w: Plus a little rant

We need to talk. If you’re a software designer or a programmer of some merit, we really need to have a quick discussion about your application.

Over the past year and a half or so, I’ve been scraping the landscape for text-based programs and trying them out for fun. It’s one part hobby, one part attempt to build something like a directory of available software, and one part pruning away the things that don’t work at this point in history.

In that time, I’ve come across a few programs or program suites that follow a Very Bad Trend, and I’d like to stamp it out before someone comes along and decides it’s a Cool and Hip Trend That Should Be Followed. People are occasionally dumb like that.

Back in August 1990, there was an RFC for “Choosing a Name for Your Computer.” Consider this an RFC for “Choosing a Name for Your Program.”

Do not, under any circumstances, give your program a one-letter name. It’s not clever, it’s not a gimmick, and it’s not innovative. No one will think you cool if you whittle “quasi-graphical multicolor modularized file manager and ftp client” down to “q.” You are not being sassy or stylish or minimalist by reducing “terminal-based weighted action priority list manager” to “t”.

There are already at least two programs named t — this one and this one. And there’s at least one j and at least one z. And at least one r. And and least one e. And at least one w. And of course there’s X, which you’re probably using now.

I said “at least” each time because there might be more. It’s hard to be sure. Because a single-letter program name is a sentence to death by obscurity. Have you ever tried to search for a program called “u”? Go ahead, give it a try. You won’t have much success; probably you’ll be lost in a maze of text messages and Twitter postings left by 14-year-olds complaining about their parents. If you get that much.

If you give your application a single-letter name, you’ll be consigning it to the backwaters of software popularity. It’s difficult to find, hard to distinguish and doesn’t show any real creativity.

Sure, you might be a minimalist and feel it’s necessary to express that by calling your program “c”. But that’s been done too. And think about it: Out of a planet of 7 billion, a substantial proportion of which can create and develop software, are you really expressing your individuality by picking one out of 26 possible letters, and plopping down your magnum opus at that spot?

What you do on your own system is your business. If you want to scratch out an alias of “q” that redirects to “quasi-graphical-multicolor-modularized-file-manager-and-ftp-client” then more power to you. I wholeheartedly support your right to abbreviate. I do it all the time.

But calling your latest iTunes knockoff “i” is just a wicked bad idea. Don’t do it. Not only will your program never, ever see any measure of subscription, but that little tiny “i” link is really hard to hit with the cursor. 😛

Oh, I almost forgot: Here’s w, which is part of procps-ng in Arch and shows up as w.procps in Debian’s procps.


Shows load average across users. Has a few options, but mostly does what you see there.

Never heard of it? Well, now you should know why. 👿

6 thoughts on “w: Plus a little rant

  1. John

    Actually I use w quite a lot, usually to check which display or tty a session is using, but on at least one occasion for the idle time field.

    I always assumed it was short for who (it sure does seem related – look at the description lines from their man pages). I further assumed that it got such a short name out of its age and ubiquity. But now that I think about it… I have no idea where I got these assumptions. ls and rm are short names from age and ubiquity, but they’re actually required by some standard (POSIX, right?), and I rather doubt w is.

  2. CorkyAgain

    I would add that if you give your program a one-letter name, you’re more likely to clash with an alias the user has created.

    For example, I’m one of those vi/vim users our host likes to poke fun at, and I have ‘e’ wired into my brain as the “edit” command. So “alias e=vim”.

  3. Pingback: Links 23/6/2014: New Releases of Opera, MakuluLinux, Netrunner | Techrights

  4. darkstarsword

    I too use w all the time – seems like it’s a fairly standard UNIX command and I always assumed that it got into one of the earliest UNIX OSs along with all the other obscurely named commands which gave it the right to claim that name, but I don’t have anything to back that up, and googling “history of the w command” doesn’t get me anywhere for much the same reasons you outlined in your post 😉

    I’m not sure if it’s mandated by POSIX or not – I just wasted half an hour trying to navigate the ieeexplore searching for IEEE Std 1003.2-1992 (POSIX Shell and Utilities), but it’s a damn mess and just makes me angry at standards bodies and scientific journals for hoarding information and demanding people pay them enormous amounts of money for the privilege to access it, then somehow failing to use that money to keep their papers organised or make their site slightly usable!

  5. Pingback: z and z: A tale of two z’s | Inconsolation

  6. Pingback: e: Like a diamond bullet | Inconsolation

Comments are closed.