Understanding RIM's tablet platform app strategy

Yesterday RIM reported their quarterly earnings. The results were mixed to slightly negative and the shares were down 10% in after-hours trading.

I’ll work through the smartphone market data at a later time but for now what I want to focus on is RIM’s strategy which really means understanding RIM’s intentions or their approach to the market. For that we have to go straight to the source: what management actually says. Trouble is, management often speaks in a jargon that is unfamiliar to people (sometimes unfamiliar to anyone outside the company).

Strategy analysis involves translation. So here follows the interpretation of Jim Balsillie’s remarks during the conference call (transcript sourced from

What Balsillie Said My interpretation
First of all, what we announced is Gingerbread. This is not Honeycomb. I don’t know what the number of Honeycomb apps is, but it’s not very many. Whereas Gingerbread they’ve got lots of them. RIM will support Android apps in the PlayBook tablet. As there is not one Android, I need to clarify that RIM will not support Android tablet (Honeycomb) apps but Android phone (Gingerbread) apps. So RIM’s tablet will support Android phone apps even though our tablet does not actually have a phone in it, but requires tethering to a BlackBerry for cellular data.
You’ve got the volume of the handset apps, so if you’re looking for the tonnage of apps, or some kind of long tail stuff, you’ve got it. This support means there will be many apps including apps that are not very popular. We don’t like supporting unpopular apps. We like a few good quality apps but this approach allows us to offer all kinds of apps.
At the end of the day, people are going to want performance. However, many apps will not run smoothly.
You’re just not going to get things like gaming and multimedia, you’re not going to get the speed going through a VM interface. If you want content, or Flash type stuff, or you’re looking at AIR-type, evolving web-type assets, that’s what you’re going to do. Games and graphics will not run well through emulation. For processor intensive experiences we will have native execution options or browser apps.
There’s no compromise here. You’ve got the tonnage of apps. And you’ve got the performance. Do I think the tonnage is overplayed? Yes. By supporting emulation of Android apps we can offer a large catalog of so-so apps while having a small portfolio of high quality native apps.
But if you think it’s about having a couple hundred thousand apps, there you go For people who value large selections, they can go to one store.
Do we believe it’s about super high performance? Yes But for those people who value quality, they can come to our store.
I’ll just offer that observation that today broadly in the world without any specific comment on us or our competitors, there are multiple ecosystem patterns that need to be considered Beside emulation and native we also think there are many other ways to deliver applications.
Do we believe it’s about full web fidelity? Yes. For example, Flash.
These are concepts that were really relegated as not technically possible, which we’re doing here. This is a no compromise environment. Because of Moore’s law, Flash is becoming viable on a mobile device.
If you want to work on Android, great. Do we think people will want to migrate web assets? Yes Then there are browser apps.
Do we think they’re going to want super high performance native assets with the SDK? Absolutely We will have our own native SDK.
You think they’re going to want to use their Flash based stuff for an offline Flash/AIR type environment? Yes And support Flash.
I’m just not interested in these sort of religious application tonnage issues.  I really think we put that issue to bed. We have a new platform but catalog sizes can no longer be used against us.
And if you think the whole world’s going to want to develop for Gingerbread, fine. Do I think that’s going to happen? Then why is there a different environment for a tablet? And you know about the performance issues and you know about the app volume issues, cause it’s tough. And that’s why QNX matters Google’s ecosystem is fragmented, Android apps are not able to rely on competitive hardware and there is too much competition. That’s why developers should target QNX (which is unfragmented, built on fast hardware and there are few apps on it).
Is this stuff going to go more in the browser and HTML 5 and more native? These are going to be strong trends. But if you want these app players for different VMs — and don’t forget we have 25,000 BlackBerry 6 apps Browser apps will increase in value and we also have legacy BlackBerry apps.
So, at the end of the day, we believe this is going to be about performance. It’s going to be about enterprise greatness. But we believe in a future with native applications running on QNX.
Things like multi-threaded capability, symmetric multiprocessing. We believe it’s about an uncompromised web. We believe it’s about enterprise security. True multitasking, not with suspension — and that matters because you’re going to want to run these things in the background. Because QNX is a robust operating system with PC-like performance. It also supports a good browser.
But I’m out of the religious war on tonnage, which I’m delighted. However, before QNX takes off, we will provide access to catalogs of apps which are built for other platforms.
  • Alberto

    Poor Rim, where went all of his talent or even his common sense. At the end, Rim tablet users can chose between sub par mobile apps or non exclusive web and flash apps. Why any body will buy this product. I think this strategy will kill 90% of high quality native apps

    • It depends on their native SDK and the experience given by native QNX apps. If they're miles better than Android, which isn't hard to do really is it, it might be a decent enough stop gap. Some decent in-house written apps might swing it for many too.

      Do they have a native SDK yet though? Last I looked it was either Adobe AIR (Ugh) or web based technologies.

      • Alberto
      • David

        Ask OS/2 how well supporting someone else's platform helps the long term app situation.

      • 2sk21

        So true! Having good support for Win 3.1 was one of the reasons why OS/2 never took off.

      • Ravi

        And having such good support for DOS apps is why Windows never took off…
        And having such good support for 16-bit apps in Windows 95 is why Win32 never went anywhere…
        And having such good support for PowerPC apps is why Apple's transition to Intel didn't work out…

        Oh, wait a minute…

        On the other side of the scale, we've covered OS/2, having good support for x86 apps on Alpha Windows NT didn't help Digital at all and Linux binary compatibility was the beginning of the end for some x86 Unixes.

        Compatibility is a strategic tool, like any other. Applied judiciously it enables new opportunities. Applied thoughtlessly and it destroys them.

      • No your're wrong.

        Dos and windows market was both in Microsoft control. Microsoft migrated words and key software to windows and then wait quietly to kill dos. Whatever pc sold with dos or windows 95 it was the same for Microsoft. In the end, one day, Microsoft told companies to stop selling pc with dos but with windows 95 (dos 7)

        Apple was ppc and intel Mac at the same time. They killed one to make the other. You didn't have choice in the end. And beside whatever the CPU it was still money for apple

        It was NEVER a good idea to support an ADVERSE platform in your own product. It killed a lot of products. OS/2, sgi irix and others.

        Linux never had that problem. It is unix and it's not commercial. It plays with total others rules.

        Linux was cherished and wished by people like me. We wanted to install it, support it and work for it.

        Totally a different story than commercial platform as os/2 and now rim tablets.

      • kevin

        In each of the cases where it worked, the same company was in control of both OSes, and was clearly moving from one to the other.

        When it didn't work, it's because one company tried to include an OS that was from another company, where the OS wasn't already declared to be on the way out.

        I think we all know which category fits RIM's attempt.

      • addicted

        And DOS, like Android, was available on MS competitors' platforms…Oh wait.
        And 16bit Windows apps, like Android, were available on MS competitors' platforms…Oh wait.
        And PowerPC Mac apps, like Android, were available on Apple's competitors' platforms…Oh wait.

        Those examples are completely irrelevant because Android is not just available on competitors' devices (unlike your examples), but they are first class citizens there. Customers will look at the Playbook, and be like "99% of the apps here run better on the Samsung….let me get the Samsung instead".

      • Ravi

        You are, in fact, getting the history wrong. Apple is the only case where a competitor did not have high-fidelity support for the existing applications being transitioned to a new platform.

        Microsoft was not the only DOS vendor when they managed the transition to Windows. There was also DR DOS, which was considered by many to be technically superior. And as the later settlement revealed, the failure of DR DOS was also driven by Microsoft's illegal actions – not just because of Digital Research's missteps in the marketplace.

        And as for 16-bit Windows, we started by talking about OS/2 which was Microsoft competitor that, arguably, ran 16-bit Windows apps better than Windows itself. I'd also highlight that, at the time of the OS/2 – Windows split, Microsoft and IBM had equal rights to what had been built so far. Consider the counterfactual: if OS/2 had won we'd be talking about how Microsoft obviously shouldn't have tried to take over IBM's turf (because IBM had more control over the "PC standard" or something like that).

      • Ravi

        IBM's problem with OS/2 was not the support for Windows apps… their problem was that they couldn't keep up with Microsoft's treadmill as they kept adding new features, so their bootstrapping value proposition of "A better Windows than Windows" couldn't be maintained.

        RIM, arguably, has some of the ingredients to make "A better Android than Android". QNX is a real-time operating system and they might be able to use that to provide a smoother user experience than Google can on top of Linux. RIM has also been licensing J2ME from Sun (and now Oracle) for a long time. If they put the pieces together correctly (pickup all of Google's good ideas and code while using all the Sun IP they want), they might be able to pull off a better mobile JVM, too. And to top it off, they are a hardware manufacturer, so they could ensure the whole package was integrated.

        Do I think they're doing that? No. I see too much public flailing for them to have any sort of plan they've actually thought through. However, as Nokia did, RIM has an interesting opportunity if they want to go after it.

        Do I think they're doing this? No. They're flailing to much to have

      • No

        IBM could follow the treadmill. It was just flawed by concept. You have not to support your opposition.

      • kevin

        At the time, most people thought IBM had some of the ingredients to make "A better Windows than Windows". IBM was after all software experts as well as in control of the hardware.

        In any case, RIM's problem with QNX will likely be that they won't be able to keep us with Google's treadmill as Google has been pretty quick at adding features, and until RIM bought QNX, one would be very hard-pressed to say that RIM was a software company.

      • addicted

        So you think that IBM, a behemoth, was unable to keep up with MS, but RIM, a relatively tiny company with little cash and shrinking profits will have no trouble keeping up with the Google machine?

        Something doesn't seem right there….

      • Ravi

        Well, suppose RIM were clever and managed to hire Cyanogen and some of his favorite collaborators? Still think they'd be so obviously outmatched? I'm not saying what RIM needs to do is easy, I just think it is far from impossible.

        And the "treadmill question" assumes without evidence that Google and RIM have to be adversaries. I don't think that's inevitable at all. At bottom, why would Google care whether RIM, Motorola, Samsung, HTC or someone else makes the most successful Android tablet? Isn't Google's bigger concern that successful Android tablets are made? Doesn't RIM have something different to throw against the wall in terms of their existing customers and enterprise relationships? Does anyone have a strong reason why Google wouldn't embrace RIM the way they tried to embrace Nokia? Yes, RIM might have their distinctive changes (using QNX instead of Linux, a different JVM and support for legacy Blackberry apps), but is that really so different from Samsung's TouchWiz and HTC's Sense (and the Flyer's stylus / Scribe API) or whoever's making "3D" tablets and phones? And even if it is a step further is that step so far that it would overshadow the value of having RIM as an ally?

        I should be clear here. I agree that RIM is a basket case, they're too confused and disorganized to be anything else and that they're not going to succeed. Deciding to work on Android support in secret and then haphazardly announcing it (without even trying to build a relationship with Google) is just flat-out stupid. I'm merely arguing that their embrace of Android isn't *self-evidently* foolish. RIM's problem isn't that their strategy is fundamentally broken, it is that there is no evidence that they are making the sort of moves and building the sort of capacities required to execute it competitively.

      • ______

        If your point at the end of the voluminous prose you have written on this page was just that there's more nuance to supporting competing platforms than is implied here then:

        1 > Even if you were right, the point is really weak and you could have stated in 10 words what you took hundreds for. But
        2> Pt. 1 is moot because you are wrong. RIM has run a closed system, like Nokia. Nokia jumped from the frying pan in to the fire. RIM may one day follow suit, but there are no indications that they are ditching QNX for Android. They will not be just another Android-based HW manufacturer and hence have no meaningful relationship with Google. Therefore, Google is a competing market/OS. Supporting a competitor's apps (and badly, at that) isn't a winning strategy. There's no logic and no precedent. This will not succeed. Embrace of Android is foolish, either in whole or in part (for an integrated company like RIM)
        Finn schoolboys, extreme cold, pissing, pants – those words were part of a poignant insight provided by a Finn recently. RIM would do well to take heed.

      • WaltFrench

        RIM is apparently offering VM-like "boxes" in which java and Android apps can run. They clearly INTEND to offer a native SDK, but I will take a stab and say it won't be above beta quality for a year. Another stab: the Green Box for Android will interoperate with the native OS only badly. A single app in background will require keeping the whole VM in active memory, hogging the resources for a user with multiple browser windows and say, a music player. That same background app will not have a good way to notify that same user that she can review the results of its background work.

        More importantly, developers will recognize the Green Box as an interim-only solution, and will put minimal effort into tuning their apps for the Playbook's larger screen, UX, etc. I've been thru several similar interim steps — one now, in fact — and conversions (as has Apple) and it's a lot of work for a fair amount of heartbreak. Consider how badly the whole classic/carbon/cocoa transition figured in the destruction of a previously strong Adobe relationship (tho I won't say fully caused it).

        And most importantly, I see zero enthusiasm from Google for this move. They had the opportunity to bring in RIM as a full Android partner, and instead announced that Honeycomb would stay locked up in the current Open Handset Alliance. Google has a huge interest in ad revenue, IP, strategic control of the platform, trademarks and partnerships that I perceive as threatened by this move; I cannot imagine how RIM will ever get Android updates within a year of release, if Google doesn't first sue them or otherwise block them from using Android.

        So this looks to me a desperation move by RIM because they lack that native SDK. I see so many ways that it could end up badly, the least of which is making them look like incapable fools.

  • sandro

    Wow. What a strategical mess.

    • HTG

      Your telling me… this is what UBS thinks about the whole deal..

      "We also remain somewhat unclear about RIMM’s PlayBook strategy. While the focus appears to be on enterprise customers, RIMM will be launching the product via several retail channels (thus targeting consumers) at a price point at par with Apple’s iPad 2. For a challenger product, with an arguably considerably smaller ecosystem, the notion of price parity with the leading product in the market is baffling."

      I am completely stumped by everything RIM is saying and doing (Note, RIMM is the stock code for the company)… Also note that the above quote is polite analyst speak in a published document – you can only imagine what the analyst is saying directly to clients! Baffling = TAOOTFM!!!!

      BTW there is no such words as 'strategical'… it is 'strategic mess'.

  • Pingback: RIM bringt Android-Apps aufs Blackberry Playbook. —

  • Not surprising at all. On one hand you have this new post-pc play that can make your offerings redundant. On the other you are saddled with the evolutionary cruft of your own making.

    Very few have managed to transform base metal into gold. And casts what Apple has accomplished in very favourable light.

    Truly testing times for any leader/organization, almost feel sorry for RIM.

  • Imon

    Wow. Seems like Rim's problems comes from an incoherent top down management approach, catering to a vocal minority of enraged 'web geeks'.

    Thank you for the translation. I was almost otherwise, lost.

  • Jon T

    There is only one other CEO whose utterances I have read and heard which measure up to that degree of goobledegook, and that is Ballmer. The speak their own variation of English, which destroys the language.

    Quite shocking that people can rise to such positions and not express themselves, simply, and straightforwardly.

    Glad you bothered though Horace, because when I read it I couldn't make head nor tail of it.

    • stephenreed

      Some of what Horace translated is intentional marketing obfuscation by Balsillie. E.g. "At the end of the day, people are going to want performance." translates to "However, many apps will not run smoothly."

      Rim management is speaking to industry-savvy analysts, so that explains the use of jargon. RIM is trying to portray itself in the best possible light, so that explains the obfuscation.

      The decline in the RIM stock price occurred because of what the analysts understood from the conference call – thank's Horace for sharing that understanding with us.

    • gslusher

      One would hope that someone in RIM would have the guts to show a video of Balsille's talk to him, along with taped reactions. Maybe he'd get the point. Perhaps the Board should insert a clause in his contract that he is not to speak in public, as, every time he does, he damages the company. Maybe someone should anonymously drop a copy of Carmine Gallo's book, "The Presentation Secrets of Steve Jobs" on Balsillie's desk.

      There's this "fad" (long-running) of CEOs and founders speaking for their companies. VERY few can get away with it. Obviously, Jobs does a good job, as did Dave Thomas for Wendy's. (I'd also count Harlan Sanders, as he founded KFC, but continued as the public voice for years after he sold the company.) Gates was so-so. Ballmer is a disaster and Balsillie makes Ballmer sound like a master speaker.

  • FalKirk

    On the one hand, you have to sympathize with RIM. The Xoom runs all the Android phone Apps and 16 and counting native Apps and they're being crucified for their lack of native Apps. Imagine how much worse it would be for the PlayBook to enter the market with only a handful of native Apps. The howls from the Analysts, the pundits and no-nothing commentators (like me) would have reverberated from Wall Street to Main Street.

    Having said that, this multi-platform strategy raises some serious questions:

    1) How will this affect the PlayBooks performance?
    2) How will this affect the PlayBooks battery life?
    3) How will this affect the PlayBooks customer experience – ease of use of the operating system?
    4) How will this effect developers of native programs for the Playbook?

    I would think the answer to all of those questions would be "negatively." And always, always, always, RIM will be forced to answer the question: "Why should I buy the PlayBook instead of the iPad?" I'll withhold judgment until after the PlayBook hits the market and is reviewed. However, the lack of a strategic direction expressed by management in their earnings call does not bode well for the PlayBooks future.

    • CndnRschr

      But the Xoom plays Android 2.3 apps too – they just run like crap. My bet is that they'll equally run like crap on the PlayBook.

    • Hamranhansenhansen

      XOOM doesn't run native apps at all, except for the Web browser. Android apps all run in a virtual machine, they are not native. They are Java, not C.

      The PlayBook native C SDK is the bigger story, but RIM buried it with the Android story.

      • WaltFrench

        RIM buried it cuz it's nowhere near ready.

        Google was able to create one in a remarkably short time; props to their engineering talent. Still, it took them 5 years to get one with smooth and/or 3D graphics, decent keyboarding, etc.

        RIM was in denial in 2008 still, and has published nothing resembling a strategic roadmap of a unified tablet/handset OS. I suspect it's worse now that their handset business is even more relegated to low-end, low margin sets; very hard to span the > 10:1 range in CPU power, whereas Apple only has to cover about 6:1 from the iPhone1 to the iPad2.

        In this regard, RIM's situation is very much like Nokia's: they need to keep the low-power stuff going, but the growth path from there is incompatible with what they need yesterday at the high-end.

      • WaltFrench

        The Dalvik VM is one sweet piece of software engineering. Many apps — even those that don't ask the system functions to do most of the heavy lifting — become indistinguishable from native.

        However, on Android, that JVM is tightly integrated to the OS. On Playbook, whatever VM RIM uses will have to be adapted, with some impedance mismatch, to the QNX OS.

    • Stripes

      (the following assumes the limits are the technological issues, and that the execution by RIMs programmers is very high quality, actual world results may fall short, but will not exceed)

      The android apps are already java, so there is no technical reason the playbook would run them any slower then a normal android phone. The OS (QNX) even supports most of the APIs hat Limux does (and QNX is very fast, or at least it was when I used it a decade ago). However supporting multiple application environments especially at the same time will REALLY put a hurt on RAM usage. They will either need to put more RAM in then other tablets, suffer in performance, or invent a very interesting new technology.

      Customer experience could suffer, if each set of apps operates very differently less of the learning will transfer. If each store is independent it will be harder to decide where to look for an app. If there is one merged store it will be harder to decide which app might be slower or faster or just work "weird". Also apps the alter things at the system level are not likely to translate, so putting swype on might change the keyboard for android apps but not change the keyboard for AIR, web, or old BB apps, or the new QNX apps.

      Battery life may not suffer too much, more RAM requires more power, but not a huge amount. However slow apps tend to not just annoy you and waste time, while you are watching them think they are busy sucking your battery dry. It isn't universally true, but it is frequently true, enough so to make it a rule of thumb.

      Devloper direction will be murky. We have seen poor guidance from the top, ther is no reason to beleve thatRIM wil, produce useful clear and "true enough" statements about which APIs and languages to use for what. Worse some of the APIs may be incomplete "oh you can only get at the built in SQL from AIR not QNX (as a made up example)". If you look at Apple's track record with carbon, cocoa, and cocoa-java over the last decade you can see some of what RIM faces, except RIM has many more APIs in the soup and unlike Apple it has control over NONE of them, and NONE of the languages involved.

      The android strategy may still Bethesda best available to blackberry, but it still isn't good.

      • WaltFrench

        “The android apps are already java, so there is no technical reason the playbook would run them any slower then a normal android phone.”

        I can count several reasons. I won't say they're definitive, but I'd wager that several will be apparent.

        First and foremost is that the approach puts another layer between the Android OS software and the hardware. With the Android Green Box virtual machine running alongside others programs, RIM will have to supply a supervisor that mediates between the Green Box and who has right to draw to the screen in a certain location, who gets to save data in a certain spot of disk, etc. This ain't kid stuff, and the fact that I cannot imagine this as RIM's strategic direction, they seem unlikely to invest in the VM interface.

        The second reason is that all java is not the same. Android 2.2 and better includes a very fine piece of software engineering that runs java (apparently, other non-native code, too) up to 3X the speed of earlier releases. If RIM incorporates the Dalvik VM, they would apparently have to load on many other Android features that are sewn together with it.

        This is a non-trivial engineering challenge at the same time RIM is simultaneously struggling with their completely independent handset OS transition and with some future native SDK, not to mention the little challenges such as getting Flash acceptable on the device, which bested Moto, a more focused firm.

        This looks like Nokia in an alternate universe where they instead tried to do everything themselves, then put in 4 (!) different crutches when they found they were years away from success.

        PS: LOVE that auto-correct on Bethesda!

    • davel

      As the other reply noted more VM's require more cycles and RAM. This impinges the platform.

      But the insistence on flash will kill the battery all by itself.

      I find it interesting that a year after Apple threw its hat into the ring there are still only a handful of competitors. Shipping that is. In the meantime Apple is on rev 2 of its design.

      • CndnRschr

        I'm not a programmer but with the various VMs involved and the need to run launcher apps to contain the non-native apps, the user experience is bound to be both more complicated and flakey. If a user launches an Android 2.3 app that uses Flash and launches an AIR app that uses Flash, there is a possibility the respective VMs are insulated so the device has two independent instances of Flash code running. Perhaps this can be safe-guarded by low level architecture, but this predicament is true of other emulators due to memory space protection, etc. RIM might get around it by making the VMs mutually exclusive (only one VM at a time) but that could also make the experience of switching apps slow and disruptive.

        I do doubt that PlayBook users will run much Android stuff though. As Jim Balsillie rather revealingly admitted, the capability was added to guard against the perception of no apps. RIM better hope that Android isn't the primary app source for the PlayBook. Put wheels on it and they could claim it could run skateboard functions too….

      • Hamranhansenhansen

        They also insist that mobile users want to run background computation, as if they are running encoding jobs, or are part of a cluster. Which is like saying "no battery management!"

        The only thing we know about PlayBook battery life is 1) QNX is not made to run on batteries, and 2) a prototype they were showing said "4.5 hours remaining" with almost a full charge. Those estimates are usually optimistic, so it will be interesting to see what happens when battery tests are done. Remember, a 7-inch screen has 48% of the screen area of an iPad, so there is no excuse not to get 10-hour iPad-type battery life under real world use.

      • CndnRschr

        Yes, but a 7" tablet also has 50% less space for a battery than a 10". Looks like the PlayBook is a "chunky monkey". They certainly are not playing the Samsung game of going after the thinness of the iPad2 (and failing).

      • unhinged

        The reason why competitors are struggling to catch up is because Apple is close to a decade ahead of them. Apple was working on the iPad and forked the project into the iPhone, which was released in 2007, so I think we can reasonably assume a two year lead time for the phone and thus an even earlier start for the iPad.

        The competition are actually achieving great things given that they mostly came from a standing start when the iPad was released, but they look poor in comparison to Apple. This is no accident.

  • My.translation: I am high as a kite!

    • WaltFrench

      I would seriously consider hallucinogenics and/or narcotics in his place, as well.

  • sphvgt

    Another great post!
    I like the translation part. It's really impossible for me to understand the left column.
    The translation make understand why their shares where down.

  • I think you misuse the term emulation (and all of it's negative connotations).
    As i understand it, RIM have promised to include a java virtual machine to run the Android apps. This is the same system as Android itself (RIM won't use 'Dalvik' but presumably other JVM's can achieve similar performance). RIM will need to mimic the android API's so the apps inside the JVM can talk with QNX underneath, but there's no particular reason for this whole approach to be slow.
    It might have a horrible user experience (iphone apps on the ipad is really the best you can expect) but there's no reason for this to be slow.

    • asymco

      What is a better word? (One that is understandable by people who are not developers.)

      • Sander van der Wal


        The Android environment is hosted on QNX, next to the PlayBooks native UI. The emulation bit is due to the fact that the Android UI is (mostly?)implemented in Java, and Java runs on a virtual machine.

        Android could have been written in C and still be ported to run on top of QNX. That would loose the emulation, but it would still be there next to the native UI.

      • WaltFrench

        “Hosted” is not bad although I prefer apps that run “in a compatibility box.” That emphasizes the difficulty of communicating in and out of the different environments, say when an Android app needs attention but is in the background. Android per se provides notification services, but those might be very inconsistent with what the user is doing in her native browser, for instance.

        BTW, a good share of the low-level Android OS *IS* written in C. Wikipedia says,

        The Android operating system consists of 12 million lines of code including 3 million lines of XML, 2.8 million lines of C, 2.1 million lines of Java, and 1.75 million lines of C++.[22]

    • handleym

      "RIM won't use 'Dalvik' but presumably other JVM's can achieve similar performance."

      For those scoring this on technology rather than business strategy, this is FAR from true.

      Writing a functioning JVM (ie one that simply interprets byte code) is easy, but adding the JIT smarts that give it speed is far from easy (not to mention that some of the various techniques that can be used for this are probably under patent). Put this together with the fact that this seems something of a last minute decision by RIM, and it seems highly likely that, yes in fact, this JVM will run a lot slower than Google's JVM.

      (Yeah, yeah, for the pedants in the crowd. The software being discussed in not literally a JVM.
      A JVM interprets java byte code, which is a very specific byte code designed and controlled by Sun, now Oracle. What Android's environment does is to translate Java into a DIFFERENT byte code, which is then run by a piece of software that functions like the JVM, only interpreting this different byte code.
      This matters for legal purposes, and to some extent as a technical matter, but as a large scale issue, both a JVM and a google byte code interpreter are the same sort of thing.)

      • WaltFrench

        I would imagine that RIM has to bring in significant parts of the Android libraries into their Dalvik equivalent. Apps would invoke standard Android functions (APIs) such as "write 'Abc123' at location (100,240)" because otherwise the app would have to know too much about each different handset.

        RIM has said their APIs will be “highly” compatible. Inevitably, however, the Android apps will end up slinging pixels at not the actual hardware, but a buffer controlled by QNX, to make sure that an app running inside the compatibility box (thinking it has ownership of a certain screen area) doesn't smush all over the user's current browser session. That will be a performance penalty, especially for graphics as the untranslated gobbledygook said. Flash would appear to be unworkably awful in the Android Green Box. Many games will suffer performance penalties that the Playbook's otherwise superb hardware won't be able to overcome; I'll wager that only one of the Playbook's CPUs can be devoted to any compatibility box.

      • Ravi

        RIM isn't adding a JVM at the last minute. They've had licensed J2ME support for a long, long time and they needed compatibility with their legacy apps. Obviously handling Android apps means they're changing the scope of what they're asking the JVM to do, but a lot of the core infrastructure was already there.

      • handleym

        Except that, Ravi, as I pointed out, Dalvik is NOT standard Java byte code. That's how Google gets around having to pay various Sun license fees. The Java language part is the same, but the byte code (and hence the byte code interpreter) are different. Having a standard Java byte code JVM won't help them.

      • WaltFrench

        That's a major component of Google's claims that it doesn't need to pay various Sun license fees.


        I am NOT following the Oracle suit against HTC at all closely, but I seem to recall that after the byte code specifics, a big part of the complaint was about Android's use of communication features, which the Playbook would sensibly want to lock out anyway.

      • Ravi

        I don't agree with how you're looking at this. They talk about re-compilation,so I don't think they're proposing to run Dalvik bytecode unmodified. And don't forget, Google isn't the only company that can write a good mobile JVM. As I recall, in the days before Froyo (when dalvik finally got a JIT), at least one mobile JVM provider (and possibly more) touted how much faster their JVM was compared to Google's.

        In RIM's place I'd worry more that their previous-generation JVM (and initial PlayBook plans) probably didn't need as much native integration (or JNI support) as Android does (because Blackberry phones supported a fairly limited flavor of J2ME, as I recall). That's certainly a fixable problem, but depending on their existing JVM architecture that could easily be a significant stumbling block.

      • handleym

        "RIM won't use 'Dalvik' but presumably other JVM's can achieve similar performance."

        You appear to be right, technically. Looking around, as far as I can tell, RIM will NOT in fact run unmodified Android apps out the box. Developers will have to "port" them to Blackberry, though it's not clear how whether this is simply a "recompile" to standard Java byte code, or requires more extensive app modifications.

        Either way this would seem to mean that
        – yes, the technical issues I raised, of them not having a high quality dalvik engine are moot
        – to be replaced with the WTF issues of massive customer and developer irritation. People who hear "it can run Android apps" and then discover, "well, yeah, it sort of can, in some technical sense, for some limited set of Android apps" are going to be mighty pissed off. And are Android developers going to even bother to make the ports to Blackberry, given the ongoing pain that is being a Blackberry developer?

      • Justin

        I don't think it's a last minute decision. A playbook from December 2010 was shown running an Android app. Video's on youtube if you're interested. The commentator accidentally mentioned that the Playbook will 'support android apps' which is where the speculation came from—although it seems RIM was trying to keep that part a secret as an 'ace up their sleeves' before launch.

    • Hamranhansenhansen

      He is not saying they will suck more on PlayBook than on Android. He's just saying they will suck in the exact same way they suck on Android. Apps running in a virtual machine do not perform like native apps. They will not be as powerful as native apps. Android does not offer 3rd party developers access to the native C API. The native C API on Android is closed. That is part of why iOS developers laugh when Google talks about how open it is. Android is not even open enough for them to port my C code there from Mac, Windows, DOS, PlayStation, Xbox, Wii, and so on. That is part of why iOS apps are so much better than Android apps.

      • Ravi

        Android's C/C++ APIs are most certainly *not* closed and they're getting notably more open with Gingerbread.

        I'll add that you're radically underestimating the power of VM technology. You almost certainly use at least one high-end VM every day (and quite possibly many such VMs) and you don't even notice.

  • neilw

    I think there's a partly sensible (though risky) strategy struggling to get out from behind Balsillie's incoherent ramblings. Playing from behind as they are in the tablet space, RIM felt that they could not ever get traction if they had to start from scratch on the app front. So they gave quick and dirty access to a large library of pre-existing apps, even if they won't necessarily be great on the Playbook. Sort of (but not completely!) the way Apple launched the iPad with the ability to run iPhone apps, even if they don't always look great.

    In their best case scenario, the available of a large library ("tonnage", LOL) creates short term viability, so they're not dead before they even get started, and over time the draw of higher quality apps leads developers to produce native apps for long-term competitiveness.

    In the worst case scenario, it's an incoherent mess that produces and inconsistent and unappealing user experience, leading the market to form a bad impression of the platform and hence abandoning and/or ignoring it.

    I lean towards the latter, but we'll see how it goes. In the meantime, note to RIM: don't let Balsillie talk in public.

    • I'm glad someone's brought up the analog of iPad's "emulation" of iPhone apps. If RIM would state clearly that they consider the native SDK the way forward and frame Android (and Flash, for that matter) as a legacy transition mechanism, I might even admire the approach.

      • David

        There is some significant differences..
        1) iOS apps were all native
        2) Devs could/did/can make universal apps that work on both
        3) iPad had around FIVE THOUSAND native iPad apps day one
        4) Provide a built in migration path from iPhone and iPod touch
        5) Clear first mover advantage
        6) Proven technology

        It goes on and no. There is, IMO, no conparison

      • 7) Apple's in the US, RIM is in Canada.
        8) "Steve" is two letters longer than "Jim"

        100) the iPad is a better product

        None of that is the point. The PlayBook is a new platform, and it is a sensible strategy to provide a migration strategy from a successful platform to help establish the new one. If Balsillie had the courage to say outright "our platform is better than Android, but until it is mature we want our customers to benefit from the app tonnage today" we wouldn't be having this fun picking apart his PR-speak.

        The real story is that Android's openness enables competing platforms to use this strategy. Only Apple could leverage the iPhone ecosystem to bootstrap a tablet, but anyone can leverage the Android ecosystem to do the same.

      • davel

        I agree. It is a new platform and the corporate world rarely has the courage to say the emperor has no clothes.

        I take issue with how open Android really is. Just recently Google closed off Android 3.0 to everyone. I like their openness.

        As far as leveraging the open Android platform as easy as pie. Why did all these great hardware engineering companies not flood the market pre-Xmas with these wonderful gadgets if all they had to do was minor tweaking to the open easily extensible code?

      • Today is a good day to revisit this, in light of the BusinessWeek piece about Android's "openness" myth fading…

        If RIM hadn't made this move, would Android fragmentation alone been enough for Google to play their hand?

      • CndnRschr

        Unless you want to leverage Honeycomb which is not available to RIM (yet). RIM has added the capability to use a phone OS on a non-phone tablet. It will be competing directly with Honeycomb Android tablets.

        Moreover, the risk to RIM is that developers focus on the "tonnage" and continue to spit out Android code making the catch-up period for native PlayBook apps even longer. This was the problem Apple faced with Windows (before the Intel switch). Software emulators were developed that allowed some applications to run but they were awful compared to native Windows. Still, the software houses didn't have to port in order to sell to Mac users so they didn’t bother.

        The PlayBook may either become the Rosetta Stone or stage actor of tablets. Lots of versatility but is there a character of its own?

      • Yeup. The danger of being just a carrier for other platforms is exactly why RIM should have a clear message about the long-term focus on QNX and the benefits to developers of PlayBook-native deployment.

        Balsillie seems to be gesturing in that direction ("You’re just not going to get things like gaming and multimedia, you’re not going to get the speed going through a VM interface."). Unfortunately he, and RIM by extension, suffers from an almost Balmerian inability to articulate a clear point of view.

      • David

        I think that that is the point. It isn't reasonable to compare Apple supporting iPhone apps and Rim's dart throwing. I think it won't work because it almost never has. Many have tried support other systems including the competition and none that I know of have succeeded.

        An analog would be migrating from BB to Playbook. What they are doing, I think, simply won't work.

      • WaltFrench

        “…anyone can leverage the Android ecosystem to do the same.”

        Interested in a wager?

        It is unusual in this business for major rollouts to be announced in quarterly earnings calls, to say the least. This announcement should have been in Toronto's finest auditorium, with Eric Schmidt or Sergei endorsing this approach.

        Instead, Google announced a couple hours earlier that Honeycomb, the Android Tablet OS, would be out of RIM's reach.

        Google has a huge arsenal of weapons to prevent RIM from undercutting current Android Handset Alliance partners, and if I were Moto, I would be outraged if Google didn't use it. Already, "closed" is the new "open." The only restraint I see on Google is Oracle, since apparently RIM licenses the Oracle mobile java and would do so for the Playbook; Google would not want Oracle to focus its assault on them, so may simply make it impossible for RIM to get upgrades and help with their Android effort.

        If Jim B had the courage to say their platform was not ready, it would get the attention and interest that HP has whipped to an anti-frenzy. HP keeps churning out those highly profitable ink cartridges but RIM is seeing itself cut off from all the high-margin stuff and the investors are getting nervous about this little fantasy of developing a homegrown OS in 3 years while the competition is 6-10 years ahead of you.

      • Hamranhansenhansen

        It's interesting that you think supposedly "open" Google wants to limit Dalvik use to their own partners, that Google wants to stop RIM from using Dalvik, while on the other hand, everybody except Microsoft uses Apple WebKit, and not a peep from supposedly "closed" Apple. Apple simply outcompetes the users if its open source projects, including Google (WebKit in Android and Chrome) and Adobe (WebKit in AIR and Creative Suite). RIM has Apple WebKit at least twice on PlayBook: once in the PlayBook browser, and again inside Adobe AIR. And yet, no challenge from Apple on that.

        HP is also not only an Apple WebKit user (it provides not only the browser but the whole user interface in TouchPad and HP phones) but also HP is one of the biggest Apple Bonjour users, in their printers.

        Another wrinkle is I have heard the supposedly "closed" PlayBook is a better HTML5 client than Android. HTML5 is the most open API ever created, so it will be interesting to see if Apple and RIM provide the #1 and #2 HTML5 environments. (Apple is already #1.) You would think with Google's "open, blah, blah, blah" and Web heritage, they would have done a better HTML5 environment on Honeycomb.

      • WaltFrench

        In the days when IE was everything, Apple suffered greatly from having a substandard browser. I think WebKit depends on others' contributions, even if it helps them compete, to maintain Apple's internet capability near the forefront.

      • Hamranhansenhansen

        iPhone was a new platform at one point, and did not emulate any other platform. Apple offered sandboxed HTML5 apps and built-in native apps on a product that was so good, 3rd party native app developers lost their minds pleading for a native SDK, which they got a year later.

        OS/2 also ran Windows and DOS apps and that didn't work out well for them.

        Running BlackBerry apps makes sense, but Android? It is marketing only.

  • pk de cville

    BOTTOM LINE: It seems like RIM is playing the hand it's been dealt; no, make that the hand it picked for itself while gorging on easy to have profits and sleeping at the wheel.

    Now, its time for the crash and crashin' it will be.

    They will ship a system that EMULATES Android (Good luck on battery and UX on that.), and introduces QNX. For the sake of 'tonnage' (thinking of a garbage barge here), they are risking turning off people at the intro when the Walt Mossberg review hits: "QNX promising, Emulation Sucks"…

    Has anyone thought of the dangers of over promising and under delivering? (see Newton)

    Seems like that's the gamble they're taking. It might work if the emulation is excellent, but we'll wait for the reviews.

  • I have red that playbook only works with RIM phones, it seems to me that it is some sort of tablet for the club not for the masses.
    The multiple programming options are some sort of this type of speaking: don't be afraid to remain sticked with us, even if it is a niche device, playbook wont have niche app because we will support android and flash and web phone apps, even if they will give a poor user experience.

    I have alse red that RIM is not pushing to much support toward developers that have to pay a lot and certify themself with a lawyer before having access to the SDK and develpment tools.
    That's to say: we want only a few of apps from big developers no more, the rest will be emulated just to have numbers to show at meetings.

    Poor strategy in my opinion, for an expansion to new client for sure, and perhaps even for convincing existing phone customers.

    • gslusher

      According to RIM, the PlayBook can tether to some (maybe not all) BlackBerries to get cellular access. One problem, though, is that this is the ONLY way to get email from RIM's servers onto the PlayBook. WIthout tethering, the PlayBook can't do the one thing that RIM keeps harping about–reliable, protected email. Thus, pretty much the only market for the PlayBook is current BlackBerry users. In contrast, NONE of the people I know who have bought an iPad had an iPhone. (Several who have iPhones have said that they were putting off buying an iPad until ATT/Verizon allowed tethering.)

      • Hamranhansenhansen

        RIM originally required a notarized proof of identity in order to sign up as a developer. They have since relaxed this requirement. Apple does this by requiring developers to have an "Apple ID," which typically has a credit card linked to it, although I don't know if that is a requirement.

        You can tether an iPad to either an AT&T or Verizon iPhone via Wi-Fi since the launch of iPad 2. Before that, the feature was only on Verizon. It also enables the Wi-Fi iPad to use the iPhone's GPS.

      • addicted

        Since you need to pay $100 to publish apps to the App Store, I would assume that a developer would necessarily need to link a credit card to their account, to pay the $100 fee, if nothing else.

  • CndnRschr

    A defining attribute of true leadership is knowing when to say no. RIM seems to have said yes to everything in terms of programming environments. This offers choice but will mean that lazy developers will not bother to learn the native environment and will instead port over their code and run it sub-optimally. Meanwhile, the best developers will have their work buried by the free and easy garbage. Choice is usually good. In this case, it will obfuscate good.

  • You nailed it. When I read Balsillie's comments yesterday, my first thought was, "They threw the entire BlackBerry developer eco-system under the bus". The sad thing is the developers were doing OK. They were gaining traction and actually making more money than Android's developers. When you consider the far fewer number of developers actually doing BB work compared to Android, a per developer analysis might have shown them doing pretty well.

    I can't see any existing BB developers jumping in joy of this announcement.

  • James Katt

    RIM is saying:
    1. The future for tablets is a PC-Like Operating System. And since Windows 7 for Tablets is not yet available nor wanted, RIM is going to evolve their own QNX OS into this PC-Like Operating System for Tablets.

    2. They are hedging their bets by including everything but the kitchen sink (that they can get) into their tablet. This includes a player/emulator for Android apps, Flash apps, Web apps, and even legacy Blackberry apps.

    3. And since the Blackberry OS for their smartphones is totally unlike the QNX OS for their tablet, and since the competition is so strong, they want to first sell to Blackberry owners by forcing their tablet to be tethered to Blackberry smartphones.

    This is RIM's entire strategy in a nutshell.

  • Marc in Chicago

    "There’s no compromise here."
    "This is a no compromise environment."
    "We believe it’s about an uncompromised web."

    Translation: We've compromised the user experience.

  • This is interesting.

    On one hand, you have Jim Balsillie who seems to want to urge developers to not write Android apps for the PlayBook. He wants a future without needing to support these apps. Regardless of this position, he builds in support for Android apps in the PlayBook "in the meantime", imagining that they'll move to QNX.

    On the other hand, you have Steve Jobs, who wants to urge web developers to not write web apps in Flash. He wants a future without needing to support Flash. Consistent with his position, iOS doesn't support Flash "in the meantime", knowing that if they did, nobody would ever have reason to leave Flash as a technology.

    RIM: if you support Android apps "for the meantime", QNX won't take off. It just won't. This is a wimpy move on RIM's part.

    • WaltFrench

      Oh, give me a few nanoseconds more and I'll come up with a dozen other reasons not to support Flash. First being that Adobe is treacherous in its over-promising. nVIDIA may be locked into Flash for the foreseeable future but nobody else has to look at the Xoom fiasco and volunteer to get sucker-punched the same way.

  • Pingback: Android Apps on PlayBook – Brilliant Idea or Big Mistake? - Finding Out About()

  • Slap Shot

    Let me echo what has already been said: what a mess, What a Mess, WHAT A MESS, !!!WHAT A MESS!!!

  • Pingback: Weekly Poll: Android Apps on PlayBook – Brilliant Idea or Big Mistake? : Casa Semplice()

  • Senator Gronk

    RIM talks like they're throwing checks, but they're not even on the ice.

  • Guest

    The length of his explanation, his delivery, all speaks to why BB, Moto and others are having trouble being seen as equal with the Apple Experience. Steve Jobs would have put that into 1 paragraph and spun that into gold. Apple sells product with every word, picture, and video they put out there. The RDF exists. This load of BS that has had to be explained by Horace and others is proof that they just dont get it. You have to sell the experience.

  • Joe_Winfield_IL

    It's surprising that RIM perceives apps as the sole driving force behind iOS and Android's market success. Of course having proper "tonnage" of apps is a bonus, but there is so much more to it. Besides the battery life, UX/UI, form factor, etc., I have one HUGE issue with Playbook.

    This thing doesn't do email! It's a Blackberry that can't read and write email! Sure it can be tethered to a BB phone to get the mail client, but they have artificially restricted themselves to their existing customer base. Also, users will have to always have both devices in order to use the Playbook. If the phone battery dies, so does your email. Not having email is like not having a power button or a web browser.

    Just like the initial Android tablets, RIM is asking its customers to buy a half-baked product and wait for upgrades. They are also asking their customers to have a separate phone contract, just like the first Androids. Of all the great features that iPad introduced, the #1 attribute is that it is a TRANSACTION. You can buy the product and never send money to Apple or a wireless carrier ever again. It is easy to give as a gift, it requires almost no account setup, and it doesn't need a future software/hardware update to get the advertised features.

    So for the same price as iPad we get:
    1/2 the size
    probably lower battery life
    no email
    few native apps
    poorly run virtualized apps
    very weak media software

    Tough call on which to buy. But hey, at least the cameras are better. And there is nothing cooler than taking snapshots with something the size of a book.

    • davel

      No email?


      • Joe_Winfield_IL

        See my response to @Bobegan on the second page of comments for a more thorough explanation. I guess it's unfair to say that it doesn't do email. More accurately, it doesn't do Blackberry email. Bad fit for a Blackberry device.

    • KenC

      At least Palm had the good sense to kill their Folio before it launched!

    • addicted

      "there is nothing cooler than taking snapshots with something the size of a book"

      Considering that this is likely to be bought by BB users (lack of BB mail without tethering with a BB), even the cameras are worthless because people will use their phones to take pictures.

    • Justin

      It will be shipped with Microsoft Cloud 365 which includes Cloud Exchange Server. However Blackberry e-mails must be tethered to a secure Blackberry device to ensure security in a workplace environment. Useful for for of us in IT as we do not need to do *any* additional upgrades, updates, or configurations to roll out Playbooks in the workplace since it all goes through the existing security of the users Blackberry.

      I believe the media experience will be better on the Playbook simply because of it's widescreen (16:9) form factor opposed to the iPad's 4:3 form factor that leaves huge black bars at the top/bottom of the device. (effectivly limiting the viewing are to a 7" screen anyway) Nice to finally get stereo speakers on a device too, watching movies and listening to MP3's on my iPad is painful.

      But I realize these features don't matter to a lot of consumers too.

  • Pingback: Weekly Poll: Android Apps on PlayBook – Brilliant Idea or Big Mistake? | Programming Blog()

  • What part of native don't you understand? Its not likely to be Dalvik b/c of the Oracle litigation but RIM is more likely to release something based on IcedRocket or OpenJDK.

    I have seen no indication from RIM that they would run in an emulation mode. And I would jump all over them if they did. It would be just dumb with all the options and tools available.

    It also quite clear that none of you have used a Playbook. It feels good to hold and is very snappy.
    The Playbook does true multi-tasking unlike the iPad (which i own btw)
    By contrast the new Samsung and HTC tablets, while very nice in a lot of ways, feel more delicate and not as quick (I am not sure why).

    For those of you who say it can't do email, you are just plain mis-informed.
    It does email over wi-fi just fine, and also via a tether to a bb phone for 3g/4g (which i think is a mis-step)

    RIM has a communication problem. You bet. But lets keep the facts straight.
    RIM also reminds me of what got Nokia into trouble – International growth while loosing North American share.
    This is a real problem.

    • Joe_Winfield_IL

      "It does email over wi-fi just fine" – sort of. It does not do Blackberry's own email over WiFi. Balsillie has said repeatedly that RIM is targeting enterprise users (core BB users) over consumers. These users, the very target customers of the PlayBook, will be locked out of their email unless tethered. The company has put themselves in a position where they have a better product offering to new customers than to existing ones.

      I'm sure it's a beautiful, well built device. I hope it does well, if only to keep Android from consolidating the market of non-iPads. It has plenty of positive features, but I have a hard time envisioning the target audience. For the same sales price, iPad targets students, consumer use at home, business users, medical professionals, frequent travelers, and computer neophytes. Other than enhanced security for businesses, I don't understand what QNX/RIM/BB can offer. Because tablets are typically used more for content consumption than for communication or content creation, Apple's media chops give it a huge leg up vs. their competitors to the north. The physical keyboard alone keeps Blackberry relevant in the smartphone business; I'm struggling to see the tablet equivalent.

    • technicalconclusions

      The problem is this. I was interested in the Playbook months ago when it was first introduced. I think QNX was a big step in the right direction for RIM. The problem is, much has happened over the past 2 months, most notably, the iPad 2. The hardware doesn't seem so special anymore. More importantly, I see nothing which makes me want to get excited about this anymore. If they shipped a product back in October, things may have worked better for them. The more I hear Balsilli speak, the more I am concerned for RIM's future.

      Worse, they should be going after the consumer market and not just assume they will naturally inherit the enterprise market. What RIM doesn't realize is that many people that do use a Blackberry for work no longer do because they want to. They do it because that's all their company will support. The playbook is different enough from the Blackberry such that enterprise adoption of this product is anything but guaranteed. I'm seeing pressure for IT departments to support iPhones and Androids and nothing else. The enterprise market is slow to react to user demands, but Blackberry is no longer the preferred device by users.

    • davel

      Since you have played with both. How fast is it compared to ipad2?
      How is its graphics?
      When does it ship?

    • WaltFrench

      Oracle IS upset with Dalvik, but not the fact that it is one first-class piece of language translation software. If RIM licenses mobile java for Playbook (seem HIGHLY likely), no issue with Dalvik. And if they don't, the use of an alternate JVM interpreter wouldn't help keeping Oracle's hackles down.

    • don

      Yea, my 90 year old grandmother was all set to buy an iPad, but decided not to because of the lack of "true multitasking". That's a real deal breaker that is causing Apple to lose billions of dollars in sales.

    • ______

      Who cares if it is native, non-native, some half-baked scheme in between? The point is, why support a competing, smartphone grade, probably-outdated-in-a-few-months OS? Dumb, Dumb, Dumb.

      As for true multi-tasking, I'm happy using my tablet that does what I'm telling it to do (for the last YEAR) rather than masturbating over some true-multitasking mythical, fantastic, tablet Goddess that never appears. In the following year, I KNOW that my iPad 1 (now handed down to other family) and iPad 2 will steadily improve with SW to the extent that is reasonable.

      Can't say the same about any other tablet. I only have to look at the various phone HW devices and their upgrade record over the last couple of years.

  • David

    I suspect it will be slower. VMs always extract a penalty over native. VMWare is about as fast as you can get and has virtually unlimited amounts of power, CPU and memory and stills runs a bit slower. More than usable, a bit. Mobile doesn't have the benefit of of unlimited resources and must be constrained. Imagine a scenario where the Flash/Air runtime, Qnx native, html/css/js, java BB and Android Dalvik are all running. The GUI collisions alone will be impressive, let alone the resources allocated to running all these environments. Android only recently started supporting HW for the GUI. Will Playbook support this? If not, the apps will be doubly slow. Try running VMWare Unity for apps to get an idea.

    How much time could they have possibly had to optimize all this? Back in the day, OS/2 ran Windows 3.1 applications. It ran them either full sceen, or seamless(you just see the app, again, think Unity). You could run each app in a vm or run them all in a one session. Will playbook run each app in its own machine or a shared instance? What about Air? It can get messy.

    I suspect that there is a reason battery life stats haven't put in an appearance.

    Furthermore, what's going to happen when someone try to run an Android native game on Playbook? Will it work? Poorly? At all?

    Finally, let's give them the benefit of the doubt and say that it works well? What's the motivation to write native apps? OS/2 never got apps because it ran Windows apps so well. And when Win95 came out, endgame.

    But if it doesn't work, why bother writing anything because it doesn't work and few will buy it.

    • David

      That's supporting HW acceleration.

      • WaltFrench

        Yup, a Honeycomb feature that RIM will have to emulate in their VM for all the reasons you originally cited, and almost certainly what JimB meant when he said graphics in the Android Green Box wouldn't be its best feature.

    • I think those are great points, but all things being equal, there really isn't a "native" mode or app in Android. Everything runs inside a VM…right?

      I think what you are saying is that multi-tasking could do them in?

      • David

        My understanding is that Android either has or will soon make a native C++ set of APIs available for games. To my knowledge, there is very little VM based gaming. Java, which I do most of the time, is dead on the client except for development tools.

        I don't think the multitasking is that much of a selling point. People, I think want to browse and listen to music. I've always thought that the card-based metaphor of WebOS(and now Playbook) isn't that compelling to must people. From what I've seen, most people don't seem to get multitasking very well even in 2011.

        Not multitasking, but running these different platforms. I think that there is a price to pay to support these environments. At the least, they won't behave the same way resulting in a disjointed experience. Cross-platform tools have to either support the lowest common denominator of all their hosts or create their own widgets. Look at Air as an example of the latter and Java the former.

        So, you'll have Android looking apps, Air looking apps, Qnx native apps, and Java BB apps. All not looking alike and probably not very well integrated. If you click on a mail link, will your Android FF browser launch or your Torch? How will they exchange data? Will cut and paste work across all three?Can your Android app access your Playbook contact list? Will Swype work at all? Will those Android trojans work? 🙂

        One of iOSes greatest strengths is the consistency of the user experience. What's more frustrating than when Linux didn't allow you to cut and paste across different apps. So, this isn't a problem of multitasking(really) at all, but of integration.

        Could RIM have resolved all of this in this amount of time? If they did, that's one great piece of engineering!

      • I worry about the same disjointed Ux. We agree on that for sure.

        I guess i am less worried about the horsepower to run multiple code instances of different types than what i thought you were talking about in your earlier post.

        All in all Tablet makers will be judge based on their user experience, not specs, like in the Windows which by definition defined the Uz allowing one to compare (almost) one machine vs another by memory sizes, cpu speed etc etc

      • David

        Well, less the horsepower and more the responsiveness once several of those things of running. A few fast mockups won't necessarily represent how fast everything is should some take advantage of what they are offering. I don't (yet) know enough about how much a given runtime or vm would take it. However, if Android is not as smooth as iOS because it lacks HW acceleration, how well could Android apps run on Playbook if it doesn't support this. It's not just an issue of raw CPU.

        And in all Flash cases, we've yet to see a comprehensive review of the battery impact. I think the reasons are clear. The battery suffers. If it didn't, Adobe and all of Apple's competitors would be making hay of out Steve Jobs from the moment flash was first released on Android.

      • WaltFrench

        “ Android either has or will soon make a native C++ set of APIs available for games.”

        Google has announced support for C (C++?) code, but has strongly emphasized its purpose as special-purpose, high-performance code chunks that won't interfere with how the Dalvik VM is controlling the environment that it operates within.

      • Jay


        "If you click on a mail link, will your Android FF browser launch or your Torch? How will they exchange data? Will cut and paste work across all three?Can your Android app access your Playbook contact list? Will Swype work at all? Will those Android trojans work? 🙂 "


        Hi David,

        I don't claim to know all the answers but I can certainly answer some of these for you as I have tried a live demo of the Playbook for a short period of time and have experienced the tethering in action. Questions regarding development I don't know any answers to.

        – If you click a mail link on your Playbook it will launch on the Playbook, not the Torch. There's a video out there of the integration in use with deleting mail, synchronizing, clicking links, and the auto-wipe features upon bluetooth disconnection.

      • Jay

        – Data is exchanged via secure bluetooth. The blackberry service is a secure environment within the workplace, the Playbook passes all data through the Blackberry to ensure this security without requiring IT to make additional changes / upgrades / configurations to support the Playbook. Any user who already has a secure Blackberry device can purchase a Playbook without 'special permission' from IT and can begin using it immediately.

        – Cut and paste will only work on the device that is in use.

        – According to the RIM announcement on Friday: Android apps will NOT be able to access your playbook contact list, or anything else on the playbook. Android apps will run within a secure sandbox environment to prevent this from happening. I assume this goes back to security within the workplace.

      • Jay

        On a side note, I personally see multitasking as a selling point and the single reason I did not upgrade from an iPad to an iPad2. Even basic things like Safari browser supporting tabbed browsing. I'd like to check my GMail in one window, and have another window open to surf the internet at least, or to research something for an email response. That's just within Safari.

        I basically bought the iPad for movies on the go and casual internet browsing. It does neither very well unfortunately. (4:3 aspect ratio brings the actual widescreen movie viewing area down to a near 7" screen anyway, lack of stereo speakers compound the issue.) Internet browsing is weak unless you're using wikipedia or just reading text.

        I assume most of these issues will be resolved with iPad3. I suspect Apple is hard at work with something to fight back with, or will soon drop the price of their iPad to undercut the Playbook once it's released.

  • Vito Positano

    It has taken about 4 years for the British cliche "Having said that,..," first used by political talking heads, to become a widely used cliche in the US tech industry.

  • Sergio

    "You’re just not going to get things like gaming and multimedia, you’re not going to get the speed going through a VM interface."

    I still don't get why they decided to call it Playbook in the first place…

  • Steko

    You missed the best part where he blamed the FCC for not shipping anything yet.

    • ______


  • What I hear is a guy who is resigned to a world he is neither interested in nor prepared for. After years of playing their own game, this rambling monolog admits RIMM is playing someone else's game, learning by watching since they can't find the rulebook. Ballmer-esque.

  • synth

    Like he said, it's a strategical mess.

  • vinner57

    He sounds resigned to his fate and contemptuous of where the market has gone. It's like Steve Jobs pissed in his pool. It's a 'vision' to appease marketing and the analysts. and try and maintain the value of everybody's stock options. The customers and developers can go f*** themselves.

    RIM have a loyal following 'in the enterprise' but as that IT head honcho said the other day "the tyranny of the consumer" (code for Apple) has come and I can't believe this is going to end well.

  • Well, RIM is targeting their bread and butter corporate IT manager/decision maker.

    This is the market where end users have little say and purchase decisions are made primarily to protect and enhance IT carriers while paying lip service to business benefits. Rim is giving those iT managers all the check list items they need to justify the purchase of the rim tablet. You already starting to see corporate resellers touting rim tablet support in their press releases. So in this market Rim can expect success At least initially. But as a consumer device, their tablet will likely be DOA.

    • WaltFrench

      Ummm, no.

      Maybe ManuLife volunteered to sponsor the Home Team but IT otherwise needs results. This incoherent development approach means in-house developers still have no idea how they can build first-class apps that interact with other corporate resources such as email, browser access (think MSFT's online help pages), etc.

      Or, they could write their game… uh Customer Relationship Management system in Flash, the recommended approach just 48 hours ago.

      So, why did RIM wait so long before announcing a roadmap (never mind all the potholes and questions here)? What was their Plan A that was going to be ready before April 19th? Their native SDK is reportedly not even up to alpha stage.


    • davel

      My understanding is that Apple has addressed many if not all of the critical objections from IT with regards to iOS and their hardware.

      In the new world of business IT doesn't have that much to say regarding what products come in house. Vendors have long ago gone to the people writing the checks.

      Every company has decision makers/influencers running iOS. RIMM has a lot to be afraid of because their platform just is not that good and the competition is good enough for office.

  • Pieter

    I find it very funny that Balsillie first claims that using a VM will not give good performance and then proudly announce support for 3 VMs on the Playbook: Flash/AIR (Flash VM), Android (Dalvik) and BB apps (Java VM)…

  • CndnRschr

    Actually, Google is an even better example if you consider its rate of acquisitions (that often seem to vaporise) and the number of beta products that simply drift quietly into the night. The difference (compared to Acer) is that Google at least innovates (Google Voice, Street View, Maps) whereas Acer seems to shuffle the same old, low cost components. Acer is the new/old Dell.

    • Hamranhansenhansen

      The problem is, Acer and many others just go component shopping and then build whatever they can out of those components. Out of whatever is out there. Apple designs a useful device first, for example, iPad in 2003, and then later buys, invents, or custom-builds the components they need. Acer's process leads to adding more components, while Apple's leads to taking components away, revealing only the essentials.

      Until somebody outside Apple starts designing, Apple is going to continue to own consumer electronics. Consumers respond to well designed products when given the choice. My roommate is on his second iPhone, and did not know the switch on the side mutes the ringer. He doesn't need more features!

      • The difference is that Apple starts with the question, "How should the computer experience be like?" and then they work backwards to get as close as they can.

        Acer starts with the components they have on shelf and then tries to build forward.

        The advantage of starting with a vision is that it is much easier to make tough decisions. It is easier to say no. It is easier to prioritize. It is easier to identify important new technologies.

  • jsk

    Sounds more like RIM's tablet strategy is "We think consumer really only want text based email on their phones and tablets are fad that will pass quickly. That's why we're not investing a lot of money in a stand alone device and giving our investors pie-in-the-sky lip service to keep our stock price from tanking."

    Seriously, if Android support is such a selling point, why wouldn't a consumer just get an Android device in the first place?

  • Pingback: CrackBerry Poll: How do you feel about RIM’s Announcement to Support Android Apps on the BlackBerry PlayBook? | 101 Best BlackBerry Apps()

  • Hamranhansenhansen

    No, Ballmer is incoherent, and so are almost all Microsoft people. They have a lot of their own shorthand verbiage that comes from their isolation from the rest of tech, both physically due to being in Washington, and technically because they are the only system that is not UNIX compatible. Further, they have to lie to themselves a lot to pretend they are innovators rather than just cloners, so they say outrageous things, like when Bill Gates said Mac OS X was a copy of Windows Vista. Straight-faced! Who can understand such pretzel logic? Only another Microsoft person.

  • kevin

    For Playbook to be a big hit, it needs to be attractive to mainstream consumers. Being able to run Android apps holds out that promise. But if it isn’t seamless with the use of the other kinds of apps on the Playbook, and thus degrades the user experience, it will be quickly panned and rejected by those mainstream consumers (while being loved by all the tech geeks).

    The tablet is successful thus far because it is seen as consumer electronics, not as a complex computer. Until Aple’s competitors truly grok that, they will be competing for scraps.

  • Bigfoot

    My translation of Balsillie's comments: I am a deeply insecure human being who simply cannot help blabbering incoherently whenever presented with a situation in which my abilities as a manager are scrutinized by others. In reality I don't have a clue about how to build a product that even comes close to competing with the iPad. However, investors seem to be expecting me to do something. Therefore, I have decided to pretend that I actually know what I'm doing, and to try to bullshit my way through any situation in which I actually have to explain to others what my nonexistent strategy is. Deep down inside I know that I cannot possibly fool anyone with my bullshit, but the urge to do so is stronger than me. I must remember to discuss this with my therapist the next time I see him.

  • Steve White

    Having followed this story, here's what I don't get — if QNX is the future (as far as RIM is concerned), why not 1) make it very, very easy for developers to write apps in QNX and 2) more importantly, start the ball rolling by having RIM itself write some great apps using QNX that are included on the PlayBook? And showcase these great apps every chance you get?

    When Apple launched the iPhone, it had a number of built in apps. Ditto for the iPad. Apple continues to build marvelous apps (e.g., GarageBand on the iPad) as a way of showing developers what can be done on an iOS device. Steve Jobs isn't shy about showing these apps off. "See what we did?"

    RIM should be doing the same with QNX. Start getting developers thinking about how they could build a PlayBook app in QNX that would excellent, that could get into the market early, that could make some money. Get an iOS developer thinking about a port to QNX rather than to Android. Show off how cool QNX is, how polished a QNX application is, how an end-user will appreciate a QNX app.

    RIM should have a GREAT e-mail client on the PlayBook, written in QNX of course, that is truly enterprise worthy. Show all your Blackberry customers that THIS is how you check your e-mail on a tablet. Ditto a web browser. Ditto other business and enterprise oriented apps. Ditto a great video viewer for those long airplane flights. It doesn't matter if you only have a few apps on day 1 of the PlayBook introduction if those apps are slick, well done, and useful. That was the situation Apple was in on day 1 of the iPhone. They managed.

    Dump Android. LAUGH at Android. Point out how full of holes Honeycomb is. Point out how foolish it is rely on an Android app. Why on earth would you associate yourself with all the crap-ware in the Android app market now just for the sake of 'tonnage'? If your market is business and enterprise, then GO at that market with some class. Does Apple associate itself with crap-ware? Of course not, Apple is angling for the top 10% of the market. RIM should do the same — go after the folks who fly twice a week for their businesses, the analysts who are constantly on the go, the market makers, public servants, academics, etc. You're not selling to the cheapskates who demand that everything be 'open' or 'free', you're selling to the people who appreciate, and will pay for, value.

    In other words, go after the people who traditionally have bought Blackberries (and who have lately been buying iOS devices). But NOT by forcing them to tether to a Blackberry phone just to check their e-mail, and NOT by allowing them to download Android crap, but by offering them a complimentary experience that is first class in every way. Give them an experience that makes them WANT a PlayBook.

    That means a singular focus on the best SDK available for the device, and right now that's QNX. Lead the way by writing superb apps, and get the developers to see that they can do it as well, and make money.

    Someone above said that leadership involves saying 'no'. Well, Mr. Balsillie: no to Android. No to Flash/Air. Push QNX. Support HTML 5 in every way possible as a back-up. Stop worrying about whether you have all the boxes checked, and start obsessing about delivering the best product possible.

    • I 100% agree with this. If RIM would come out and say this is why you need QnX (and QnX is a great OS). This is what QnX can do for your life. Create a great email/messaging application suite. Create a few unique and innovative enterprise based applications. Create a reason to have a PlayBook and not some off Android or iPad device…

      But this kitchen sink approach is a looser strategy if you ask me.

      • What do you think of this idea?

        Release an unfinished dev kit. Tell developers why you think QnX is the future and then tell them that you are going to be buying up Apps. Do you want to create a great email/messaging app? Do it. If a lot of people like it, we will buy it and you will get a fat pay day. Just like how Twitter bought Tweetie.

        Lot's of problems to this idea, but RIM isn't in the position to play it safe.

    • chandra2

      "..If your market is busine"..If your market is business and enterprise, then GO at that market with some class.."

      I agree. That is the confusing aspect. This move is really to check the boxes. I am sure they will have some native QNX apps upon launch. It will be foolish to ship without a native Mail client, Browser, address book, calendar, maps etc. ss and enterprise, then GO at that market with some class.."

    • davel

      Great post.

      Unfortunately few companies have the vision that Apple does. That is why they wait around for Apple to design the hardware specs, the software specs and the pricing AND THEN they copy it.

      Seriously, this has been the business model since the iPhone came out. This was the business model last year when Apple introduced the iPad and everyone went home to take it apart to figure out how to build a copy.

      Everyone is keeping the Apple pricing after getting a who cares response to their higher pricing plans.

  • For an extra $499 Jim Balsillie will come to your house and physically glue an iPad to the back of your PlayBook so it runs iOS apps too. No compromises!

  • airmanchairman

    Brilliant translation and cutting analysis, though understated here (makes it even more succinct).

    They should have had you on stage, like those religious evangelists in foreign countries, signing and translating every now and then for those of us that do not live on Planet RIMM.

  • berult

    RIMM bases its survival strategies on lock-ins.

    Consumer wise, they're going after their home-based market. Call it the ethno-centric, Canadian 'pride and joy' consumer market; roughly 35 million nationalistic one of them, an exploitable market large enough to buy them time on the consumer front. An exploitable market indeed for, to tie the consumer up into their emotional sphere of influence a simple illusion of covering a lot of software ground along with some hardwired goodies will sooth the Canadian ego just fine …so they account.

    They're simply playing their home field advantage to the fullest …to survive for another day and, fingers crossed, sock it to 'em in glory days to come ….so they discount. Nokia gets a mere 4,5 on the same home field scale, motivation enough to outsource their development strategy to other 'promised lands' I reckon. It's a matter of cross-over pollination of national vanity into a failed product's marketing strategy. RIMM's on the rope but Canada ain't…!

    …their products may be simpletons, but so are third-world myriad of consumers' needs…RIMM believes; the CEOs' esoteric discourse is simply adding coded poetry to coded insufficiency. Murky, …sure; opaque, …not on my homely watch!

    Business market wise, there are no subtleties here. Lock-ins have been RIMM's bread and butter for… , so no need for me to wander out on a beaten path other than to personally lock-in, in a spate of poetic justice, RIMM's survivability to that of its proselytizing ITs…

  • Hoo boy, it would have been really convoluted if he had announced a Blackberry OS6 emulation layer. Oh, wait! You're supposed to use your Blackberry to tether with. Will Blackberry offer a bundle? Buy a Torch 2 and a PlayBook for one price and get one data plan? There seems to be no sales model for this thing announced because no one has told us when they are coming out. I am sure developers are as confused as Balsillie; do I write an Android app, an AIR app or use the SDK and write a native app? I also agree that they are beta testing these things on the backs of the users. The only other compelling tablet is the HP TouchPad because it works on a truly scalable mobile OS that is in the wild today, there just aren't any apps for it.

  • chandra2

    How many VMs does it take to run an Android flash based game on the Playbook?

  • k.t.

    This reminds me of the "fed-speak" by alan greenspan, who turned out to be such a fraud. The morale of the story would be to never trust someone who speaks gibberish, I guess.

  • Andrew

    Alternative Translation:

    I'm a hardware man myself. I love all those qwerty keys, the buttons, sliders, trackballs ………. never really liked touchscreens, but that's what kids want these days. When we've got the hardware right, we hand it to the software boys to make it work. They are taking far too long – they explained all the difficulties, but I didn't really follow – so I'm telling you what I remember, except for the things I'm not supposed to talk about …….

  • Pingback: Linkbait 13 | Programming Blog()