4 ms·
Would you make the same argument for programming languages? Should anyone feel free to code in whatever "vernacular" style he please, because living up to a def
by aaronem 13y ago
Would you make the same argument for programming languages? Should anyone feel free to code in whatever "vernacular" style he please, because living up to a defined standard makes it "hard to express [one]self"?
No doubt someone will attempt the argument that there is a qualitative difference between a language in which one speaks to a computer, and that in which one speaks to other people, but in this case I contend it's a distinction without a difference. The very reason why coding standards exist, after all, is because what we write in programming languages must be understandable to other people, no less than to the computer, and supine permissiveness as regards writing style impairs the legibility of one's code for the other people who must also read it.
Granted, it takes more effort to live up to the coding style required by whatever project one happens to be working on, just as it takes more effort to express a thought in Wallace's "SWE" than in whatever corrupt vernacular happens to be one's native pidgin. In exchange for this additional effort, one produces a result which is both more generally legible and more likely to be taken seriously. (On that latter point in particular -- would you have read this far if I were writing, not in that acrolect Wallace and his latter-day imitator so deride, but rather in my native Southern American English? Don't seem any too likely to me.)
- sp332 13y agoIf you're writing authoritative, reference-style code, then formal standards may be appropriate. If you're part of a huge project, writing code that is readable by everyone else on the project is simply a pragmatic thing to do. But standards can go too far even in those cases. I once worked on a project where we weren't allowed to depend on order of operations for math operators - we had to put parentheses everywhere. While it may have avoided some bugs, the coding style didn't allow for much self-expression. When you're writing code for a smaller audience, or just one-off code for your own use, standards are much looser. If you're writing a one-liner in Perl, maybe there aren't any standards! And that's the point of the article. It doesn't make sense to apply strict rules about what you can say and how you say it when you're just updating your facebook status or making a blog post. I mean you could if you wanted to, just like "you can write FORTRAN in any language." Yes, I would have read your post all the way through, and that judgmental attitude you reference is just what the author is arguing we should get rid of.
- aaronem 13y agoGranting arguendo that the phrase "self-expression" is useful, you seem to assume such activity is impaired both by the imposition (even the self-imposition) of standards, and by the quest for precision in one's use of language. This assumption strikes me as highly questionable; if anything, lack of standards and disregard for precision in one's effort seems likely to make accurately saying what one means more difficult, rather than less so.
- sp332 13y agoYou can be extremely "precise" without conforming to a strict generic format chosen by someone else who didn't know anything about your project. I don't know why you consider conformity to externally-imposed rulesets to be equivalent to accuracy of expression. Sonnets can be beautiful and meaningful, but imagine what a pain it would be if everything were written that way.
- aaronem 13y agoWhy assume, as you seem to, that all standards originate outside yourself? Do you not require of yourself a certain level of quality and precision, the degree of both depending on project and audience, when you write either natural language or code?
- sp332 13y agoYes, and I feel like I have to translate when I'm writing for a general audience. Anything I write for a general audience is stilted because I have fewer options available. English is a mix of many different influences, each of which was novel at some point in history. If people want to make up new rules to experiment with (like doge and ragecomics), that's not a bad thing. The more people we have who feel free to experiment, the more options we'll have in deciding what we want language to look like in the future.