The Switch: Afterword

After moving my wireless contract to Orange (.PL) the only thing that stopped working were SMSes. Really... Murphy's Law... I can receive SMS messages just fine. I can even send them out. But any message longer than just a few words (3-5) displays as a set of little squares on the recipient's handset. And if a recipients are lucky enough to have an iPhone, they get the message in Chinese...

I submitted this problem to Orange customer support and after roughly two weeks got this reply:

W odpowiedzi na złoszenie techniczne dział ten informuje co należy zrobić, aby problem został rozwiązany. Przy problemach związanych z odbiorem wiadomości SMS prosimy, by odbiorca wiadomości wykonał w kolejności:
1. przełożenie karty SIM do innego telefonu na czas ok 3 godz.;
2. wykasowanie 2-3 wiadomości SMS;
3. wysłanie do siebie samego wiadomości: text.
Część telefonów / kart SIM niepoprawnie odbiera wiadomości typu LONG/graficzne/z polskimi znakami. Pozdrawiamy. Orange.

In English this means:

In reply to the tech support request, we advise the following steps to be taken to solve the problem. When a recipient has problems receiving SMS text properly, he should:
1. move his/her SIM card to a different handset for about 3 hours;
2. delete 2-3 SMS messages
3. send to himself/herself the message: text.
Some handests / SIM cards do not receive properly messages that are LONG/graphic/with Polish accents. Regards. Orange.

So indeed, when I want to send an SMS now, I should call the recipient 3 hours in advance, tell him to get another handset, move his SIM card to the other handset, delete a few messages, text himself. And then they will be able to receive my messages.

Surely I will be doing so. Be prepared :)

Mac Tablet Musings

Earlier this week I had a brief discussion with Houston Neal who posted an article on potential role of the iTablet in EMR applications. There are areas where I would fully agree with Houston. User interface (or its clean, streamlined design) has been the primary strength of Apple's OS for years. User interface has been the winning aspect of the iPhone. It has been both intuitive and cute. It is funny thing with the Apple Tablet (or iTablet or whatever it will be called once delivered...)... Apple has a long track record with pen / touch interfaces. Starting with early Newtons and now with the iPhone and iPod Touch. People love it. It seems pen / touch interface to the OS X should not be that far away. But may be Apple does not want to follow Microsoft's path of just adding a couple of tablet functions to the plain old operating system?

To be honest Microsoft has not been very successful with its Windows Tablets... They were stars of one or two seasons... back in 2004... And they have a niche, but regular users seem to prefer QWERTY keyboard and mouse. I am such a user. I had one of the best Windows Tablets - the Compaq TC-1000 (look: I always refer to it when I post in the Apple tablet....). The Compaq was fantastic for web browsing, but had some issues: poor WiFi, poor battery and slow processor. I believe HP would make good money re-releasing the TC-1000 now, with 8-hours battery, good WiFi/3G connectivity and Intel L9400 or better...). Then I had the Lenovo X61 convertible, but this was not a good pick... soon I swapped it for a plain X61, that was much smaller and much lighter and had significantly better battery life... I was not expecting that much overhead was imposed by the convertible / tablet design. But hey... the issues are almost identical with convertible laptop computer as they are with a convertible car... It is all about the stiffness of the frame... and making it really stiff means adding weight and bulk... By the way this was the strength of the TC-1000 design - the entire computer was in the "screen" part and keyboard was just an add-on. And this is what is envisioned on Houston's page: keyboard is just an add - on.

In the meantime (since first using the TC-1000 back in 2003 I think) I drifted towards classic permanent qwerty + screen design. All that was said about the cute factor of Apple touch interface can be repeated how pathetic it is with its real productivity. Whenever I see people gliding and sliding their fingers on iPhones or on touch pads of their iMacs, I feel sorry. There is nothing like a real keyboard... full stop. I am writing this blog now sitting in an armchair, with my QWERTY laptop where it belongs - on my lap :). It would be much more difficult do do that with just about any kind of tablet, as they are either too imbalanced with most of theit weight in the screen part or the screen joint is just not stiff enough causing the screen to wobble...

Also after spending a couple of weeks with the iPhone, I decided to switch to BlackBerry, and from perspective the 8900 Curve is by far the best phone I have ever had. Again the primary reason to switch was the lack of a physical keyboard on the iPhone. I write a lot and I just plain simple need real QWERTY. Even if it is a small one. But I agree - there are types of applications - mainly those who retrieve data, like EMR, that may be good candidates for tablet design. Let us see what future brings to the table... I just hope they won't replace the keyboard with a wheel...

Memory Leaks

Memory leaks are plaguing our gadget - heavy life. Most of the electronic devices we use, from game consoles to mobile phones to digital cameras to personal navigation devices to netbooks and beyond. All of them are computers in fact. A microprocessor, memory, storage and software running on top of the classic Von Neumann architecture. Processors are faster and faster and software becomes more and more complex, very often running away from initial frames it was designed to stay within. Software running away usually means weird behavior of an application or execution slowdown or loss of functionality. Software running away usually means memory allocation problems hidden deep inside. Software programs spend most of their time allocating and releasing blocks of memory. When there is continuously less memory freed than allocated, we have a condition called memory leak, ending up in a situation when all available physical memory is eaten by an application and no more can be allocated causing the program to stop or the entire computer to freeze or reset.

When it comes to memory leaks, a lot of them are linked to the general architecture a given platform (hardware and software) is based on. Architecture of software design and implementation and how much an operating system helps in detecting leaky applications and prevents one faulty application from affecting the other running in parallel. I would even risk a statement the memory handling abilities of a platform may be the main factor related to it's commercial success.

Some of you remember Windows 95 and how much of a bad image Microsoft owes to this operating system for its famous and notorious crashes. Much more stable Windows NT was not popular at that time and it took years for Windows XP to regain the faith of the users. Looking from a perspective, Windows XP did an excellent job convincing users a computer platform can run for months without resetting. But things seem to get worse now, new wave of applications seem to eat memory faster than anytime before.

There are two software environments that in my opinion are having significant problems handling memory. One is Symbian, known as one of the most difficult platforms to write programs for. The other one is Garmin, with its proprietary platform. I have Garmin on board in my car and it is not unusual that it just resets when navigating to a destination or just occasionally when I touch the display trying to invoke one of supplementary functions. At 3GSM World Congress I was surprised to see the new phone / navigator by Garmin on display, showing a "not enough memory to calculate the route" message. Really not something to be proud of in 2009. Of course the memory / OS problems did not prevent Garmin from reaching the top spot in the personal navigation devices market. But these problems may significantly suppress Garmin's aspirations to enter mobile phone market with the NiviPhone line.

While iPhone did it right (I never experienced a freeze or reset using iPhone), BlackBerry on the other hand is not without issues. So called battery-pull procedure is well known among the BlackBerry community, being offered as a remedy for many software glitches. Something goes wrong on a BlackBerry? Remove the battery, wait a few seconds and put it back... Forced restart of a platform. Good thing is it seems you are unlikely to lose data with that, but really... this is not something users should be doing. And applications like the QuickPull are just the proof of how bad the state of memory management really is in the industry....

I wish developers paid more attention to defensively design and implement their applications. I wish marketing managers gave development teams more time to stress test the software. And in the end... I wish users started picking their devices allocating more weight to device stability versus features....

The Switch: Episode 5

As promised. And as anticipated. It looks like this is the last episode of The Switch saga. I am now a happy customer of Orange, using the BlackBerry Curve 8900, including it's UMA feature. The entire story started back in December 2008, when I decided to jump onto the BlackBerry platform and started learning all the pros and cons and curiosities of this setup. As you may remember I selected the BlackBerry Bold 9000 smartphone and a two - year contract with Era (T-Mobile) network. In a matter of days I figured out two things. First, whatever application runs on BlackBerry, it heavily depends on what data connections are configured and running. And the connections and settings are a maze on the BlackBerry platform. I started downloading various applications just to see them failing to work. It never happened before on a Windows Mobile platform. Nor on the iPhone. WiFi was enough to run everything. And outside of WiFi coverage, the usual cellular data over standard APN was sufficient too. But not on a BlackBerry, that relied heavily on the [blackberry.net] APN connecting it to the BlackBerry servers. I soon learned BlackBerry without the [blackberry.net] APN connection was almost useless. And the solution was simple enough - subscribe to the BIS (BlackBerry Internet Service). I decided to shell out $15 a month to get BIS. Soon after it was turned on, many applications, like Google Talk and Facebook started working. And soon after I discovered, BIS did not solve all my problems. And soon after I found out, the real reason was the configuration of BlackBerry connectivity, intentionally screwed by T-Mobile. After long communications exchange with T-Mobile, I learned they were fully aware of the situation and absolutely unwilling to change the configuration. The decision for me was easy. Terminate the contract. But where to go?

I started checking the other MNOs around and discovered several interesting points:

• Both Orange and Plus (Vodafone) were offering the BIS service for 40% of the price @T-Mobile. Yeah I know I always pick the lonmgest queue :)
• Orange was offering UMA on the upcoming Curve 8900. Very interesting. In the end this was the deal maker.
• None of the MNOs could offer a BlackBerry for try-and-buy and none could confirm officially things like GTalk, Facebook, E*Trade and other applications would work.
• BlackBerry BIS was available only as a two - year contract with all the risk on the customer (me).

One thing I knew was T-Mobile setup was expensive and not working. But I could not be sure Vodafone or Orange configured the BIS service properly. So not knowing the pool was filled with water I decided to jump. I selected Orange, as UMA seemed to give me a better reward for taking similar risk. Thanks to the MNP (Mobile Number Portability), I was able to keep my number. Orange sent me the Curve 8900 a week before the number transfer procedure, furnished with a temporary number. This was a great idea, as I was able to configure the handset and all the associated services. The number was transferred last Wednesday night and that day I could confirm everything worked as [initially] expected. The end story is everything to recommend. The Curve 8900 is a marvel by itself. Smaller and lighter than the Bold, with even better screen and very similar keyboard. There also is a 3,5mm headphone jack, one of my key device requirements. The lack of 3G is nothing to regret at the moment. EDGE is not a speed demon, but works just fine to match the mobile screen and processor, while giving substantially more battery life with much smaller package overall. The Orange mobile plan gives me more minutes than comparable one at T-Mobile, for the same amount. And most importantly, the way they configure the BIS service is exactly as designed by RIM. Everything works, including the Internet Browser[BrowserConfig] and IPPP for BIS-B [IPPP] service books the whole war was about. With Orange I save enough money to cover the penalty I had to pay T-Mobile for prematurely terminating the two year contract. And I do not regret the switch a bit. It is a short sighted strategy trying to screw customers. In the end it does not pay off.

UMA Revisited

My new BlackBerry Curve 8900 arrived surprisingly last Tuesday. I had been told by the Orange sales representative before, it could have taken a week, but in this case the week was only two days. First impressions - the best phone ever, surely the Bold 9000 beater. Yes it does not have 3G, but who cares :) ???.

At the time I write this (Wednesday morning, as Sunday is planned for joyful snowboarding in Austrian Alps) the BlackBerry services are not yet fully functional, so I cannot comment if The Switch from T-Mobile to Orange has been successful, but one thing I tried was UMA.

I posted on UMA before, as I has been one of technologies / services I consider very smart. In short: it allows to talk via Wi-Fi. Elaborating a little more on the subject, UMA is a solution implemented by a MNO (Mobile Network Operator), that allows certain (UMA compliant) mobile phones to log on to the MNO's network over Wi-Fi. So instead of connecting to the MNO's base station, the phone uses Wi-Fi access point and fixed Internet broadband connection to create a secure tunnel to the MNO's network. It is very different at the physical level, compared to traditional GSM or 3G/UMTS, but is virtually indistinguishable at the logical level. UMA-capable phone logged via local Wi-Fi access point to the home network works as usual. It rings, receives calls, allows you to dial and connect the same way as it was using GSM/3G radio. Even data services work in the same fashion, mimicking packet data sessions to MNO's APNs, you normally use.

On lightly loaded local networks UMA does not require any setup. Just let the phone connect to the local access point (give it a password and that is it). On networks that are more utilized, you may experience some degradation in call quality. This is where QOS (Quality Of Service) policy may be needed. Imagine you are at home, and your home's Wi-Fi connects to a router than to a DSL modem and to your local ISP (Internet Service Provider). Your DSL is likely 1Mb/s or faster. As this probably is A-DSL (asymmetric), your uplink is slower than the downlink, so may be in the range of 512Kb/s or better. This is still plenty of bandwidth. Voice call plus all signaling around it should never exceed 100Kb/s, so the pipe is fat enough. But once you start downloading something from the Internet, or - better (or worse) - uploading some pretty heavy files, like your recent digital photo session from a camera up to the Flickr, you may have difficulties having a conversation over UMA. This is because large Internet packets flood the pipe and the tiny voice packets are stuck in this traffic jam. To resolve this problem you will need a router capable of prioritizing your voice traffic. This may not be an issue at home, as high traffic bursts are not that frequent, but consider one for business environment.

Apart from the above mentioned potential QOS issue, UMA is a fantastic solution. It frees the MNOs from taking an extra load on their network. It connects to the home network all those terminals usually wandering into the costly areas of national roaming partners. It gives coverage in all those places coverage is missing - as is the case in my home, where I could talk leaning out of my window, catching the stray GSM signals, but yesterday I could walk anywhere inside my house and have an uninterrupted conversation, thanks to UMA.

The problem with UMA has been the classic chicken - and - egg problem. Not many MNOs decided to implement this technology (T-Mobile USA pioneered it in the US while Orange seems to be the leader on the Old Continent). At the same time just a few dull handsets supported UMA. Last December, when I decided to go for BlackBerry, I was not aware BlackBerries supported UMA. And today this has been one of the key factors behind my decision to switch to Orange. UMA now gives me now true in-house mobility (something only a few people living on the edge of cellular coverage can understand). UMA calls are less expensive too. And when roaming abroad, if I succeed connecting my phone to the local hotel's Wi-Fi, I am not roaming any more... All calls even in Russia or Peru are local over UMA, so this is an advantage for frequent travelers watching their phone bills...

With more and more handsets supporting UMA (iPhone should have it, don't you think?), this technology can really fly. It should also be a huge cost saver for new challenger MNO's, who live on national roaming.