"It helps to know some music theory if you write and perform music, but a lot of very successful songwriters and performers get by very happily with just enough music theory."
This is an absolutely bogus comparison. Music does not need to be maintained, music does not need to be troubleshooted or debugged (and thus, reasoned about), music does not solve a problem(1).
Music is an artistic expression, while computer software isn't.
(1) Writing soundtracks for movies, for example, can be considered solving a problem with music. It also requires music theory knowledge because soundtracks have to convey particular emotions at particular moments.
I used to compose music solely by playing around on an instrument until I hit upon something that sounded good, but I had no idea why it sounded good.
I can still compose that way, but after studying music theory, I can compose directly from my head to a sheet of paper without any instrument, as (1) I can now envision in my mind what the music would sound like, and (2) I understand and can employ reusable patterns and concepts of composition that I know will sound good and work together.
Similarly, when I started programming, I had little overall vision for what I was doing. I just started writing code, and stopped when I had built up a pile of spaghetti statements that did what I wanted them to do. Now, with better understanding of design concepts and reusable patterns, the development process is much more clean and structured rather than poking around with guesswork.
Making a noise doesn't need to be troubleshooted or debugged. Making music does, unless you're trained in musical theory well enough to not make the mistake. There are many compositions that come across my desk for advice when the author knows there's something missing, something wrong, something out of place but they can't tell what.
You can have clashing chord progressions which are "good enough to ship", but to a keen listener might throw the whole performance away. It's not a perfect analogy (none of them ever are) but it might be more apt than you think.
"It helps to know some music theory if you write and perform music, but a lot of very successful songwriters and performers get by very happily with just enough music theory."
This is an absolutely bogus comparison
I disagree. There are very successful songwriters and performers who have no training, but almost all of the good songwriters and performers did have (usually classical) musical training.
The same is absolutely true in software: There's a lot of very popular crap out there, but the software which is universally recognized as good -- code like TeX -- almost always comes from authors with solid computer science training.
There's some good stuff in TeX but it's been buried in an avalanche of shit from people who aren't Knuth. You can get TeX in a couple megabytes (http://www.kergis.com/en/kertex.html), but if you install something like pdfTeX or XeTeX or TeXlive, well I hope you've got a big hard drive.
> Music is an artistic expression, while computer software isn't.
Unless it is? The sentiment that the pursuit has to be entirely about doing a job or providing utility is frighteningly inhuman. I need to go for a walk after reading that. Or maybe watch some demoscene.
This is an absolutely bogus comparison. Music does not need to be maintained, music does not need to be troubleshooted or debugged (and thus, reasoned about), music does not solve a problem(1).
Music is an artistic expression, while computer software isn't.
(1) Writing soundtracks for movies, for example, can be considered solving a problem with music. It also requires music theory knowledge because soundtracks have to convey particular emotions at particular moments.