Wednesday, November 8, 2017

Apple's AR Strategy

Back in 2015 I wrote a post titled Apple's Next Big Thing.  In it, I mused about Apple's strategy with wearables and, while doing so, predicted the introduction of AirPods.  While they not yet have all the features I forecast - foremost among them the current lack of an 'always on' capability - I think the general ideas expressed in the article are still valid: that Apple is adding and integrating ever more wearables to its product line in order to seamlessly blend the virtual world with the real one.  In some not-too-distant future, the iPhone will rarely leave your pocket - yet manage all your computing needs.

Once one realizes this, Apple's "next big thing" becomes quite obvious: AR glasses.  Apple already laid the groundwork by its introduction of ARKit.  This software - which already has thousands of developers creating new experiences for hundreds of millions of iPhones running iOS 11 - is paving the way for the day the glasses arrive.  Thousands of AR apps will be ready to project their content onto the glasses rather than your iPhone's screen.

It doesn't take a rocket scientist to predict AR glasses.  After all, Google already developed a pair a couple years ago, Snapchat tried its (limited) hand at it.   But as with all other portable electronic gadgets, it will be Apple to do it right.  So what is "right"?  To answer that, we again look back at Google Glass - what were its shortcomings?  Well, there was the now infamous recording feature that caused its wearers to eventually receive the label of 'glassholes'.  So Apple's version will avoid recording video - at least until AR glasses have become ubiquitous.  Google Glass also had notoriously short battery life.  Why?  Because Google tried to do too much "on-board".  Instead of being a companion device, Google tried to stuff a complete computer - i.e. CPU, memory, video controller, storage, networking, etc. - into the temples of the glass frame.  Since glasses have to remain light in order to be worn comfortably, the amount of battery capacity they could stuff into the glasses was too limited.  Therefore, I predict that Apple's AR glasses will have just enough electronics to let it receive instructions on what pixels to project onto the glasses and to send onboard sensor/camera data back to an iPhone.  It might even use its existing AirPlay tech for the display aspects.

Friday, December 11, 2015

The Politics of Fear

We are a nation with enough guns in circulation to arm every man, woman, and child.  Yet we are fearful of a handful of terrorists on the other side of the earth.  We are so fearful, we let our leaders trample on our privacy in the name of keeping us safe.  14 deaths by terrorists cause us to contemplate closing our borders and hearts to refuges fleeing for their lives.  A handful of deaths by terrorists scare us so much, yet the deaths of thousands every year by guns and other "American" violence leave us unmoved.

It is a strange and sad thing to behold in a country founded by immigrants whose battle cry seemed to always be "give me liberty or give me death."  Has "give me liberty" been reduced to "let me own guns"?  The right to bear arms was meant to help us defend all those other liberties & freedoms we cherish.  Such as the right to privacy.  Freedom of speech.

And politicians - some running for government, many already in government - use this fear to their advantage.  Mr. Trump is a veritable virtuoso at it.  Don't let them.  Terrorism is no different than all the other aberant deeds we witness every day.  We deal with it - we don't hand over our liberties to politicans promising illusory safety.

Friday, October 2, 2015

Apple's Next "Big" Thing?

Apple has recently stepped into the wearables arena with the introduction of the Apple Watch.  While the success of the watch is still unclear, its capabilities give us a hint about Apple's thinking in this arena:

The Apple Watch is a "companion" product.  Although it can operate independently, it only reaches its full potential when within bluetooth/wifi range of its master, the iPhone.  With the watch, Apple is selling convenience and the watch is only the beginning.  What could be Apple's next foray into the wearables market?

Bluetooth in-ear headphone(s).  It makes perfect sense.

Siri, iPhone's voice assistant has come a long way since it was introduced on the iPhone 4s.  Over time, Apple has not only improved its voice recognition dramatically, it has also broadened the ways in which you interact with her.  You can hold down the Home button on your iPhone or the crown on the Apple Watch to talk to her; if you plug in your iPhone into a power source - or if you hold your Apple Watch up to your face -  you can get her attention by simply saying "Hey Siri", followed by your spoken question or command.  Heck, thanks to a new low-power chip in the iPhone 6s, you don't even need to plug in your iPhone to chat with her.  But the form of Siri's answers aren't always convenient.

When you ask Siri a question on an Apple Watch, she's a mute.  I guess Apple Watch's designers felt that her speaking the answers - as she does on the iPhone - would be too obtrusive in a setting where you chose to interact through the watch rather than the iPhone.  So Siri, on the Apple Watch, displays its answer on that tiny watch screen.  Which isn't always optimal - e.g. when you're driving.

Using a headset would be the perfect solution.  But why should Apple come out with a new pair, when there are loads of wired and bluetooth headsets on the market already?  Because none of them have the feature Apple just introduced with the iPhone 6s: the aforementioned "always on" listening for "Hey Siri"!

With an unobtrusive headset that always listened for "hey siri", life would be so convenient!  No raising of a watch to your face a la Dick Tracy; no more fishing your phone out of your pocket to look something up!  With the headset always in your ear, you simply ask what you want, preceded by a quiet "Hey Siri" and Siri whispers you the answer.

There are so many things that could be facilitated by such a device.  I can imagine, for example, a situation in which I need to interact with my Chinese in-laws.  I could simply put my iPhone or my arm with Apple Watch on the table and whisper to Siri "Hey Siri, begin translating."  When I subsequently say something, the Chinese translation pops out of the iPhone or Apple Watch, for my in-laws to hear.  When they respond, the watch or iPhone pick it up and Siri whispers the translation into my ears!  This is getting pretty darn close to Star Trek's universal translator, isn't it?  The technology is all here!

And there are some promising signs that Apple is, indeed, thinking in that direction.  A year ago, Apple bought Beats Audio.  Granted, this was likely about acquiring that company's music talent and streaming service, but Apple likely also picked up some serious audio expertise in the transaction.   Then there was yesterday's news that apparently a shell company (i.e. a company that doesn't really produce anything and whose ownership is hidden) copyrighted the name "Airpods" in several countries.  The law firms handling this company's business just happen to be the same ones Apple has previously used when using shell companies to hide its tracks.  Hm....

Hey Apple, if you hear this: will you give me a pair of Airpods, if my predictions prove to be right?

Thursday, November 6, 2014

E-Trade Bill Pay - A Study in Bad User Interface Design

I love E-Trade.  I’ve been a customer since the late 90’s.  Heck, I was a customer of E-Trade’s predecessor/progenitor: K. Aufhauser.  Why do I like it so much?  Because I’ve always found E-Trade’s web site to be head and shoulders above the rest.  Everything in its user interface (UI) was arranged in a clear-cut manner and there was a consistent “theme” and usage pattern to every page across the site.

Alas, recently the management at E-Trade must have decided that their site wasn’t good enough and they began “improving” it.  A few months ago, E-Trade programmers modified the online bill-pay feature - a feature that was already perfect, as far as I was concerned!  The original page had two “columns”: a wide one on the left, which simply listed all of ones registered payees  with a text field in which to enter the amount to be paid and a calendar widget through which to specify a date of payment, and a fairly narrow one on the right, which showed a historical record of scheduled and previously made payments.  Very simple.  Very elegant.  In my particular browser window, I could see/edit 20+ payees and see a record of 20+ payments.  Whatever I didn’t see, was just a single scroll away.  I wish I had a screen shot of the original user interface.  But since E-Trade’s first modification came without notice, I didn’t have a chance to get one for comparison.

This first modification was not as usable as the old one, as far as I was concerned as things had begun to “spread out” - apparently the UI designers felt that white space was more important than information.  But, alas, I don’t have a screen shot of that either - since that particular experiment didn’t last too long.  It was quickly replaced with the even uglier, even less usable user interface that presents itself today.  Since I’m just about fed up (especially since E-Trade is not threatening with an overhaul of the entire system - based on what I’ve seen in the Bill Pay area, I shutter to think what awaits me now), I decided to take screen shots this time and elaborate on what I feel are its most annoying shortcomings.

As an aside, I should mention that I have 25+ years of software development experience - much of which in UI design - so I may be a little more critical than the average user.  But my comments are solely based on my use of the system - which has suffered greatly from the changes that were made.

Let’s start off with a screen shot of the screen that precedes the “Bill Pay” page:

Note the look of the tabs for both E-Trade’s overall tab group and the “Account Details” tab group (I put a red “1.” next to both): the outline and internal shading of the tabs is identical and the fonts are the same (the lower tab uses a slightly smaller font size).  There is a consistency of theme there.  Also note the size of the font used in the table that shows transactions and the row height for each transaction.  The original Bill Pay’s “Payee” table where you entered payments used to have row heights that were similar.

Now, let’s go to the “Bill Pay” page:
Quite a jarring change, wouldn’t you agree?  The tabs for the “Bill Pay” tab group (red “1.???”) look like they were drawn by a 10-year old and their shading (none) is completely different from the E-Trade tab group shown just above (red “1.”)!  Even the fonts are different (and larger, wasting space). In other words, the page’s theme isn’t consistent.  And the tabs are just plain ugly.  But that’s a trivial shortcoming compared to the others.

The row height (see “2.” on image above) for each payee is humongous.  As mentioned before, I used to be able to see - WITHOUT SCROLLING - about 20 payees in the old Bill Pay.  Now I see seven!  Why is that important?  Because most of us try to pay our bills once a month.  It is, therefore, nice to be able to simply click on the “Amount” field for each payee (and maybe specify a later date) and then hit “Submit”.  In the old Bill Pay, I could do all this without ever scrolling.  Now, I can’t avoid scrolling to pay everyone, scrolling to re-check the amounts I entered, and then scrolling some more to get to the “Submit” button.

So why is each payee row so tall?  Well, the only answers I could come up with are (a) a love for white space and (b) to accomodate the drop-down menu that lets one specify the money source (account) for each payment.  I won’t discuss (a), but (b) merits further inspection:

Let’s first review how we got to this “Bill Pay” page in the first place.  E-Trade, on its main page (“Account Overview”) lets one pick an account to work with.  That leads to the “Account Details” page shown in the first image above.  From there one clicks the “Bill Pay” tab to get to the second image.  So by virtue of how one got to “Bill Pay”, one has already decided which account to pay with!  So, why is there a need to have an “Account” combo box for each payee when most folks just want to pay their bills from the account they’re in?!

Why did E-Trade lower usability for the most common use case in order to satisfy a really infrequent one???  Poor thinking.
A much better design (which E-Trade actually uses in the “Account Details” image 1): simply put a single Account drop down at the top of the “Bill Payee” page, instead of one for every payee row!  For the common use case, users get to see/pay more of their payees without scrolling and for those who need to pay some folks with one account and others with another, they would simply hit “Submit” more often (once for every account from which payments need to be made).

Other than the oversized rows, the new design simply wastes a lot of screen real estate that should be used to show information (see labels “3.” in image above).  The most glaring example of this is in the tab group on the right.  The most useful one in that group is the “Activity” tab (don’t ask me why E-Trade designers didn’t make that the first tab, given that it’s the one most people need to look at most frequently!  Or why they do some inane auto rotation that, if you leave the tab group alone, auto-rotates focus through all the tabs - very annoying and zero value!)  Anyway the “Activity” tab shows historical payment activity as well scheduled payments.  It’s basically a table with three columns: “Send on”, “Paid to” and “Amount/Status”.  If the developer of this screen had done his job, this table would stretch with the available browser space, instead of fixing it at some arbitrary width (see all the white space in the image, for christ sake!) that requires the column content to spill over a single line - again reducing the number of rows that can be shown without scrolling, and making the whole thing just less readable.  But the waste of space doesn’t stop there.  Why, in gods name, is this tab group not the same height as the payee panel to its left - i.e. why doesn’t it make use of the web page, like the payee panel???   Now, in addition to only seeing SIX activities (due to row height), I need to scroll the “Activity” tab independently of the web page (see labels “4.” in image)!
(the fact that in Mac OS X you don’t see scroll bars until you hover over them doesn’t help the situation).

Just poor, poor design.

Sunday, February 17, 2013

iWatch - What is Possible?

Analysts and Apple fans have recently become enamored with the idea of an iWatch.  Bruce Tognazzini, a respected ex-Apple employee has an excellent blog post in which he describes all sorts of smart uses/features for such a watch.  As anticipation for such a device is building, so are the expectations for its features.  Some have come to see it as essentially an iPhone with a small screen.  I think those folks will be disappointed.  Think about it: we're talking about a device that needs to be thinner than an iPhone and, at most, 4 square inches in size.  There is no room for much of a battery!  Add to that the fact that, historically, people don't expect to have to recharge their watches every day, and you begin to see how impossible a "mini-iPhone" iWatch is.

So the features of an iWatch must revolve around power: how to minimize consumption and/or how to provide more of it without taking the watch off.

What can be done to keep power consumption to a minimum?  The two largest power hogs in a mobile device are its screen and the various radios driving communications (wifi, cellular, bluetooth).  Screen power consumption can be minimized by using a technology used in e-Readers (and in the recent "Pebble" smart watch): e-Ink.  With e-Ink, power is only needed when the screen needs to change.  But the technology, by its nature, has two drawbacks: slow refresh rates (making video impossible) and lack of color.  If Apple deems color and video playback indispensable, its option is OLED.  Reducing power on the various radios is easy: simply limit communication to one low-power channel.  I believe that is Bluetooth 4.0.  Of course, if the iWatch doesn't have wifi built in, it will not be a "connected" standalone device.  Apple's choice.

So, if the iWatch uses an e-Ink display and Bluetooth 4.0 for communication with an iPhone, what could such a device do?  Well, it could tell time! :-)  Also, like the Pebble, it could act as a remote display for all the phone's notifications.  For notifications requiring action, the watch's microphone could be used to tell Siri (located on the phone) what to do.  Similarly, the iWatch could display (or speak) results of any spoken Siri query.  The built-in speakers and microphone could also be used to make or answer phone calls (like today's Bluetooth headsets).  The bottom of the iWatch and/or the watchband could contain sensors that keep track of basic biometric info (e.g. pulse rate) - perhaps something similar to what Nike's Fuelband does - and send it to the phone occasionally (where an app can let you analyze the data over time).  Finally, an e-ink based iWatch could display OCR codes from Passbook.  When at a "favorite" store, the iWatch would automatically display the store's pass.  No need to fish for the phone or wallet!  Some analysts/users propose that the iWatch should have an NFC chip for payments.  I think this might be a future addition to the watch - there aren't enough (any?) NFC-based cash registers out there to warrant the additional complexity (and power drain) in the watch.  How long between charges?  Well, we have a couple examples that give us a clue: the Pebble smart watch lasts 7 days.  The Nike Fuelband is said to last 4 days.  So Apple, with its engineering prowess and deep R&D pockets, should be able to produce a watch that lasts at least as long as the Pebble.

If Apple wants to endow its watch with a color display and the ability to carry on FaceTime video conversations or watch Youtube videos, e-Ink is out of the question - an OLED display becomes necessary (it uses less power than LCD and can be put on a flexible substrate to allow for a curved display).  If the watch has a built-in accelerometer, intelligent power management software could turn off the display whenever its wearers' arm is hanging by his/her side and gets activated only by the touch of a finger or by an incoming notification.  Perhaps, this way, the iWatch could still last 5-7 days between charges.

But no matter how hard Apple tries, an iWatch will have to be charged on at least a weekly basis.  What can Apple do to make this as painless as possible?  Some manufacturers have begun offering wireless charging pads (and Apple owns a patent on doing this over the distance of up to 1 meter) that might be useful here.  If one didn't have to plug in yet another device and, instead, simply placed the watch, along with ones iPhone, on such a pad for charging at night, people might not mind!  Heck, if they can pull off that feat, maybe an iWatch only really needs to last as long as the partner iPhone!

Saturday, January 5, 2013

Let's Save The E-Reader!

Dear Mr. Bezos,

I recently saw statistics which suggest that the e-reader has had its hey day in 2011 - apparently sales for 2012 are down for the first time in the technology's history!  That saddens me.  I'm an avid reader and a great fan of the Kindle line of e-readers (we have two in our household and I gifted a couple more to relatives).  But I understand why sales are down: the tablet is eating the e-readers' lunch!  People apparently prefer these multi-purpose devices over the single-purpose e-readers.

But that should come as no surprise to anyone - history is full of examples where multi-use devices eliminated superior single-purpose competitors.  Most famously, the cell phone: a multi-purpose device that doesn't do all that great a job making phone calls.  Yet it has largely supplanted land-line phones.  It doesn't take great pictures, yet it has largely eliminated the point-and-shoot camera!  It does a so-so job browsing the internet, yet more and more web surfing is done via mobiles rather than desktops.  The situation is identical for e-readers versus tablets: tablets are inferior text-reading devices - I don't think anyone who has used an e-reader would prefer an LCD-based tablet for any lengthy reading - but people are willing to put up with a mediocre reading experience, because they can also use the same tablet to surf the web and enjoy rich media (e.g. audio and video).  Last, but not least, they can use tablets to run useful general-purpose apps (e-mail, facebook, Words-with-Friends, etc.)  People don't want to fill their lives with single-purpose devices!  So the e-reader is headed for the museum, right?

Not so fast!  Why, exactly, is the e-reader a single-purpose device?  I understand that e-ink technology prevents it from being a video player, but why can't it be a good web browser?  Why can't it be a great audio player?  Why can't it host useful apps?  There is no technical reason!

I have a 3rd-generation Kindle (the one that still has a keyboard).  Sure, it has a rudimentary web browser, but because it lacks a touch screen and because the CPU is too weak to render pages in a reasonable period of time, surfing the web is neigh impossible.  Newer generation Kindles have added touch screens, but have not addressed the slowness issue - so surfing the web on a Kindle is still much more painful than on any tablet.  But this should be a simple issue to address: simply improve the software and use a quicker CPU (even if that is at a slight cost to battery longevity).  It's not rocket science!   Making the Kindle an excellent audio player should be even less of a problem.  It already plays music quite well - the Kindle's user interface is simply too horrible to operate the audio player in a convenient way, so most people don't bother - or even know it exists!  So get some decent UI developers (which are also sorely needed in other areas of the Kindle's operation) to make the audio player easier to access/use!  Finally, there is the need to run apps.  Kindle e-readers are based on Linux and Amazon has had a "Kindle Developer Kit"(KDK) available for a long time, but to-date, only a handful of apps exist.  Why?  It can't be the programming language - the KDK is based on Java, the most widely used programming language in the world.  I suspect developers aren't flocking to the Kindle platform because the API is too limited which, in turn, is probably because the CPU is underpowered and can't handle sophisticated apps.  So the solution is a brawnier CPU and, possibly, a beefier API.  Again, no rocket science involved.

Making these relatively simple changes, would allow folks to use the Kindle a lot more.  Sure, it still won't let them watch YouTube videos or Netflix content, but they'd be able to peruse the mountains of static content on the Web, to visit social sites such as Facebook & Google+, and to read e-mail.  They could listen to their music while they do these things.  E-readers could, for most users, be the tablet-of-choice, since they would be easier on the eyes while reading and surfing the web, yet still allow for decent audio and app experiences.  Users who want to watch video or enjoy color on the go could still do so with their cellphones.  Tablets could be headed for the museum! :-)

I'm an avid reader and a news/web junkie.  Because the Kindle is not usable in the latter contexts, I've gotten used to using my iPhone and, when my wife isn't looking, the family iPad - the Kindle is collecting dust most of the time.  But my eyes are killing me!  I'd much rather surf the web using e-ink! If future Kindles had the capabilities I outlined, I would immediately buy e-readers for the whole family!

In summary, please build a Kindle with the following specs:

  • E-ink, touch-sensitive display in a variety of sizes: 6" is fine for books, for newspapers and magazines with their natural multi-column layouts, a bigger screen would be very nice.
  • Hard buttons for power, page forward/backward (I have no idea why such a useful feature for reading was removed in the latest Kindles - it allows for one-handed reading!), volume, and maybe wifi (to preserve power more easily)
  • A CPU and software fast enough to support fast web surfing and sophisticated 3rd-party apps
  • An intuitive user interface for reading, surfing the web, playing apps, and listening to audio
Best regards,
A reader and web surfer with tired eyes

Wednesday, December 19, 2012

Apple TV - My Predictions

For almost a year now, rumors of an impending Apple TV have been hitting the various news & technology sites.  More recently it's even factored into many financial analysts' views on the future of Apple stock.  But in all this time, I've not seen a detailed description of what such a TV might be able to do that is different from today's existing hardware.  Many people think that it's main feature (aside from Apple's vaunted ease-of-use) will be the ability to subscribe to individual cable channels - a feature that will depend entirely on Apple being able to strike a deal with the cable companies.  So I thought to myself: hey, you're into technology and are using most of Apple's existing gadgets and various existing media/streaming sites - why not use that knowledge to describe what you think an Apple TV might look like.  So here I go :-)

First, the obvious: physical design & price.  I think Apple will continue its existing design language and make the TV look like a larger iMac or Cinema Display.  The new iMacs are razor thin at the edges and clad in aluminum like all Apple devices are these days.  I expect a minimum or non-existent bezel around the screen - i.e. the front will be all black with the back enclosure aluminum.  The stand will likely also be an aluminum contraption similar to the Cinema Displays.  There will probably be two or three sizes (42", 55", and maybe something even larger, to accommodate its more affluent customers who have gargantuan rooms).  Price-wise, I expect the 55" version to be around $2500-$3500 (based on the fact that the top-of-the-line Samsung LED TVs cost that much).

Next, the user interface.  Apple already has an "Apple TV" product that might give a hint as to how one might interact with the actual TV: basically a grid of icons that get you to the content you want.  Since the remote on the existing Apple TV product, while sexy, is somewhat lacking in basic usability (if you ever tried to enter a search keyword using its onscreen alphabetic keyboard, you'll understand what I mean), the TV will come with an iPodTouch-like remote control that duplicates what you see on the TV.   Essentially "mirroring" (a feature that newer Apple devices already support) in reverse.  There will be an app for existing iDevices that can do the same thing - so any existing iPod, iPhone, or iPad can become an additional remote control for the TV!

What about content?  Well, some of the content will be exactly what you already get with the existing Apple TV box: access to iTunes content, Netflix, Pandora, etc.  But, no doubt, there will also be real-time TV channels to choose from.  How these channels get into the Apple TV is an interesting bit to conjecture on.  It could be that there will be the traditional coax connector in the back of the TV and either software or a hardware "module" from your local Cable provider that will emulate the cable box (sort of like the failed CableCARD standard of yesteryear).  But I don't see Apple allowing something as retro as a coax cable to come into its beautiful devices - Apple is renowned for simplifying the cabling to its devices - why would they want to propagate such an antiquated connection?   I think Apple will do all the work in its data centers and simply stream the content to your TV.  The channels you "see" will depend on what deals Apple can strike with which content providers and what you subscribe to.  So if Apple strikes a deal only with Comcast and your current cable company is Cablevision, the programming you'd see on your Apple TV will be Comcast's rather than Cablevision's (so you'd use Cablevision only for your Internet).  If Apple strikes a deal with both cable companies, you might even get a choice on which cable company gets to bill you!  Although, either way, you'd simply be paying through your iTunes account :-)  Finally, depending on what deals Apple can strike with these cable companies, you'd either continue to have to buy "packages" of channels or you might get to pick individual channels.  I hope it's the latter - I only really watch a couple channels - why am I subsidizing all the other crap that's in my "package"?
The channel guide will be on a "time line", initially centered around the current time.  But you can slide backwards in time and access TV programs that have already happened (i.e. no more DVR needed).  Maybe those shows will even have advertising automatically (or, more likely, for a price) stripped.

What am I missing?  Oh yeah, since it's an iOS device, it'll have all the existing benefits of being part of the Apple ecosystem: any Apple device can stream content to the TV (AirPlay, Mirroring).  Apple might also figure out a way for multiple iOS devices to share the same Apple TV - thus allowing for multiple player games on the same big screen!

Well, that about does it.  What do you think?