I was shocked when I read this article on Engadget. I actually had to read it twice after realizing that the words came out from the mouth of RIM’s co-CEO Jim Balsillie:

Now, instead of pleading for mercy at the feet of disgruntled consumers, RIM co-CEO, Jim Balsillie, calls the post-launch scramble part of the “new reality” of making complex cellphones in large volumes.

Let me get this straight: it’s not just RIM’s Blackberries who suffer from this problem. I still remember iPhone’s 2.0 firmware problems, and at my work at 7syntax, I fight every single day with obscure and undocumented bugs on the J2ME platforms.

Is it becoming a standard? I really hope not.. And before I start to rant, let me just say that I still dream that the current trend of opening the mobile platforms could help us a little.. (android: hint hint)

Ladies and gentlemen the set list is out! Although I'm not a big fan of, oh, Metallica's last 7 albums I can spot many personal favorites in there that I'm dying to try out. But I keep wondering: is Cliff Burton a hidden character that we'll be able to unlock and use in Expert mode? This would hit a very sweet spot with us, older fans.

Anyhow, my personal set list is:

• Battery
• Disposable heroes
• Fight fire with fire
• Master of puppets
• Orion
• Seek and destroy
• Whiplash

The songs from the other bands seem adequate enough, but no Danzig?! Come on...

I have to admit it: I can’t do any kind of development without my trustworthy pencil and a good amount of paper.

It’s not just to have something to scratch while I think how to solve the problem at hands. I find it very difficult to write some code without first drawing the idea and the pseudo-algorithm behind it.

To do that I use my own sub-set of the UML: boxes, polygons, arrows and labels. This allows me to quickly specify the exact behavior of the solution before writing any actual code.

Other method I often use after drawing the algorithm, is to define the set of tasks I need to finish the implementation. It is often required to refactor or change some other code or structure, so I write a set of labels with an empty checkbox at the front. As I write the code, I start checking the boxes, and that gives me a warm feeling that I’m getting the job done.

So, which methods do you use to be more efficient while writing code ?

Torch Mobile Iris Browser is based on WebKit rendering engine for Window Mobile devices, like my old i-Mate Kjam, a HTC Wizard clone. I've doing some tests and it's a damn small, mean and interesting implementation of webkit. It renders all those web 2.0 goodies nicely, like ajax, javascript et al. I couldn't help getting some memory alerts - one problem of having only 10mb available - but overall it's a nice browser, replacing all other existing options; Internet Explorer and Opera. handivi obviously runs on Iris quite fluid and with all javascript support. Here's some pictures from the "blurry camera man" :)
   

Last week Apple announced their AppStore had reached the number of 500 million application downloads. This is a really impressive accomplishment, even more because Apple launched this service at a time when most people were claiming that web applications were the only way to go on mobile, something I can't simply agree with because it depends a lot on what you are developing.

Following Apple's footsteps, Google launched the Android App Market - which appears to be evolving a bit slower than expected - and there have been lots of rumors about others creating similar services for Windows Mobile, etc.

What I'd like to see though is providers enabling m-payments through their AppStores. Like for example being able to pay for an Evernote subscription through my Apple iTunes account without needing to provide any credit card information to Evernote Corp.

The flow I see is my downloading Evernote and the application asking me if I'd like to subscribe to a premium account. By acknowledging it, an iPhoneOS billing wizard should popup, allowing me to authorize the service to charge my iTunes account for a yearly subscription of the service. This flow should also apply to Android or any other mobile devices that provide access to such services.

With this, 3rd party developers should be able to avoid splitting revenue with mobile operators who, on average, keep 50% of the charged value: say you charge $1 USD for a virtual gift on your customer's phone bill. In this case the operator will keep $0.5 USD and you'll get the remaining 50c.

Knowing that Apple, Google, etc. all have activation processes in place that require use of a credit card I think it should be relatively easy to experiment with this concept, what do you think?

Since I started to use Twitter I feel some insatisfaction with the available desktop clients. The balance of pros an cons of each desktop client always tended towards the negative side, which made me continue to use the web interface... until yesterday. Yesterday I met Gwibber.

Gwibber is powered by python and webkit, so far supports brightkite, digg, facebook, flickr, identica, jaiku, laconica, pingfm, pownce, rss and twitter.

Take a look at the default theme:

I like to hack and customize things to work the way I like most. The simplicity of Gwibber's internals are like 'hack me, please'. Gwibber delegates all the boring tasks to webkit+html+css+js, so there is no complex and boring python code to render messages and support themes. If an user wants to create a new theme, he just needs to edit html+css, couldnt be easier. Creating a new plugin to support a new microblogging site/service is also trivial.

Try it yourself: http://launchpad.net/gwibber

Hi @all! My name is Rúben Fonseca and I’m the handivi mobile lead developer. Today I’m talking about some improvements on the handivi client for Blackberry.

We certainly feel your pain mind_booster. The more I dig into Blackberry, the more I realize that it’s a solid device with a stable OS, with a great potential for developers. Since OS version 4, Blackberries support a subset of the MIDP 2.0. Soon we realized that our J2ME application could be easily adapted to the Blackberry.

Unfortunately that was easier said than done. The first try to hack a Blackberry
version resulted in a huge fail - the gap between the Blackberry MIDP 2.0 subset and the official profile was too deep.

However, we felt upbeat to try different approaches. I spent more time studying the Blackberry Java platform and found that the missing parts of MIDP 2.0 are implemented using a custom Blackberry API. Last Tuesday was the Blackberry day and I think we made a huge step forward.

Handivi now runs fine on the Blackberry emulators, and we hope to try it on real devices soon. Stay tuned for further mobile developments!

We're now affiliated with the European Startups initiative, which is a portal and a tool targeting entrepreneurs interested and/or engaged into European Web start up scene, from Ivan Trajkovic.

For now we're the only one on our country category index, but for sure that more portuguese startups will be there also - at least we expect so.

Check out our page there and the blog aggregator from all startups.