Investigating Software

Investigating Software

Sunday, 12 June 2011


Do you ever examine what you carry around with you every day, and wonder if you actually use it? For example, in my pockets I've got a 'smart' phone, wallet (credit & debit cards, cash and ID), keys, Travelcard (Oyster) and some coins. Every now and then something gets added, if its unlucky it stays. Over the years, I've noticed, that the criteria for being kept is usually convenience or enablement. That is, the items that don't get chucked or deposited somewhere about my home are usually 'tools' that make other 'things' easier like a smart-phone - I can just look up something or text someone at any time. I could just wait until I got back to my office, or see the person later but it can be easier to just act in the moment, and do it there and then.

Enablement items, are things that mean I -can- do things, that without, I'm stuck. For example: door keys. The smart phone fits into this category also, if I want to meet up with someone at short notice when I'm out, then a mobile phone is really the only practical option. Cash is another enablement device. In the same theme, I wear a watch. It enabled me to know the time, with the added convenience of a date function. The watch allows me to be less tardy, as it's a much better timekeeper than my brain.

The great inventions of our and recent times have tended to be enablement devices also. Aviation, antibiotics, motor car, personal computers, MRI, the Internet/telecommunications etc are all enablers. We can do things our ancestors couldn't, and we can do them affordably. These are all tools. They extend our reach and our capabilities.

We use tools everyday in software development. Some, like the things in my pockets, are junk, and I soon ditch them. Some I keep and reuse. But in the long run, it's the convenience and enabling tools that stick. It often seems to take me (at least) a while to notice which tools are helping, which are not helping or even hindering. Part of the problem is that some tools look like they work or at least -should- help but don't. But still I've noticed some tools or groups of tools that usually help, and in one form or another these have survived.

These include the following:

Randomness: - I have several scripts and tools that help me produce randomness in either text, numerical or event form. For example I have a script that produces a random series of UTF-8 codes, another that randomly clicks on parts of the user's graphical interface. This is an enabler. I noticed that I am not random, even when I think I am being random - I'm not. So these tools fill that gap and let me see what normally I couldn't: how a piece of software behaves with a variety of inputs [that are more diverse than I can dream up at the time]

Accurate - Record keeping: I use Blueberry TestAssistant to record everything I do on Windows. This screen recorder records the screen as well as logging all key presses. This enables me to keep accurate records of my work and findings, without much effort. This frees me to spend more time finding new issues, rather than being bogged down in manual note keeping. It also allows me to go back in time and search for issues I might have missed first time around.

No comments:

Post a Comment