Technology and giving back

I am a strong believer in the concept that technological innovation can and should be a strong driving force for the good in the world. The Internet, while full of porn and cat pictures, has given us a great way to share knowledge, experience and rally people to a common cause. Still, doing charity work in the classic sense requires spare time, money, or other resources. But now, thanks to technology, you can help the world with nothing but a smartphone (though, I mean, charity is one of those things where more is better).

The Vodafone Foundation, in cooperation with the Garvan Institute of Medical Research, released an Android application called DreamLab. The idea is simple - as you probably know, every smartphone is essentially a tiny computer which is powerful enough to rival PCs from a couple years ago. When the phone is not being used, all that power is wasted. DreamLab puts that power to use for good to fast track cancer research.

I have been using the application for a while, and I personally think it is great. It turns on only when the phone is plugged in and fully charged and you never even notice it otherwise. It does require a network connection, but it is easy to set up data limits if you don't have unlimited Internet access.

According to the app, Garvan is already working 686 times faster than it could have without it. But there's no reason to stop at this point, right?

You can download the application here.

If no one makes a program I need, I'll just make it myself

Oh yes, the option to write your own software is one of my favorite perks of being a programmer.
For example, I noticed that the following patter has become way to common in my life:

• I stumble upon promising movie (book, game, article, whatever) on the Internet, which I am too busy to check it out immediately.
• Wait a couple of days and forget all about it.
• I have some free time and have no idea I want to do.

For some reason, I was not able to find any application to help me with this problem. Apparently, the common solution is to just make a list of Evernote or something, which sounds like too much work to me. So, I made my own application which handles this problem for me. You guys are free to use it too, if you want. It is open source, too. 

I will probably be improving it a bit in the following weekends, but I'm also open to feature requests. Enjoy!

My latest project

Today feels like a special day. The reason is that the project I have been working on for a couple of months now has been released for alpha testing. I encourage everyone to check it out at http://visitmeadow.com

The whole event planning thing was sort of a pet peeve of mine for a while. It was the worst when I had to plan a vacation or something of that sort, but even making arrangements for a barbecue sometimes felt like a mess. I mean, we would lose important information and agreements, forget something important. We would also have to use a ton of different applications. I mean, at least there's a facebook group to chat. A facebook event, if we want a convenient place to store the time and location. Then evernote if we need a check-list or something. Dropbox if we need to store a file or several. Some table to keep track of expenses. The list goes on.

So, naturally, I was very excited to make an application that should help deal with this type of problem, and I sure hope we are not the only people who keep having it.

The project is also quite a pleasure to work on from a technical perspective. Getting to choose any tech stack was like a breath of fresh air after a couple of projects with legacy requirements. The only decision I'm not quite sure about is AngularJS. I mean, I really like the framework, but it seems like it is in some sort of transition to version 2.0 which will break backward compatibility.

Anyway, the project is far from over and many important features are yet to come. Will try to keep you updated on any interesting developments!

The Pragmatic Programmer

I read a lot. It's a habit I've developed early, and I have no intention to abandon it any time soon. I read books, I read the news, I read blogs, I read various articles and tutorials. Basically, anything that manages to catch my eye.

When it comes to programming, most of my reading comes from the Internet. The main reason (apart from convenience) is that in such a dynamic area I like to get my knowledge from the most up-to-date sources possible. Books on programming, unfortunately, tend to become outdated quickly. Not all books, however. There are a few that will likely remain relevant for many years to come. I've recently finished reading one of such books - "The Pragmatic Programmer: From Journeyman to Master" by Andrew Hunt and Dave Thomas.

I found it to be an amazingly useful book. Even though it was published waaay back in 1999, it is still very relevant because it does not focus on any specific technology. Instead, it examines the core process - understanding what your clients want, and turning that requirement into working, maintainable code. As stated in the books description, it covers topics ranging from personal responsibility and career development to architectural techniques for keeping your code flexible and easy to adapt and reuse.

Of course, a lot of the stuff described is pretty obvious. However, I would not necessarily call that a flaw. It's still invaluable information for beginners, and for the more experienced programmers, the book helps organize the methods and concepts into a neat structure.

To sum up, I would say this book should be a must-read for anyone who wants to be a good programmer. It won't magically turn your bad code into awesome code, but it might as well give you a push in the right direction.

My projects

It occurred to me today that I have been really busy lately. Whenever I'm not working on freelance projects or doing consultancy work (apparently, it's not so easy to escape my previous career), I am working on one of my own projects. Which seems exciting to me. So I decided to make a brief summary of the stuff I finished already and the stuff I am currently working on.

I publish my finished software under the Bearson Software trademark (link). I mostly work on it alone, with the occasional help of a designer and a marketing specialist. So far, this is what I have:

Certification manager 2015 (Github). This is a niche tool for MRO (maintenance, service and overhaul) service companies in the aviation industry. This was my first project, and it lead to an order from one of such companies for a improved and customized version.

The main purpose of the application is to a) keep track of employees and their personal information; b) keep track of training and certification of the employees. Other functionality includes import and export of data (xls/xlsx) and report generation.

Pure list (Github, Google play). This is an Android to-do list and task manager. It provides basic functionality for this type of app (multiple lists, item reordering, marking items done, reminders) plus a home screen widget. The main selling point is the clear and intuitive design, which is customizable with several themes.

As for the current projects, I have several ideas that I am testing out before making a fully functional version. I will try to update this post as soon as it takes some defined shape.

On choosing a price for software

Good evening, fellow readers!

Yesterday, I finished a large commercial application and finally launched it. I spent about a month developing it, and then a couple of weeks setting up a website, testing payment methods and so on. As of yesterday morning, there was only one decision I still hadn't made - how much should I charge for a licence? I didn't want to make it so expensive that no one would buy it, but I obviously didn't want to underprice it either. Now, I may be wrong, but perhaps the decision would be easier if the application was targeted at a consumer market (it is not).

As an economics major, I knew the basics. However, that does not help much because, a) software is different from other products due to the fact that it does not have any variable costs (i.e. creating one instance of the application costs me the same as creating ten million); and b) economic theory works best when your customers are perfectly rational decision-making robots (which, I assume, my customers are not). So, naturally, I had to do some research.

Unfortunately, no source told me how much my software should cost. However, there was some advice, that I found useful. Here it is:

1) Perceived value is not the same as objective value
From a logical point of view, if a product saves you 5 hours of your time, and you value your time at $20 per hour, you should be willing to pay anything under $100 for that product, assuming you have nothing better to spend that money on. However, how many of you actually make such calculations? For most people, purchasing decisions are made based on the perceived value of a product, which can be influenced presentation, marketing pitches and testimonials - and, surprisingly by the price.

2) The price sends a message to the customers
A product for $10 is perceived differently than a similar product for $100. Underpricing software is dangerous not only because it can decrease revenue, but also because it can trick customers into thinking that the application is useless or low-quality.

3) The product is more than the product 
People pay for the experience of dealing with a business, not just for the software itself. So, the price should depend on the objective reality of the whole operation - starting from the marketing pitch and finishing with the support that is provided.

4) Price tiers help - but at a cost
By offering several price packs, not only do you give the client a choice, but you also provide a point of reference (also, importantly, one that does not involve looking at the prices of competitors). However, multiple prices mean more complexity for the marketing, a more complicated support infrastructure and, potentially, more confusion for the clients.

All in all, choosing the price is still no easy endeavor. Fortunately, as with most things, you can test different options and see what works best. I have decided to go with A/B testing, where several groups are made the same offer at a different price to see which responds better.  

I love Material Design

Good evening, fellow readers!

One thing you should know about me, is that I really enjoy working with pretty GUIs. Don't get me wrong, if the need arises, I can work on a rotten potato with a command line, but a properly designed user interface just makes me unreasonably happy.

Unfortunately, I did not like the recent developments in mobile design. Sorry, Apple, sorry, Microsoft, but your minimalistic flat interfaces just do not appeal to me. Google seemed to follow the same pattern for a while, but things changed in June when Material design was announced. The concept instantly caught my attention, but, as usual, I had to wait before I could use it with my own hand before I passed judgement.

And after several weeks of use, I must say that my expectations were met. Material-design apps seem to pop to life, and are all-around a pleasure to work with (developing them is a bit less pleasant, but more on that some other time).

The style seems to be catching on, too. Apart from the Google apps, I have already seen a file manager, a calculator and a task manager with Material, and there is a Linux distribution on it's way (check out Quantum OS, it looks fun).

Anyway, good job, Google designers! Keep up the good work!

Working with tables in Java

Good evening, fellow readers!

Today, I was working with some tables for my application, which reminded me how confusing they were when I started learning the language. Guides help, and so does the documentation at Oracle, but I would like to point out and explain some important aspects that were not clear for me personally.

My specific area of focus will be the implementation of tables in Swing. The first important thing to understand is that tables in Java were designed using the MVC (model-view-controller) pattern. The JTable class is responsible for the "view" part, and the TableModel class is responsible for the "model". "Control" is handled by various Listeners, but I won't be touching that now.

The only job of the JTable class is to display the table on the screen. The data to display, along with details of how it is obtained, sorted, and displayed, are handled by a TableModel associated with the JTable. It is possible to create simple tables without having to worry about a TableModel (or knowing that such a thing exists). When you create a new JTable, unless you specify a TableModel, an associated DefaultTableModel is automatically created.

If you have to display anything more than the most basic set of data, it is best to create your own model. This can be done by extending one of the two classes that implement the TableModel interface - AbstractTableModel or DefaultTableModel. The AbstractTableModel implements the majority of the methods defined in the TableModel interface and leaves three methods (getRowCount, getColumnCount, getValueAt) for the user to implement. It's the basis for any custom table model a programmer might create. The DefaultTableModel class is such an implementation. It is a subclass of the AbstractTableModel with the three methods implemented where the data is stored in a Vector of Vectors.

The decision between the two mostly comes down to the data. If you are OK with storing your data within the outdated Vector cl, using the DefaultTableModel is a lot less work. If you decide to build from the AbstractTableModel, you can choose to store the data however you want to but you'll need to implement the methods that allow the data to interact with the JTable seamlessly.

At the very least, you will have to implement the getRowCount, getColumnCount and getValueAt methods. As you may have guessed from the names, the getRowCount method should describe how to find out the number of rows in the table, the getColumnCount does the same for the number of columns, and the getValueAt is responsible for describing how the data is distributed among the cells. Here is a simple example of a custom AbstractTableModel:

import java.util.ArrayList;
import java.util.List;
import java.util.Locale;

public class EmployeeTable extends CustomTable {

private List dataList = new ArrayList();

        //declare the names of the column headers
private String columnNames[] = {"Name", "Phone"}; 



//in the constructor, retrieve an ArrayList of the data to be displayed
public EmployeeTable (){
super();
DatabaseReader readData = new DatabaseReader();
dataList = readData.read();
}

// the number of columns will be equal to the number of column headers we declared
@Override
public int getColumnCount(){
return columnNames.length;
}
  
        /*
         *to count the number of rows, we count the number of entries in the ArrayList.
         *The will be a row for each entry.
         */
@Override
public int getRowCount() {
return dataList.size();
}

        /*
         *This part is a bit tricky. When drawing the table, your JTable class will go through
         *each cell one by one. For each cell, this method will be called. We provide
         *instructions on how to get the data for the cell that is currently being drawn. First,
         *get the data from the current row, then get the data specific to the current column.
         */
@Override
public Object getValueAt(int row, int column) {
myData d =  dataList.get(row);
switch (column) {
            case 0:
                return d.getName();
            case 1:
                return d.getPhone();
        }
return new String();
}
}