Caveat Emptor

16 Oct 2009

Now I will disclose one of the most secret of secrets of software business: All software is bad. Except maybe for very rare specimens that are even worse. It does not matter whether it is commercial or open-source, young or mature, big or small, it is bad. The quality is universally low. I'm not talking just about the bugs, but about all software qualities: performance, scalability, understandability, flexibility, visibility, reliability and security.

I cannot remember if I have seen good software in my entire career. I mean a software that was appropriate for the purpose. Software that worked as expected. Worked not only on the day one, but even 10 years later. Software that could be evolved without undermining its basic architectural principles. Software that was intuitive and easy to use, well documented, secure, ...

The reason for this situation is not technological. It is not that we software engineers are ... ehm ... idiots. We are not. We are doing our work well, considering the circumstances. The reasons are purely economical. It is just not profitable to create good software. Bad software can be very successful on the market. Quality is not high priority when making software purchasing decisions, but features are. Quality is difficult to understand and it is usually not directly visible. However features are outright visible and can be presented in in an impressive way. Quality can be usually seen only after the system survives first few years under production load. That's the point where the defects will manifest themselves, usually in a spectacular way. But at that point the software is already purchased and strongly hardwired in place.

From this point of view it is just a plain waste of money to invest in quality. Increased quality will not increase software sales and therefore not increase profits of software companies. In fact it may even harm their business: higher quality means lower motivation to purchase support services. Quality is not a competitive advantage. Spending more money on quality is a competitive disadvantage.

Therefore, caveat emptor.