Lair of Peppe

Every software is different. Therefore you shouldn’t use the same approach to testing all software that you create, but refine the testing processes at all times.

Taken from Mauro Pezzè’s and Michal Young’s book “Software Analysis and Testing” here are the questions that a quality manager should have the answers to:

• When do verification and validation start? When are they complete?
• What particular techniques should be applied during development of the product to obtain acceptable quality at an acceptable cost?
• How can we assess the readiness of a product for release?
• How can we control the quality of successive releases?
• How can the development process itself be improved over the course of the current and future projects to improve products and make verification more cost effective?

Source: Pezzè M. and Young, M.  (2007) Software Testing and Analysis: Process, Principles and Techniques. John Wiley & Sons

I am interested in always enhancing my work flow when developing software. That includes taking new tools into use and making the work techniques better. As I work with Vaadin, my focus is right now in making the testing process of Vaadin applications better. Testing of Vaadin applications is possible right now with Testbench, which records usage of the web app and replays it by actually clicking automatically around in a browser. It is just a bit cumbersome and a tad slow. The goal is to write a paper on testing Vaadin applications as well as provide some kind of framework where you can test your application with JUnit, and more importantly, without actually needing to run a graphical user interface. To do this, however, I have to dig quite deep into software testing and analysis. A lot in that field is quite new to me. To remember important concepts and to revive this blog, I’ve decided to write here short texts every now and then that I find on the topic. Hopefully this journey will end in me providing a tool which would make verification of Vaadin applications easier.

Today I had to do a lot other stuff than on this project. I actually did some of my normal work for a change, went to school to attend to a lecture and went to see some stand-up in the evening. I could only squeeze in work on this project between 12:00-15:00 and 22:00-02:00. It wasn’t really enough. I finished some basic views for all three main objects (players, tracks, rounds) but then I noticed a showstopper bug in my software regarding db4o. To add a new round, I create a round object, fetch the players and a track and set these to the round, fill in all the scores and tell db4o to store the round. Every time I stored a round, db4o also stored a new copy out of the players and round into the database. I ended up pretty fast with hundreds of copies of the same objects, which pretty much made the whole software useless. I have now spent a couple of hours trying to figure it out, but in vain. The view pages are somewhat limited, and I still have my home screen completely undone and it can’t really be done before the db4o problem is solved. I guess I can squeeze an extra hour to this before the deadline 12:00 tomorrow, but it’s 2am now and I want some sleep!