Working Up

by Dwayne Phillips

Want to know what computer software is doing? Go to the source of knowledge regarding it.

I run a bash script daily to put my Internet viewing log online. (See here.) This script runs through a series of folders, us old guys still call them directories as in the Unix days, copies files, runs a (secure) File Transfer Protocol, another old-guy thing, and puts it all online.

I don’t remember the details of all this. If I want to refresh my memory, I look at the text of the script, a.k.a., the source code. It lists each individual, tedious step that I have the computer do for me everyday. That is why I have written a computer program of sorts—this is repetitive, tedious, and error prone.

At my current job, I find things in the software that don’t quite make sense. One of the programmers will often tell me, “Let’s look at the source code to see what we are actually doing.” Such readings of the source code often bring surprises. “This is what we are doing,” he’ll tell me. “Is that what we are supposed to be doing?” he then asks.

And now we have the title of this post. The source code, as we programmers have called it for many years now, bears that name for a reason:

the source code is the source of all knowledge regarding a computer program

If there is any question about what a computer program is doing, go to the source. Read the source code and learn. Forget the documents, forget the conversations, forget all else. The source code is the definitive answer, i.e., it is the actual definition of the software.

It is fortunate that we have the source code.

It is unfortunate (in my humble opinion) that in many other parts of life, there is no source code. There is no actual definition, or we don’t want to reread the definition. See, for example, marriage vows.