Analyst Predictions

Posted in Spritely by Dimitri on September 22, 2011

I don’t understand how research firms and analysts come up with their predictions. It often seems like they follow a formula that ignores past successes and is overly optimistic of new products.

According to Gartner (via 9to5mac.com) tablets are set to take off over the next 4 years. I don’t think anyone debates this.

However, here are their predictions for the OS’s that those tablets are going to run:


QNX (better known as the Blackberry Playbook OS) has been a flop this year, and didn’t even exist last year. But they predict that in 2015, it will ship on 26 million tablets, or 8% market share.

Microsoft has made no real progress in the tablet space in the past few years. Each announcement has been either delayed, or cancelled. Remember when at CES 2 years ago Ballmer showed off the “slate” that ran Windows 7? It never shipped. But Gartner believes that Windows 8 will sell 34 million tablets in 2015, or double what the iPad shipped last year.

Android Tablets barely have made a dent in the market. There are a lot of them, but none have made a meaningful impact. I don’t know if Gartner includes all Androids (for instance, the Barnes and Noble Nook Color) in these figures. However, they expect that it will start approaching the iPad in 2015.

I think that at the moment, it’s anyone’s guess. Until we can start seeing the start of explosive sales from ANY of the tablet vendors other than Apple, I don’t think it’s safe to assume that the market will automatically spread out between vendors. The only safe bet for the moment is that Apple will continue to sell tons of iPads. They ARE the tablet market at this point.



Comments Off on Analyst Predictions

WWDC Predictions Part 2: iCloud

Posted in Spritely by Dimitri on June 6, 2011

Apple’s attempt at a major cloud services offering is almost here. Tomorrow morning we’ll have a better picture of what iCloud is, and how it will affect Mac and iOS users and developers.

In this post I’m going to look into what Apple could be working on from the perspective of what I think that iOS could benefit from with a comprehensive cloud services offering. By identifying these areas that can be improved, I think that I can predict at least some of what we on Apple’s priority list with iCloud.

Problem: larger music ibraries don’t fit on smaller devices
Problem: Purchasing content from iTunes doesn’t allow that content to be redownloaded at a later time.
I think that for some users, this will be the most exciting thing about iCloud, while to many others, it will have limited usefulness. The Music companies can never keep quiet when signing a new deal, and this is no exception, so we know that there will be at least some kid of streaming or synching cloud based library. Perhaps there will be a fee for this, perhaps not, but I think that this will be only one of the important aspect of iCloud.

Problem: iOS devices require a computer when you first get them home, even before you use them.
Problem: iOS devices require a computer and iTunes to be able to back up their contents
Expect to see some of this addressed by Apple tomorrow. I’m not sure if they can pull off an entire computer-free environment just yet, but something like adding Time Capsule support for iOS would allow the device to do it’s sync and backup without iTunes, and without requiring a home internet connection with a huge upstream and lots of cloud storage.

Problem: Finding already-purchased apps on the App Store for re-download is difficult.
Even though the App Store allows unlimited re-downloading of apps, it doesn’t provide a view of all of your apps, making the simplest way to make sure your apps are on your phone is to use iTunes. A proper interface for listing the apps available to install would simplify matters.

Problem: Apps don’t update themselves from the App Store, users have to go get all updates manually
I’m sure that Apple is going to be changing this. Expect them to allow apps to auto-update in the background.

Problem: With multiple iOS devices, settings and app data doesn’t sync across the web easily.
If you have multiple devices, this can be a real annoyance. Some app developers that want to address this end up writing there own web connections that allow them to synchronize the devices. Most end up leaving it alone. What if iCloud brought dropbox-like content synching to iOS, where a simple folder in the file system would automatically be synchronized with the cloud, so that all devices would have this same folder available?

Problem: Multitasking rules prevent applications from retrieving updated content from the web without the application launching.
Apps like The Daily, Reeder, Instapaper, and any number of magazines expect that you’ll have internet access wherever you prefer to do your reading. They then all go through a process to download the latest issue, which a user must wait for. What if iCloud had a dropbox-like folder that also had a server-side API so that magazines could write the latest issue to it as it’s released, and then leave it to be synchronized to the devices by Apple? An app not launched for days could still have this morning’s issue ready for you.

Problem: Downloading large files over 3G strains networks and costs users money.
I imagine we will see a limit on the size of files that can be synchronized to iCloud over 3G. Assuming my drobox-like sync is correct, you can imagine to folders: one that can be transferred while on 3G, another that has to transfer using WiFi. By setting a limit on the 3G size, Apple can allow a small settings file to be synched while on the go, but force hundred-megabyte magazine issues to be downloaded over WiFi. I’d expect the same rules to be in effect for iTunes content: perhaps smaller files like mp3’s will stream over 3G, but movies will require WiFi.

Problem: Downloading large files kills battery.
Apple is very proud of it’s 10 hours of true battery life it gets from the iPad, and all-day battery usage on the iPhone. It’s not going to sacrifice those for anything. So if the above is true, expect the device to have some sort of “dock” mode, when it’s connected to power. This means that when you plug your iPhone or iPad in at night in your home connected to local WiFi, it will only then go retrieve the large files waiting to be downloaded to it. Since most users do this regularly (nightly for iPhones), it shouldn’t become much of a problem. Expect this to be when apps auto-update, so that any large apps don’t harm battery life.


Just imagine: you pop into the subway for your morning commute, and the paper has already been delivered to your iPad (overnight, while charging) via the iCloud API. You then open your favorite To-Do list app and add a task, which will be queued to be sent to iCloud. Since it’s small, the update gets sent out the moment you walk out of the subway. iCloud then distributes this change back down to your iPhone and Mac clients, and allows the developer access to that task via an API (to allow the developer’s website to show the task).

You then want to listen to some music, so you open up the iPod app to find your entire 50GB library available on your iPhone. Only 5GB is available for offline listening, but all the rest is available for streaming. Rather than synching via iTunes on the Mac, you use the iPod app to select which music and videos you want on your device. If the changes are small (downloading a track or album), the data is synched before your eyes, ready to be accessed offline. If it’s large, an alert notifies you that your changes will be made the next time you are on WiFi and charging.

Upon returning home, you plug in your iPad, iPhone, and drift off to sleep. Your devices then back up their data to your local Time Capsule as you sleep (maybe they’ll call it Time Machine for iOS?). They will also fetch the next day’s paper and sync your music choices.

If Apple pulls off these features correctly, they’ll be miles ahead of the competition, especially if the service is free. I think the key is allowing app developers access to all of this stuff so that we can make use of it. If Apple makes it really easy to do something that was once quite tricky, it raises the bar for all iOS apps. That bar is already much higher than that of other platforms. To users, it will just work.

Comments Off on WWDC Predictions Part 2: iCloud

WWDC Predictions Part 1: iOS and Lion

Posted in Spritely by Dimitri on June 2, 2011

Next week is WWDC, and Steve Jobs will again take the stage to show us iOS 5, OS X Lion, and iCloud. In the last 2 years I’ve attended WWDC, my lack of preparation and coffee relegated me to watch on a big projection screen in the overflow room. This year, I’m going to force myself to go early to see the keynote live. There just is no substitute for the real thing, and I think this year will be notable: in the last few years we’ve seen hardware released at WWDC, which is fun but not nearly exciting for developers as a major new OS upgrade.

What we know for sure: we will see the announcement of 3 products: Mac OS X Lion, iOS 5, and “iCloud”. I think that iCloud will touch both iOS and Mac, and will be the most exciting of the 3 announcements. I’m going to talk about that in a following post.

Mac OS X 10.7 Lion
I think I’m least excited in what they have to say about this. So far what I’ve seen of Lion has been underwhelming, and unless they reveal something completely unannounced in the betas released to developers, I don’t see it changing the way I use my Mac. What I do expect is some deep integration with iCloud, which I think will be far more interesting than the rest of the OS announcement.

iOS 5
I think we’re in for a treat with iOS 5. iOS 4’s most notable feature was multitasking. I think that iOS 5’s notable feature will be iCloud, followed by a new notification system.

iOS notifications suck. Plain and simple and horrible for anyone who uses more than 1 or 2 apps, or for anyone who messages regularly. “Katie: 2 Messages” is not all that useful, and the way that most push notifications work, stacking the most recent message and causing me to have to play “dismiss the dialog boxes” is maddening. If I don’t use my iPad for a while, I’m destined to play this game with meeting requests, CNN alerts, Words with Friends updates, etc. Each I have to deal with, at that moment, one at a time, until I’m done with them all. And god forbid I want to actually act on 2 or more of those alerts instead of just dismissing them: acting on one brings up the app in question, only to show me another notification about a different app.

This has to be fixed for iOS 5, and I think they finally will. I don’t know what form it will take, but I’m hoping it has some sort of history so that I can see and act on past events. It would be interesting to see Apple create some sort of priority structure too, so missed calls have a much greater importance than letting me know my crops are ready in Farmville (note: use of Farmville for illustrative purposes only). The trick will be in making this system as dead-simple to use as the notifications: while they are frustrating to a power user, making someone make a Yes/No decision is quite straightforward. Anything that Apple does, they’re going to try to keep this level of simplicity.

I don’t know what Apple will do in this realm, but I think that providing widgets tastefully makes a lot of sense. Android features widgets on home screens, which in my opinion are ugly: it makes each screen look cluttered and unorganized. The pictures that they show you selling the phones are rarely what actual Android phones look like: I’ve found them to be an unorganized swipe fest of random forgotten widgets.

For letting me know about the weather, or show me my calendar, or any other such simple task that is designed for fast consumption, widgets make a lot of sense. They allow you to quickly get a lot of information at a glance without hunting through your apps. I wonder if Apple will make the lock-screen a new home for widgets, and save the home screens for app launching. Or perhaps they will use the area above the multitasking bar, making the double-tap a power users dream for task switching as well as info gathering. In this area I have no idea what Apple will do, and I think they are just as likely to leave the feature off if they haven’t dreamt up an elegant way to solve it.

Xcode 4
The Mac OS/iOS development platform for most of us is called Xcode, and just went through it’s biggest update yet, to version 4.0. I think Apple will update Xcode 4 to 4.1 during the conference, and will make it the required platform for building for Lion and iOS 5 upon the release of iOS 5. This is likely a few months away, and will allow most devs sufficient time to move over to the new platform (bugs and all).

As for other announcements, we could very well see some Mac hardware announced at the show, but are unlikely to see any iOS hardware. These are, in order of what I think is most likely:

  • Macbook Air (minor refresh, speed bump, thunderbolt)
  • Airport Express/Time Capsule (if rumors of stock are to believed, and to support iCloud)
  • Mac mini (which hasn’t seen an update in a year)

I don’t think we’re going to see an Apple TV announcement. The hope with Apple TV is that Apple will open it up to developers to write apps for, but I don’t see that happening just yet.

So that’s it. Stay tuned for my iCloud predictions, where I think the meat and potatoes of the announcement will be.

Comments Off on WWDC Predictions Part 1: iOS and Lion

The Apple ship is sealed up tight

Posted in Spritely by Dimitri on April 8, 2010

We’re a mere 1.5 hours from an Apple event, and there haven’t been any leaks. Speculation, but no leaks.

It was the same with the iPad event. Tons of rumors, but nobody got it right. Sure, we can all make fairly educated guesses, but that’s all they are. Apple has a completely locked down all rumors. For a company with hundreds of employees working on these projects, that is seriously impressive.

Comments Off on The Apple ship is sealed up tight

iPhone OS 4.0 Predictions

Posted in Spritely by Dimitri on April 7, 2010

Tomorrow Apple will be unveiling iPhone OS 4.0, which presumably will be released this summer alongside a new iPhone model. Here are my predictions and hopes for what we’re going to see tomorrow.

1. Multitasking

As I mentioned in my post earlier this week, I think multitasking will take the form of some sort of usable background services for developers, rather than the traditional model of keeping entire applications in memory simultaneously. highly likely

2. New Home Screen

This goes hand in hand with creating the appearance of multitasking. I wonder if the new home screen will be an overlay over a running app, and therefore will allow fast access to it without closing your current app. highly likely

3. Widgets

I talked about these last month. I think that these widget applications will do wonders for allowing the iPhone’s home screen to provide some real at-a-glance information to the user. Perhaps they will be the only method of multitasking, as they provide a visual representation of running services. probable

4. Better file support

I don’t think Apple will implement a Finder, or anything like it. However, they need to come up with a better way to transfer documents around within the OS. 3.2 is already making pretty big leaps in this direction by providing “Open With” functionality for apps, but as of yet, none of Apple’s own apps take advantage of this. Even with this “Open With” functionality, it makes sense to allow apps to “fetch” files of the types that they require. You can look to the “photos” or “contacts” functionality in the current OS to get a good example of this: Apps are allowed to fetch photos or contacts from their respective apps. Then again, since this would require some shared disk area for each document type, I’m not confident that we’ll see it in 4.0. Instead, I predict that Apple will implement the “open with” functionality in it’s own apps, so that apps can essentially pass files around between each other. If you want a .doc file out of your dropbox, it would require opening the dropbox app, tapping the word doc, tapping “open with”, opening it with pages, editing it, then sending it back to drop using “open with” again. possible

5. Printing

Before my grandmother can buy an iPad and replace her computer, she needs to be able to print. Until this can happen, she requires both devices. I think printing support will only work directly with WiFi and Network printers (no cables), and ONLY printers that Apple authorizes and builds basic drivers into the OS for. Additionally, I bet Apple will allow the Mac and Airport Express to act as printer shares via bonjour, so enabling an existing USB-only printer just requires buying an Airport Express. possible

6. Notifications

Let’s face it, notifications on the iPhone suck. They are modal, so they distract from what you are doing and demand instant responses. They only allow one to appear at a time before they start to stack. On the iPad, which has plenty of screen real estate to show off more, they only show a single, modal dialog. I fully expect Apple to roll out a new notification system that is built into the new home screen and presents notifications in a much more usable manner. Just check out notifications on Android or WebOS and you’ll see how far behind Apple is on this one. highly likely

7. Second screen size for phones

This one’s a tricky one. The next iPhone is rumored to be double the resolution of the current one. But as the iPad showed, developers need time to develop for new screen resolutions. Therefore, Apple can’t simply announce their next device and expect there to be apps that work well on it. This one I’m fairly doubtful about, since I know how much Apple hates to tip their hand early. less likely

Enhanced MobileMe

This one’s more of a wish. They need to either add something else compelling to MobileMe, or severely drop the price, to get me to even consider it. Perhaps some sort of cloud storage for documents, like iDisk, but with heavy integration with the OS and iWork.com would do it for me, if it worked better than dropbox. less likely

Comments Off on iPhone OS 4.0 Predictions

Multiple Tasks

Posted in Spritely by Dimitri on April 5, 2010

I’ll be posting a bit on my experiences on the iPad in the coming days, but with the announcement of iPhone OS 4.0 coming this Thursday, I wanted to make sure to get my predictions filed for how I think it will handle multitasking first.

A number of blogs are carrying reports of a rumor that iPhone OS 4.0 will support multitasking in a similar manner to exposé in Mac OS X. I think that this rumor is fake, and I very much doubt that this will be the way that Apple chooses to implement this important and ever-requested feature.

Essentially the rumor states that by tapping the home button twice, the user is presented with all the “running” apps, exposé-style. Each running app is represented by it’s icon.

Here’s the first problem with this description: it is essentially describing what the home screen already is. If you remove your preconceived notion of multitasking, and instead replace it with the term “task switching”, you will see that the iPhone OS already does this: the home screen allows users to switch between tasks.

But I’m getting ahead of myself. What most people think of as multitasking is essentially 2 things:
1. Quickly switching between tasks, without losing one’s place.
2. Running multiple processes so that they can all do “work” at the same time, even if the user isn’t interacting with them at the moment

The first part is what you do already with the iPhone OS home screen. Tap the home button and you see your apps. Choose an app, and you will be in that app, and the app will (or at least is supposed to) remember where you left off.

Implementing a separate exposé-like task switcher would only serve to confuse. Why restrict it to only running processes? Why not just show the home screen?

This is how I predict that Apple will handle task switching on the iPhone, for the foreseeable future: they will bring the user back to their home screen, where they can select an app. It’s brilliantly simple: you choose an app once, and only in the same place. Want to pull up Google Maps? Drop back to the home screen, choose maps, and you are running Google maps.

For a modal OS, this is the type of interface that just makes sense to people intuitively. And it sounds MUCH like the rumored exposé screen, but with one major difference: the user won’t have to consider what apps are running and what apps aren’t.

I’m not saying that Apple won’t implement a faster way to bring up the home screen. I bet they will. Perhaps they will allow the home screen to slide over the existing app, and make it a toggle, so that users can drop back to their app should they decide not to change tasks.

Many people will argue that it’s too slow to launch apps from the flash memory each time, and it’s better to keep them running in RAM. I tend to agree with you. Even though the 3GS, the iPad, and whatever will come next have a faster and faster launch process, it still won’t be the instant-switch that keeping a program in RAM allows. But here’s the important distinction: Apple will implement this using a system wherein the user won’t know whether an app is running when they pick it from the home screen. For simple task-switching, it doesn’t matter whether it’s running or not. If it’s running, the change will be nearly instant. If it’s not, the change might take a second.

Think about all that this method of task switching affords you. You don’t have to quit apps. You don’t have to think about what’s running in the background, and if it’s affecting your foreground app’s performance, because the OS will manage that for you and quit the apps it needs to. And when you are searching your email and decide to look something up on Google, then come back to email, the second switch will be instant, as will all subsequent task switches between Safari and Mail. A resource-intensive app, like a game, would likely save your progress and quit altogether, as Apple would likely place strict rules on the resource usage of background apps.

But what about background apps you want to keep running, you ask? Like that GPS app that keeps a log of where you walked, or Pandora, which you like to keep playing music?

For these apps, I think that Apple will implement some sort of resource-saving mode that apps can go into, as well as a confirmation (much like push) that the user wants to keep the app running in the background. Here’s how I imagine it:

1. When opening an app for the first time, the iPhone asks the user whether they would like this app to be able to run in the background.
2. When the user initiates a task-switch on the home screen, or the OS decides to quit the program to free up resources, the OS sends the app a warning.
3. On receipt of this warning, the app puts itself into low-resource mode. It destroys all of it’s UI’s, frees in-memory caches, etc. It only keeps running the essential task, like playing music or logging GPS coordinates. Even actions such as polling the network or searching for the phone’s location would be reduced in frequency, to prevent the app from hogging battery.
4. The OS confirms that the app is within it’s limits for memory usage, processor time, and network utilization. If it gets greedy, it gets shut down. Greedy apps wouldn’t be approved by Apple, as this would be considered a crash, and reports would be filed.

An alternative to this, and perhaps a more likely scenario, is that an app could become 2 running processes. One would be a lightweight service, designed to perform only the most basic tasks (audio playback, network access, etc) and the other is for the User Interface. When the iPhone determined it needed to free memory, it would simply quit the UI app, leaving the helper app running.

If the rumors are true, I’ll be surprised. Then again, I’m regularly surprised by what Apple does. But the current rumor of “it’s just like exposé” feels a bit too limited in scope to be true. Apple knows it needs to keep these devices simple. In fact, that’s Apple’s whole “post-PC” strategy, to take what we’ve learned over the past 30 years about computers and to reinvent them so that the actions of the “computer” become an abstraction, and working with one of these machines is working directly with your content. A major point of confusion for many people who use the computers of today is task management, and i think it’s very clear that Apple has decided to attack this problem head-on in the iPhone OS.

What is this “HD” business anyway?

Posted in Spritely by Dimitri on April 1, 2010

So many iPhone OS games and apps have taken to postfixing “HD” to their names to represent the iPad. Technically, the iPad isn’t a traditional “HD” resolution (it’s taller but less wide than 720p). But in looking at screen shots today on the newly opened iTunes App Store for iPad, I realized that these apps really do feel HD. High Def and Hugely Different. And that’s not just a comparison to their smaller iPhone brethren, it’s also a comparison to the world’s most frequently used app: websites.

Most websites are cautious: they make sure that they can fit within a width of 800, 900, or 1000 pixels, to insure that they are usable by all screen sizes. The iPad is 1024 pixels wide in portrait, which nearly all of these HD apps fill with their content. Take the New York Times for example:

On the left, the NYT website. On the right, the iPad app for the NYT. The green areas are content.

On the website, 44% of the screen is occupied by the requested content. The rest is ads, navigation, branding, etc. And I’ve even removed the additional space added by the browser’s navigation bar, tabs, and bookmarks.

On the iPad, 88% of the screen is occupied by the requested content. It fills the entire width of the screen and nearly all the way to the top and bottom. A few navigational elements sit on the top, but the screen is essentially showing all content.

HD indeed.

Comments Off on What is this “HD” business anyway?

AT&T Femtocell

Posted in Spritely by Dimitri on March 29, 2010

Living in NYC, I have to deal with the now notorious AT&T coverage. Dropped calls. Never-ending web page loads. AT&T has finally admitted that they are having these problems, and are looking for a way to improve them.

One of the ways you can take it upon yourself to help improve your reception is with AT&T’s own “Femtocell” box. This is basically a miniature cell tower that you can put into your house, and route all of your calls, incoming or outgoing, over VOIP. It turns no-bar, low-bar, or inconsistent reception into perfect, full-bar reception. And for $150, it’s not TOO expensive: it’s about 1.5 times what I pay AT&T every month. AT&T doesn’t require a contract to use this (your minutes are used, even at home), but if you talk a lot at home and want to step down your minute package, you can also get unlimited minutes while at home for $20/month (and save $100 on the box). If you have AT&T home services already, that monthly fee drops to $10 or even free (if you have voice+internet).

Now I personally agree with TechCrunch and CNET, that AT&T should provide these boxes free or reduced costs to customers who have complained or who live in their admittedly “bad” areas, like NYC and SF. By getting these into the hands of tens of thousands of people, rather than hundreds, they will help offload the use of the local towers onto free-to-AT&T local internet connections.

But what’s surprising to me is that AT&T didn’t go for a bigger play here. For years Verizon has touted that you can get “One Bill” for your cell phone and home phone. That’s nothing. Imagine that AT&T offered their femtocell as a proper VOIP box, complete with the ability to port your home number. Doing this would give users a home phone number that would ring traditional house phones as well as cell phones, Google Voice style. And since they are doing all the switching on their end, why not allow these home phone calls to be taken anywhere? By piggybacking on the user’s existing internet connection, AT&T reduces their own costs, and makes their customers happier.

I can imagine a family that pays AT&T for 5 lines: a home line, which rings the cheap plastic handsets throughout the home, and a line each for mom, dad, and the 2 kids. When the family is on the go, their cell phones switch automatically over to using the cell network. When at home, all of their calls are routed over VOIP, and they pay a small fee to AT&T for the unlimited talk time. Additionally, the home line can be set to ring both the parents when neither of the parents are at home. This sort of scenario is happening anyway, but without the convenience of multiple lines: the rule in many homes is that kids use the cheap VOIP line when chatting with friends to save on huge cell phone bills.

And why femtocell? Years ago T-Mobile had a technology called @home that worked over WiFi. With all smartphones these days including WiFi, it makes more sense that AT&T and manufacturers would simply allow our phones to switch over to using WiFi when in range of our routers, without the need of additional hardware.

Comments Off on AT&T Femtocell

Rework vs. Strippers and Steaks

Posted in Spritely by Dimitri on March 20, 2010

Amazing interview, Jason Calacanacis (weblogs, mahalo) argues for the “traditional” VC model, while David Heinemeier Hansson argues to make simple, profitable companies. I’m reading Rework right now, and it’s fantastic. Jump to 47 minutes.

Comments Off on Rework vs. Strippers and Steaks

Making the iPad decision, visually

Posted in Spritely by Dimitri on March 20, 2010

I was having a hard time deciding which model to buy (3G or non) and came up with this handy chart.

I think I’ve decided on going with a non-3G to start. I think the chart shows that it’s not a big deal if you decide in the future to upgrade to 3G. If you aren’t sure whether you are going to use 3G, consider the getting a non-3G, then selling and upgrading if you need it.

And the $15 plan is the way to go. Keep your iPad on that every month. If you hit your cap, upgrade FOR THAT MONTH, then downgrade for the next month. You’ll save money, guaranteed.

Comments Off on Making the iPad decision, visually