out and settle onto a few platforms. And that improved Mobile Device  Management software will make the process of updating and patching OSes much more seamless  and easy than it is today.

But until that day comes, IT managers need to be able to react to changing hardware and platforms, and be able to deploy them more quickly.  So out would be companies like Federal Express, which took a year last time to roll out a ruggedized mobile device. Granted, that was to deploy and train 100,000 not-very-technical workers worldwide. And that that ruggedized device is so expensive that FedEx hopes to keep it around for six years.

An example more in-line with contemporary trends would be my parent company, SAP, which rolled out 3,000 iPads to its salespeople in just 6 weeks using Sybase Afaria MDM software.

3) Workers have much more say today. The Consumerization of IT trend brought BYOD to enterprises. BYOD is only the most visible example of how the balance of power has shifted away from IT departments. Command-and-control is out; partnerships with the business side are in.

That's analogous to Agile's emphasis on bottom-up, organic collaboration and privileging people and relationships over rules and processes.

A nice example of this attitude would be Rick Peltz, CIO at real-estate brokerage, Marcus & Millichap.

"The real estate brokers are my clients. I like to hang out with the guys in the industry. And I will put my job on the line, if I believe in something," he told the audience at IDC's mobileNext.

As a result of listening to the brokers, Peltz went ahead and asked AT&T to build an iOS and Android mobile app for them. The app hosts profiles of Marcus' 1,200 agents, and enables property buyers/customers viewing those profiles or property listings to immediately text, e-mail or call one of Marcus' agents.

And, mind you, Peltz went ahead and had this built without any explicit request from business-side executives, and purely out of his own budget.

Peltz is so focused on his end-users - the agents - that he is even being Bcc:ed on every e-mail inquiry sent to the brokers, just so that he can better understand their needs.

Or take the recently-published Mobility Manifesto (from Sybase), and its 'Universal Declaration of Workers' Mobile Rights.' Both of these examples parallel Agile Programming's emphasis on bottom-up, organic collaboration over formal, inflexible, top-down planning.

And all of these can serve as useful role models for progressive IT departments wondering how best to structure themselves in the post-PC era.