Update Report 001 – Ohra, Italy, Touch Screens, Android and Design Patterns

Posted on April 11, 2012


Since I probably will not be able to hold up to a weekly report as Alper Cugun and Kars Alferink on “Whats the Hubbub” I decided to simply create update reports and number them.

Period: October 2011 to March 2012

this update covers the period from October 2011 to March 2012.

Working on- and closing the Ohra Project

In October 2011 I was hired to do some Flex work for Ohra: working on the dialogs that help people request an insurance online.

I finished 6 months later, end of March 2012 and left behind a new framework to build the screens and dialogs faster and more efficient.

The project started ith bug-fixing and the test and release phase of two earlier projects: “Woon” and “Tijdelijke Reis”. These two products were based on the old architecture used to build about 10 other dialogs that were delivered in the past two years. While bug-fixing I found several points for improvement and started logging them down in what I call a: “shit list”. The shit list contained all issues that caused problems either due to overcomplication of processes and code or due to technical debt: code that should have been refactored at some time, but was not due to other priorities.

Based on this shit list I started designing a new framework in November and December: aiming at a way of building that would allow for faster and more solid results. One key element was to encapsulate all standard elements of input-items into View Components. Another key element was to completely separate the Model from the View: binding it via Injection and tags passed to the View Components.

This way you could easily build click-through models and then add extra constraints and refinements as you went.

About that same period, another team started the implementation of two HTML versions of earlier dialogs, to cater iPad and Android users. Not even a month later, in January, HTML started to become the main buzz.

As stated, we finished the “Woon” and “Tijdelijke Reis” products in November and started building on the “Overlijdensrisico” product that was next in January. As the new framework was still in progress as we were building, the work was divided into me building the framework and my college building the screens and dialogs. Due to holidays (I already planned and booked a 1 month holiday for December) and illness “Overlijdensrisico” got delayed. And due to political reasons the product was shelved until further notice.

In February we started building “Arbeidsongeschiktheidsverzekering voor Zelfstandigen”. As most of the Framework was already built while doing “Overlijdensrisico” we were able to make a lot of progress and build most of the screens in the first two weeks. As each product has very specific business rules that change the dialogs as you select stuff, we spend another two weeks getting the interactions and business rules right. In the last two weeks we wrapped up most of the tiny details, leaving a very nice product (and maybe one of the best products) for testing.

March was my last month and after wrapping up the last bugs, buying goodbye cakes for my colleagues I closed the project and prepared to deliver the studio unit I had rented in Arnhem.

At that time, Flex was definitely dropped as a technology to build new dialogs with: being replaced by HTML using JQuery and some other technologies.

Sonic Acoustic Wave Screens, ARM-based boards

During the same time I started building some concepts based on older ideas for Roomware and HotForestGreen, dealing with Touch Screens, sensors and Android.

I ordered and received two Sonic Acoustic Wave Touch Screens in November and ordered and received two ARM Cortex A9 based experimentation boards from PandaBoard and Calao Systems. Both capable of running Ubuntu and Ice Cream Snadwich on full HD format.

I expected the Sonic Acoustic Wave screens to be close to indestructible by human error, as they use sound waves to determine where your fingers hit the surface. One clear and serious problem presented itself soon enough: when fluids hit the table (basically my spit) those drops break the acoustic wave and are perceived to be touch points. Another issue is that the Sonic waves require specific ridges carved in the sides of the glass, to create specific breaking-patterns that can be registered by the sensors. Any dirt breaks the process, making the Sonic Acoustic Wave panels even less fit for rugged use than Resistive or Capacitive touch.

22 inch Android tablet

While I was in Italy in October I also created as quick Proof of Concept on Android Resistive Touch on a 22 Inch screen. The hack was simple and straight forward: the Resistive Touch tablets I already bought for experimentation purposes has 4 wires. The 22 inch Resistive Touch screen I had installed on a 22 inch Samsung screen also had 4 wires. With a bit of luck both would be compatible and the tablet itself would allow me to use that screen.

The experiment was successful and was posted to my blog, to Youtube and to Hackaday, where it gathered over 2000 views in 3 days time.

Arduino, transistors and soldering work

In that same time around November I picked up electronics again, all in the same line of experiments: building the simple infrastructure that would allow me to monitor a house and see what is going on inside. I built several simple schematics, including a Light Bridge and a very sensitive amplifier that would react to the electric field surrounding the human body: handy for sensing proximity.

Touch based on Proximity

Using the results on the proximioty sensor in November I started experimenting in the late hours with the possibilities of building a Touch screen that would be less sensitive to damage: allowing me to use a glass pane to cover the screen and still sense touch based on proximity of some part of the human body (the fingers).

Using some tricks I was able to build a simple prototype that would allow multitouch without the disadvantages of the Sonic Acoustic Wave panels.

Due to time constraints I was not able to finish the first prototype until now.

Android ADK book

During that same period in November where the Android Touch screen hit the web, I was approached by an agent to write an Android ADK book including and involving Arduino and the Arduino ADK board.

Design Patterns book

Postponed due to all other projects I decided to take the one month holiday of December and January to start writing the Design Patterns book I had planned somewhere in May.

In 20 days time I wrote over 60.000 words and made over 100 illustrations on Design Patterns and the other theory I cover. In the successive weeks, all other projects were abandoned as my February deadline for the Design Patterns book was approaching. Another reason for that abandonment was that discovering how these Design Patterns worked was great fun. Each section I wrote on Design Patterns was like a Nerd-rollercoster into even more knowledge.

In the successive weeks until end of February I wrote another 40.000 words in the evening times, bringing the total to 100.000 words, 160 illustrations and around 500 pages in Word.

It is now the beginning of April and I am happily typing away at what should be the first publishable version online, properly edited, containing all the major patterns and without the invalid assumptions I made about – for instance – the Strategy Pattern.

Word to Website

In that same period (Februari) I revisted “Word to Website”, which I created somewhere in 2004 and allows you to publish a Word-document to HTML and to a structured website. Using C# I hammered the basis together in one weekend: doing all the basic stuff using the Builder, Visitor and Parser Patterns to cleanup and map the chapters and the content of the Word document.

Italy and my own projects

Having been in the Netherlands most of the time, and away form my love, closing March it became more and more clear that it was time for a break from all work and be more with my girlfriend. I cancelled two potentially and pending gigs, booked the flight and went.

The idea is to spend the next period until the end of 2012 working and investing in my own projects.

This includes the hardware-projects, the touch screens, the HotForestGreen framework and the Design Pattern Book.

Posted in: Status Report