Rants And Ramblings About Mobile Technology

Once I had the vision that mobile phones would have a desktop, kind of like on a PC, with local applications as well as links to web sites sharing the same surface, and by extensively using JavaScript and HTTP communication with services. This way the web sites would not look and behave that different from local applications. At Obigo we had discussions about implementing what later became AJAX in our web browser, but then also supporting e.g. SMS and MMS communication from the JavaScript applications. I figure many had the same thoughts. I know Opera and Access did at the time.

Anyway, if the iPhone is an indicator of future phone functionality (and it is in many ways, in advance) it seems web browsing will take the second fiddle, as even companies that don’t have anything sensible to present via applications (see e.g. Coca Cola’s “magic bottle” application as an example) still prefer to make applications instead of mobile web sites.

There are some good reasons why:

• The interactivity is much much better, to the point that mobile web sites are outright laughable in comparison.
• Applications are fast (at least on the iPhone) and can sport animations, 3D graphics, sound, GPS, accelerometer, compass, camera etc etc, functionality more or less impossible to add to mobile web sites.

I’ve said the same things about MIDlets, but it’s even more apparent on the iPhone.

Let’s see if HTML 5 can shift the balance, but I doubt it. When will phones support HTML 5?

I only wish this had happened already with MIDlets, but phone manufacturers didn’t realize making phones optimized for applications is a paradigm shift and requires a completely different mindset in the design of the phone.

Posted by Anders Borg 23:10 | Permalink | 0 comments | Links to this post

All iPhone-based MIDI controllers I’ve found seem to rely on Wi-Fi communication and show up as MIDI devices on a Mac. I wish there was a solution for direct communication with synths as well, using USB instead. Of course, a wireless solution is much more practical on a gig.

There’s probably more out there:

• New iPhone MIDI Controller
• Pro XY Turns Your iPhone Into A Wireless MIDI Controller

• iXY MIDI Controller For The iPhone

Posted by Anders Borg 11:45 | Permalink | 0 comments | Links to this post

Short term for checking out the possibilities to develop applications for it. I lack a Mac though, but I’m considering a Mac Mini, so I can use existing screen etc real estate, and so it doesn’t take up space.

Regarding Android: I haven’t released my Twitter application yet, and might never do, considering there’s already better applications in Android Markets, and they are completely free. That doesn’t mean there aren’t opportunities for Android applications (as previously indicated), yet most likely in other application areas.

Posted by Anders Borg 10:42 | Permalink | 0 comments | Links to this post

(I know, off-topic again…)

Finally scientists come to their senses and skip the notion that robots need to look like human beings, and the obvious drawbacks of that (slow, clumsy, large, etc). Now they rather look at the task at hand and how robots need to be shaped for that. Two examples:

• Secrets Of Insect Flight Revealed: Modeling The Aerodynamic Secrets Of One Of Nature's Most Efficient Flyers
• Mimicking Insects to Avoid Sinking Using Surface Tension

What people seem to forget or ignore is that e.g. insects are extremely optimized for their respective biotopes, evolved during millions of years, hence anything that would make the insect power inefficient or badly adapted have been ironed out long time ago. Spider-like robots would be very practical for complex surfaces, etc.

This is one of those “I told you so, 20 years ago”, but as blogging didn’t exist then, I have no proof of the matter.

Posted by Anders Borg 09:12 | Permalink | 0 comments | Links to this post

Another one of those off-topic stories. I have plenty.

I discovered a toolkit for accessing MIDI devices from C# applications, so I hooked up my trusty 20 years old Korg M1 to my laptop via a USB-to-MIDI converter, and have had fun ever since.

As you most likely know, MIDI stands for Musical Instrument Digital Interface, and has been around for more than 20 years, and is physically implemented via the original MIDI connectors (DIN5, the same as was used for analog audio at the time) or via USB. Communication is all digital and serial, and what’s transferred is the events that occur (note on/off, pitch bend, etc) rather than the audio.

It’s extremely easy to connect two MIDI-compatible devices together, which was really the core goal: No need to know anything technical to hook stuff up.

As you might also know, MIDI was initially a rather slow protocol, clocked at 31.25 kbps. As each note requires 3 bytes (note on/off, channel, note and velocity in combination), the time resolution is approx 1 ms. You might think this is plenty, but consider a 6 to 8 note staccato chord. We are almost down to a resolution of 100 ms. Sure, there’s a possibility to compress Note on/off to only 2 bytes each, but this is still not really enough. This becomes even worse if you play a note-dense multi-track song (e.g. the M1 (and remember, it's really old) handles 8 tracks using different instruments) via MIDI. When MIDI is transferred over USB end-to-end the bit rate and hence resolution is of course much much higher. The data transferred is though the same as via DIN5.

More professional synthesizers and keyboards nowadays tend to come with USB/MIDI, but surprisingly many lower-priced devices still use only the original MIDI connectors for compatibility reasons.

MIDI supports numerous other commands than note on/off of course, like pitch bend, program change, etc, yet where synth manufacturers tend to go crazy is in the use of System Exclusive, that enables proprietary exchange of sound banks and other data, so you can re-program your synth via MIDI.

Anywho, the short term goal is to re-write my MIDI echo I “invented” a long time ago. It essentially plays back notes at set delays, with the added twist that the sound played back might be completely different from the original one. As MIDI is all digital I simply offset times, notes, velocities (essentially volume) and channels any way I want. By switching to a new channel an initial piano sequence could be played back as a soft pad at a lower volume and in a lower octave a few seconds later etc. It was an extremely effective creative tool, so I've missed it since MS-DOS came out of fashion.

I have my eyes on a Korg M50, that is way better than the M1 (it should be, considering the time difference), except for the not as good keybed, lack of aftertouch and less sturdy casing.

And I need to learn to play too.

Posted by Anders Borg 22:13 | Permalink | 0 comments | Links to this post

Information found via the RCR and Fierce newsletters:

Motorola will be using Android in new phone models, initially the Cliq, a rather normal-looking business-style phone (a la Nokia N97, Sony Ericsson (HTC) X1 etc), but with a broader appeal due to a focus on social networking via the MotoBlur software.
Motorola anchors its rebound on Android and all things social

Multiple things are to blame for Android Markets relatively sluggish development: fewer phones, no top display of commercial applications, trickier payment system. I would also blame less interest from developers (Catch 22 due to fewer phones etc) and more differing specs between phones requires testing on all those phones (a whiff of the terrible MIDlet situation). The iPhone / iPod Touch models are on the other hand essentially identical.
New phones, seller tools improve Android Market's fortunes
Can developers survive in Android Market?

There are benefits with developing for Android though:

• You are early on, and can build a brand and presence due to less competition, which is way harder for the iPhone.
• Getting a top spot at App Store is incredibly hard and unlikely.
• If you can make MIDlets or desktop Java applications, and you know Eclipse, you can quickly learn to make Android applications. The UI paradigm (activities etc) is the biggest hurdle, yet the iPhone is certainly not simpler to develop for.
• You don’t need a Mac to develop for Android. Any PC (running Windows or Linux) will do. Might sound like an irrelevant comment, but most companies don’t have nor want to invest in Macs.
• There will soon be a whole slew of Android phones from Samsung, Motorola, LG, Sony Ericsson and HTC (but not Nokia), so it’s the right time to learn.

Posted by Anders Borg 20:24 | Permalink | 0 comments | Links to this post

I don’t understand why anyone would perform a DoS attack on abiro.com, and maybe I should even consider it an honor to be so recognized that I at all get a DoS attack. Anyway, the site is up again now.

Posted by Anders Borg 12:39 | Permalink | 0 comments | Links to this post

I’ve just started advertising for “Mobilt liv” via Google Adwords. Of course only in Sweden.

I so far have a CTR of 0.06% (which in ad CTR pro lingo is “disgustingly dismal”), yet with a reasonable amount of impressions (tens of thousands in a day) averaging as the third ad on search pages. The cost was therefore minimal, but maybe I still got some visibility due to the amount if impressions.

I don’t expect an immediate return, nor that this will be a very effective method even over time, but as long as the cost is this low I can keep the ads rolling for weeks.

Posted by Anders Borg 20:52 | Permalink | 0 comments | Links to this post

Supposedly the cheapest available, but then why two displays? Remove the other one, as it’s much easier to read on one display only, and the price will go even lower.

ASUS planning dual screen Eee Reader: world's cheapest e-book reader

The only reason paper books have opposing pages is to cut cost and lower weight, not for readability reasons, but outdated solutions seem to stick. I hope Asus understands to make a one display model as well.

Posted by Anders Borg 20:37 | Permalink | 0 comments | Links to this post

No further comment required, except rejoice, until you are out of flatrate:

Spotify launches on iTunes App Store and Android Market, premium members only

Posted by Anders Borg 20:33 | Permalink | 0 comments | Links to this post

Finally the iPhone is useful (just kidding) :)!

Apple approves officially-licensed Commodore 64 emulator for iPhone

Interestingly, to get the right feel in especially games, the emulator must be equally slow as the C64, meaning emulating a 1 MHz 6510 (pretty much the same as the very popular 6502, that has a strange and impractical instruction set with hardly any registers), and it needs to emulate the (at the time) rather advanced SID audio chip.

Irrelevant nostalgia: I for instance used the C64 SID to emulate com radio tone squelch, both for serious uses (testing out com radios in a TV/radio shop) and for fun (people running to answer the com radio, but there’s no one on the other side; rinse and repeat; you get the picture).

Posted by Anders Borg 20:30 | Permalink | 0 comments | Links to this post

I’ve submitted all my reasonably useful/well-mannered MIDlets to PlayNow Arena. I certainly don’t expect all of them to be approved, even though I hope Converter Pro, EasyCall, ColorSampler and Watchdog will, as they are the newest and most interesting in my opinion. Who chats via SMS, and does the world need yet another Twitter client? Admittedly Jitter is my most downloaded MIDlet at Kalador, but that version is free/ad-supported.

All MIDlets I uploaded to PlayNow have a price tag, spanning between 2.50 and 4 Euros. No freebies this time.

All are strictly named “Abiro …” to build brand.

Posted by Anders Borg 19:16 | Permalink | 0 comments | Links to this post

In the web applications arena it’s not given that big players succeed, but rather the field consists of a massive amount of independent and entrepreneurial small companies and even hobbyists, that might not even have a commercial intent initially. One reason this works is that it’s so darn simple to develop and host web applications for a worldwide audience. Most fail of course (how many communities can a world bother with), but due to the Darwinistic and chaotic nature of this marketplace a few will succeed, that satisfiy the needs of many users.

Examples:

Twitter, still a very simple service that in my opinion hasn’t evolved a bit in two years. Supposedly now run by 20 people, supposedly has no profit, has funding enough to run in the black, yet knocks the socks off heavily-funded Jaiku (acquired by Google) and similar services in popularity and mindset. At the same time being a service type so simple that before its inception no one would believe it would have any applicability at all, and many still don’t believe so.

Piratebay, tracking torrents via an amateurish site that looks like crap, yet caters to the needs of a whole world: free multimedia content, whether legally obtained or not. Some funding, but considering the lawsuits against them, no possibility whatsoever to fend off those lawsuits financially, except due to the fact that Piratebay doesn’t actually store any multimedia content, so conceptually they are in the green.

Skype: Certainly Skype has always had a commercial intent, but it’s still been resource-efficient along the way, where the main value has been the free phone client: Free to get and free to use. Sure, there are commercial services complementing Skype, but that’s not the core reason people use Skype. As Skype is acquired by eBay (but might be sold off, according to info from eBay) the company never had to be in the black to still hugely succeed from an owner perspective.

Facebook, initially a internal service for Harvard University’s students, proved to be so generally useful that it was further developed and opened up for general use. Nowadays it’s a quite advanced service, probably the most advanced community in the world from a technical standpoint, due to the third-party application architecture. Still remaining the most popular community for working people, decreasingly so for pre-teens and teens (children don’t want to be on the same community as their parents). Certainly by now a fully commercial service.

What else will pop up from nowhere, that will be spawned by a few guys, possibly not even with any commercial intent, and yet will become a huge (as in world domination huge, at least for a while) success?

I’m sure there are big new opportunities in other areas, where timing is everything, but finances and manpower are initially almost nothing.

Posted by Anders Borg 11:45 | Permalink | 0 comments | Links to this post

I’ve long expected Windows Mobile to lose out, and also Symbian will. Both OSs/platforms are outdated, and there’s need for new technology to move ahead.

Smartphone segment keeps growing, but not for Windows Mobile or Symbian

Here’s the market share of each “smartphone” OS/platform:

iPhone Share of Smartphone Sales Continues to Increase

Please note that most Nokia S60/Symbian phones are purchased as normal voice-centric phones and that “smartphones” is still a smaller part of the total market.

Posted by Anders Borg 13:44 | Permalink | 0 comments | Links to this post

I’m experimenting with a new service called “Mobilt liv” (Mobile Life), initially aimed for the Swedish market due to the issues with allocating short numbers in many countries and the costs involved, providing (hopefully) useful and fun SMS services, starting off with some simpler ones, including getting a map of where you are, calculating BMI, sending e-mail via SMS, getting Magic 8 Ball responses etc.

The only way I know to get revenue from SMS services is via subscriptions, but that’s a moral grey area that I haven’t found enough stomach to get into yet.

The name Mobilt liv / Mobile Life obviously has a double meaning, as in utilizing mobile technology as well as living a mobile life, which the services are also intended to convey.

I completely re-designed the “shopping window” page today, and added auto-detection of PC and mobile phone browsers, so the same short URL (mobiltliv.se) can be used for both. Doing it this way means there’s no need for any xxx.mobi, wap.xxx, mobile.xxx or other special domain for the mobile site. I also added “smsto:” links to the mobile page, so that it’s easy to spawn messages without entering them manually.

The service is completely housed in a single script file, including reception of the SMS, determining what service to perform, performing that service and sending back the response SMS. No point in making it complicated.

I set the price to 3 kr (40 cents), which is very low for SMS services, and might give the wrong signals, but experience will tell if I made a mistake.

The photo in the top banner is from my home district, captured a few days ago, and maybe you notice the Swedish flag colors in the logo. All for the national vibe.

Now comes the question about building awareness. It won’t be easy, unless I can piggyback on some other popular service.

For the daring, here’s a Google translation to English. Most of it makes sense. Oddly it translates “Wikipedia” to “articles”.

I invite you to try the service (if you are in Sweden that is; if not you won’t be successful). Try e.g. texting liv här to 72500.

Posted by Anders Borg 20:37 | Permalink | 0 comments | Links to this post