Archive for the Blog Category

Two Years Later: HSPA+ Comes to Canada

Two years ago this month I wrote a piece about Bell & Telus’ shiny new shared HSPA+ network (here October 7, 2009); let’s take a reality check to see how much has changed since then.

In that piece I explained that Bell and Telus had separate CDMA networks which they would be abandoning in favour of a new HSPA+ network, which has happened very smoothly.  At the time Rogers had covered about 84% of the Canadian population with HPSA (3.6 and 7.2mbps) and had just opened their first city with some HSPA+ support (Ottawa, if I recall correctly) and over the next six months upgraded many other cities that had existing 3G service to HSPA+ (in May 2010).  Rogers has also since expanded their 3G coverage to include the province of Manitoba (through a partnership with TBayTel) and some other areas on Ontario, thus leaving the Maritimes with unchanged 3G service areas (though they promise this will change “soon”).  Neither Bell, Telus, or their subsidiary brands have offered Rogers customers a means out of their existing contracts.  Rogers has also launched LTE 4G in 4 cities in 2011: Ottawa, Toronto, Montreal, and Vancouver.  Rogers data packages are typically 500MB or 1GB for the two most common smartphone packages; LTE starts at $52.93/10 GB.

Bell and Telus have also had a busy few years; let’s start with Bell.  For years Bell had a 50% share in Virgin Mobile Canada, a MVNO that used Bell towers and cellular equipment on the “no-frills” wireless service plans offered by Virgin.  In 2009 BCE Inc. bought out the remaining shares of Virgin Mobile Canada to make it a full subsidiary company and began to offer HSPA/HSPA+ phones in late 2009.  Today, Virgin has very competitive phones and plans, and is a serious wireless provider that offers “tab” service rather than traditional termed contracts (whereby 10% of your bill each month pays off the balance remaining on the purchase cost of your phone).  Bell still maintains Solo Mobile in central/western Canada, though they’ve fallen by the wayside in light of the Virgin brand.  Bell continues to activate, sell, and service CDMA phones and has no current plans to discontinue service for these phones in the near future, though they do no longer service analog phones.  The average package with Bell and Virgin is 1GB with the typical smartphone package ($60 and $65/mo respectively).

Telus has also been busy: Koodo, a “no-frills” affordable service provider they created in 2008 to attract younger, cost-conscious consumers has taken off; as of today they offer a vast selection of smartphones with data plans starting at $5 (a flexi-plan that increases with usage) as well as a 2GB/$30 plan which far-and-beyond beats typical dataplans with Bell, Telus, Rogers, and Virgin.  Telus now offers a nearly HSPA-only line of phones, with a selection of HSPA+ handsets.  I had imagined that there would be more of these phones by now, however they’ve only really come out in great numbers in the past year after AT&T announced their use of HSPA+.  Typical data packages remain greatly unchanged for Telus, with approximately 1GB being the average allotment for most consumers (as is bundled with their $65/mo smartphone package).

The strange development in Canada in the past two years has been new entrants with HSPA+ on the AWS (T-Mobile) bands, notably Wind Mobile and Public Mobile.  Wind greatly focuses on unlimited services, such as unlimited data, messaginng, and calling within Wind-zones (where Wind has towers; the phones will still function outside of these, though will be roaming with increased airtime rates).  Rogers has launches a competing brand, Chatr, aimed at this “unlimited  in zones” mentality which is a scam: Wind has to do “zones” because they only have towers in those areas yet still wish to offer nationwide service.  Rogers has nationwide service (to some degree) and arbitrarily enforces “zones” on Chatr customers.  Wind has now accumulated more than 317,000 subscribers (here) and is still going strong.

In the near-to-mid-term future we have the Maritimes’ largest cable provider, Eastlink, launching some sort of cellular network.  Currently Eastlink offers a Motorola Canopy network for rural broadband across Nova Scotia and is partnered with Rogers Wireless to offer discounts for bundle customers.  Eastlink has cable/internet/home phone service in nine provinces, though is most predominant in Atlantic Canada (Nova Scotia, New Brunswick, PEI, and Newfoundland).  Given that Bell or Telus are the only real options for cellular service in the Maritimes, there are high hopes for Eastlinks forthcoming service.  Notably, in 2005 Eastlink began offering home phone service and made the 902 area code (for Nova Scotia, New Brunswick, and Prince Edward Island) the most competitive in North America.  Hopefully the 902 area code will once again be the most competitive in 2012 with cellular service.

 

Who wants a Palm Pre?

I recently published my review on the Palm Pre 2 and offered the verdict that it was a $199 feature phone with great font rendering (which is really true!).  I acquired the phone in early July and went to sell it in late July and held off publishing my review until it had sold to give me time to play with it and really think about the harshness of my words.  I went to sell the phone on Kijiji, a service similar to Craigslist but is heavily used in Nova Scotia, for $200.

 

Weeks went by.  I got a few offers: “$100 if you deliver it to Cape Breton” (an eight hour drive) Nope… “$100 in Halifax” …. Nope!  So I reduced the price to $175 and finally got a reasonable offer: “$175 in Halifax if it can run on Koodo.”  Koodo is a cellular brand owned by Telus Mobility which operates on the shared Bell/Telus HSPA+ and CDMA network.  By virtue of the shared network it uses the same frequencies as Bell and Telus, who chose to use the North American standard AT&T bands for HSPA+.  We’re in luck! Rogers also uses the exact same frequencies so the Pre can work in theory.  After intent Google’ing I discovered people have had success unlocking their Palm Pre 2s and running them on Bell.  Awesome!  I replied saying that it was possible and that it would need to be unlocked, a process I’m unfamiliar with.  He said he could do that himself.

 

When I met the guy to sell the phone I discovered he was a Palm Pre (original) user.  After some talking I’d had all my fears quashed that the buyer on the other end may have mistakenly thought the Palm Pre was an Android phone, a reasonable fear given the vast fragmentation in the Android handset marketplace.  He said how he’d been an avid Pre user (on Bell, CDMA) since 2009 and loved the look and feel of webOS except that it was slow on the original 2009 Pre.

 

I got a sinking feeling.  The Palm Pre 2 does have much more hardware resources than the original Pre but it is still incredibly slow compared with the Nexus One, iPhone 3GS, and even my Blackberry Bold.  It’s by far the slowest phone I’ve used since my early days on a feature phone (the LG Keybo for what it’s worth).

 

I booted the phone up and he marvelled that it boot up so quickly.  And applications open so fast.  And typing is much more responsive.  And card view is much more fluid.  And you can open Google Maps, Music, a browser, email, and messaging cards and the phone doesn’t pop-up to say that it’s out of memory.  Wow.

 

It then dawned on me that to a fish the world is wet.  From the perspective of an original Pre owner the Pre 2 is a logical successor and is so much faster.  If you’d entered the smartphone world through webOS and never left then you’d only see the Pre 2 as fast and shiny.

Palm Pre 2 Review

I am a bit of a mobile nut.  First I had an iPhone, and quickly expanded from a single “daily driver” to a selection of fashionable smartphones for all occasions.  These include the 3GS, Bold 2, Nexus One, and the latest addition, the HP Palm Pre 2.  However I haven’t used a Palm device since a late ’90s Pilot, I can already begin to draft Palm’s obituary after two weeks on the Pre 2.  Perhaps the biggest shock was how quickly I was let down by webOS as a whole; I really did have high hopes for the “new” Palm with the Pre (at CES 2009) and have continued to triumph that webOS is the most-Apple non-Apple operating system.  First, some background.

In 2007 Jon Rubenstein, a former Apple executive responsible for trimming the Mac line-up in the late ’90s and later developing the iPod, joined Palm.  Using the invaluable experience from developing and shipping mobile devices at Apple, he was crucial to the development of Palm’s saviour movement to compete in the modern smartphone market against the iPhone, Android, and Blackberry.  webOS premiered at CES 2009 and, should it have shipped the following quarter, would have given the iPhone a run for its money.  From day one the Pre (the first device to run webOS) offered multitasking; background process management; cut, copy, and paste; and the entire feature set of contemporary iPhone (3.2MP camera with video, multitouch, 3.2mm headphone port, GPS, WiFi b/g, accelerometer, notifications, a WebKit browser, App Store, and 3G–albeit Sprint’s CDMA/EVDO).  But it didn’t.

Eighteen months later (October 2010) came the reluctant rollout of the third-generation of webOS devices, the Pre 2 at the helm.  The Pre 2 boasts a 1GHz processor, 3.1″ multitouch screen with Guerrilla glass by Corning, a sliding QWERTY keyboard, a gesture area, 512MB of RAM, 16GB of internal storage, webOS 2, 3GSM (HSPA/AT&T/Rogers) compatibility, and a $449 price tag (from Palm online store available in GSM or Verizon CDMA flavours).  Compared to the 3GS (which shipped before the original Pre) this would be a reasonably comparable device with a faster processor, twice the memory, a $150 savings, Flash 10 capability, and an awesome notification and multitasking metaphor that iOS is still trying to figure out.  The biggest sale was to developers: a Javascript API that devs already know!

The most compelling feature of any smartphone is app availability.  I’m not a believer that the number of apps available for a platform is any kind of indicator of the success or usefulness of a platform; just that you need a core set of applications that are essential to you, like Facebook, a good Twitter client, a music player, etc.  The hope with webOS was that developers would port their existing web apps (that Apple has iPhone devs craft in 2007 with their web SDK) to native webOS apps with ease.  Compared to the contemporary platforms, webOS used the highest level language in place of Java (Android, Blackberry) or C (iPhone as Objective-C is a superset of C) and supported dynamic typing, automatic memory management, and other modern language features.  That was the hope, at least.

  • iOS has had this from the early days of native apps (2008) as Facebook launched a “web application” in 2007 and a native version a year later; Tweetie (now Twitter for iPhone) by Loren Brichter launched in November of 2008 (just five months following the release of the native SDK at WWDC).
  • Android has had Facebook native since summer 2009 and numerous Twitter clients, including twidroid (now twidroyd due to legal issues) since fall 2008.
  • Blackberry had Ubertwitter (now Ubersocial due to legal issues) since 2009 and an official Twitter for Blackberry client since summer spring 2010 as well as Facebook since fall 2008.
  • webOS has no official Twitter client as of now, nor does it have any particularly useful ones either (I picked the so-called crème de la crème one, phnx for $2.99 and it is pretty immature so far).  Facebook for webOS premiered in spring 2010.

webOS is slow (as research for this entry I stumbled upon this podcast, Hypercritical, which offers some explanation and discussion).  I don’t entirely understand the reasons why it’s so slow, although the most suspect reason is that the entire native SDK is built on top of WebKit.  WebKit itself is an incredibly fast layout engine complete with Javascript support that is utilized by Safari (Mac, Windows, iOS), Dashboard (Mac OS X widget host), Chrome, Android browser, webOS browser, and Blackberry OS 6 browser.

It seems to reason the use of WebKit is not the problem, but more likely in implementation.  The kind of speed issues I’m facing remind me of the transition in browsers from (traditional) Javascript interpreters to Javascript compilers.  The primary difference is that Javascript is first compiled to native code before being run: there is an initial performance hit as it must compile before anything can happen, but it is particularly efficient at repetitive tasks (as are found throughout modern web apps).  The best case study would be the evolution from Firefox 3 to 3.5 which upgraded to a JIT compiler.  The difference in speed was somewhere between 20 and 40x.  Perhaps webOS isn’t capitalizing on modern innovations in WebKit such as JIT compilation.

But where would you see this?  The best example of the sluggishness of webOS is in switching between chats in the built-in messaging (SMS/MMS) app (which doesn’t do any kind of character counter!).  When in one chat you perform a left-slide gesture on the gesture area to return to the list of chat threads then you tap on another chat to respond.  There is a noticeable time, perhaps three quarters of a second load time to show these messages and an additional half second before typed characters are registered by the text box (typically the first couple characters you type are ignored so you get partial words that you have to correct).  Even the Blackberry, which may freeze for minutes at a time when performing any menu action, does this right: everything responds nearly instantly and characters may not appear immediately but are buffered and eventually appear in the right order.  In other words, the Blackberry understands your intentions.

All of the problems with webOS stem from immaturity.  The speed issues are reminiscent of a 1.0 software release, not a 2.0 update that should be polished.  My overall judgement is that webOS has no place in the market as developers clearly aren’t adopting the “modern” API architecture and the user experience doesn’t feel competitive with iOS, Android, or Blackberry.  It’s as if webOS developers are unaware of the goings-on of other platforms and have evolved in a vacuum without the influence of the industry.  For example, all of the Twitter clients I tested don’t do “pull to refresh,” rather use a button or “shake to refresh.”  Also, none of the clients offered to complete usernames @ mentions when composing a tweet, not even offering a selection panel like Twitter for Android; the best client did, however, allow you to reply (but not “reply to all”) and filled in the user name.  The UI for phnx has two themes: one navy blue with white text, the other black with white text, reminiscent of the early days of iOS apps, where modern apps have become lighter.  These are conventions that have been offered on most clients for the better part of a year by now but haven’t ventured to webOS.  The best way I can describe the end-to-end experience is like walking into a time machine to 2009 where the Pre was released in lieu of the iPhone.  The problem is that most people have experienced the responsiveness and feature sets of modern platforms by now and that cannot be ignored no matter how hard you try.

The Pre 2 hardware is perhaps the most quirky.  In addition to being the thickest mobile phone I’ved used since 2000, the Pre 2 is also the best fingerprint magnet I’ve ever owned.  The Guerrilla glass lacks a coating that we’ve become accustomed to since the 3GS that prevents oils from adhering to the screen.  The glass itself feels grippy and almost resists your finger gliding to perform gestures.  This leads to an almost “gross” experience as the screen looks, feels, and always is dirty.  The soft-touch plastic Palm uses is commendable and feels like it would resist scratches, cracks, and dirt.  The sliding motion of the QWERTY keyboard is excellent, except that I haven’t yet figured how to hold the phone with one hand to hit the sleep/wake button without inducing the sliding motion.  The keys themselves are tricky: Palm has recessed the keys behind a bezelled lip around the keypad that makes them unusual when coming from a touchscreen or Blackberry Bold keyboard. (There is no option to use a touch screen keyboard).

Possibly the quirkiest feature is the use of autocorrect with the physical keyboard.  Out-of-the-box the dictionary is very limited and doesn’t know common text slang such as “lol” or even common contractions (such as getting “don’t” from typing “dont”) and opts to replace these with seemingly random words (ex, by default it replaces “dont” with “weird”).  The dictionary claims that it does learn but I have yet to witness it learning, but keeping the preference pane (“Text Assist”) to edit the dictionary quickly is easy with the card view multitasking.  By far the most useful feature for typing is to set up short cuts which can be any “word” or character (but not space) that is replaced with another word, for example, by default “r” will be replaced with “are” when “r” is not connected to anything and you hit the space key.  I’ve used this short cut feature to set up common contractions, “youre” → “you’re”; “arent” → “aren’t”; “ill” → “I’ll” and more.

My verdict: webOS feels like a $199 feature phone with great font rendering.  The gesture area is not really intuitive and doesn’t provide any added function beyond Android’s physical “back” button or iOS’ software buttons.  It simply is not possible to recommend this phone to anyone given that you can get a Blackberry Curve with a much better keyboard, a much more responsive experience, and a complete social package for $50 less (retail Curve 9300 is $399 from Rogers).  That said, if HP is able to find a good dev for a Twitter client (to include/offer for free) and is able to speed up the phone to contemporary levels then this would be a different story.