Talk:378: Real Programmers
In the beginning was UNIX. And it was good. And it was written by some very clever people.
One of the first very useful tools they wrote was ed, a "line-editor" (i.e. it works one line at a time). It uses some simple commands, and was created to work on very-old-school teletype machines, where you type a command, and ed types a response back.
It was a lovely bit of code. Using very little the way of resources, it allowed you to create a text document of any length, including source code in whatever language you wanted to program in.
Eventually, a more sophisticated version called ex (short for EXtended) was written by a clever man named Bill Joy. While it has some great improvements over ed, it was still a line-editor.
The trouble was, using a line-editor like ed or ex requires you to have a very good mental model of the document you are creating. Unfortunately, humans aren't very good at this, so they constantly need to refresh their mental model by printing out big chunks of the document (or program) they are working on. This took a LOT of paper using teletypes.
Eventually, teletypes were replaced with terminals. This saved a lot of paper. But the people who created the terminals began making them smarter than teletypes, so that magic character sequences could be used to move the cursor around, rather that simply going character-by-character across the line, then scrolling down to the next line, and so on. This opened up a whole new world.
The very clever Bill Joy took advantage of these magic character sequences to create his wonderful "full-screen" text editor vi. vi was the "VIsual mode" of ex. With vi, the user could see a screen-full of text at once. Entire forests were saved.
Emacs was developed at the same time as vi, using the same magic characters, and was also a full-screen text editor. I've never used it, so I can't speak to its merits, but there are many people who still find it more useful than any GUI they've tried.
On the one hand, vi and emacs are more sophisticated tools, and thus take longer to learn to use than ed. However, once you learn to use them, they make writing code EASIER, and they are therefore considered a less praise-worthy way of writing code by those concerned with defining what a "Real Programmer" is. (In other words, those programmers suffering from testosterone poisoning.)
Using cat to write a program looks like this: (Note that the $ is the prompt provided by the computer. The rest is typed by the user. And the ^D means the user held down the control key while typing the letter "d".)
$ cat | cc
The user types C code here, and ends with ^D. Assuming all goes well, the compiler silently finishes after creating the executable program a.out in the user's current working directory.
The reason this is considered a more praise-worthy way of coding is that, in those early days, doing this meant that your code was lost the instant you typed it. If you made a mistake, you would have to type the whole thing again. So doing this for code of any sophistication was considered an act of courage, confidence, and conviction. (I myself did it several times, for the fun of it, when no-one was watching, though never for a program that took more than about 30 lines of code. I was delighted that it worked all 3 times, but since I love to write re-usable code, this wasn't really something I wanted to keep doing.)
NOW PAY ATTENTION. VI IS NOT VIM! Vim was written in 1991, long after more sophisticated shells were created that made it possible to copy and paste text from one part of the screen to another. This ability greatly reduced the risks of using cat to pass your source code directly to the compiler, so it was no longer a praise-worthy stunt. Thus the line "Real programmers use vim" was NEVER considered true by any UNIX programmer.
Whether this was a mistake of the author, or the character (possibly Megan?) is unclear. It seems possible that it was a simple typo, but since I've never seen one in the strip before, I'm somewhat skeptical.
cat | ccdoesn't work on my system. My
ccis simply a symlink to
gcc; what's yours? --Lucaswerkmeister (talk) 10:16, 7 August 2013 (UTC)
I would never claim to be an emacs expert, but I'm reasonably proficient in it. Command of the form M-x (whatever) are a way of calling commands (or, really, arbitrary functions in the emacs code) by name. SO 'M-x butterfly' means that there is a function named "butterfly" somewhere, but that it has not been assigned a keyboard shortcut (or it has, but you're calling it the long way).
Also, three are still advanced Linux programmers today who swear by vim or emacs being superior to IDEs. There are specific technical reasons for this: emacs is basically an IDE construction kit that's incredibly easy to extend and customize, and is more customizable than pretty much any other program in the history of software with the exception of a Smalltalk installation. And vim has highly evolved commands to give experts a superhuman typing and editing speed when coding.
So when someone claims that "real programmers use vim," they are claiming that RIGHT NOW, vim is the best possible editor for developers of sufficient competence. There's a community of very smart people that basically thinks this.