Philosophy and Truth in Software

Neil deGrasse Tyson provides a basis for discussing truth. This foundation can be applied to software in similar ways and with striking parallels to modern science.

Objective Truth in Software

Every piece of software is an expression of a thought. This is similar to a scientific hypothesis or a guess that needs to be tested. Some software is right and some are wrong. Only when we have others use and review our software do we begin to change our software into well established mature applications (an objective truth).

Some objective truths in software are broadly applicable and are well established. Security software contain many examples, OpenSSL being a perfect example. This library is critical to secure a large portion of the traffic over the internet. It is also able to run on many different networked devices (i.e. cloud, PCs, TVs, phones and IoT devices).

Other objective truths in software are more localized and focused but are nonetheless very important and mature. Two such technologies competing in data center storage are, Fiber Channel and Infiniband. These two technologies have well established open source libraries and APIs that enable vendors to build storage systems that service compute resources in a data center.

Personal Truth in Software

This is the standard religious debate in software that we all know about. Unix geeks remember the VI vs. EMACS debate. More modern debates are, Linux vs. Windows or Containers vs. Virtual Machines, the list can go on.

The big difference between modern religious debates and software debates come from the universal objective truth that governs all software, Turing Completeness. Computability theory is a very deep subject and is an exercise for the reader. It practically means that computing professionals tend to pick a side so they only have to learn one thing instead of two, because they know the two sides are equivalent. This makes them more efficient doing their real work.

Political Truth in Software

The political truths in software are also well known. They happen at different scales from intra-personal to inter-company.

At the personal level we do this all the time. Software developers will jump to implementation to early without doing enough research. We may over-complicate the architecture trying to solve a larger problem the client does not have. We lie to ourselves, thinking what we are doing is right. Luckily, agile development methods curb these individual mistakes quickly before they become costly.

At the industry scale often powerful companies will push inferior products, competing in emerging spaces. Other products get pushed out, because the powerful company uses their market share to force the technology in a direction that benefits them. Power plays like this, set back technology in that space a decade or more, hindering innovation over that time.

Conclusion

Truth in software is still a moving target. I like to think I’ve got all the answers, but I know that I’m lying to myself. All I know is that the discussion above seems to fit my perspective on software.

My perspective is, there is no good and bad software. Software is like having a conversation with someone, only you wrote every line of your side years ago. So, bad software is just a one-sided conversation that completely missed its audience.