In Windows, a few important files for the semi tech savy are buried in the location “c:\windows\system32\drivers\etc\ …”
I’m pretty sure this path was chosen becuase Unix also uses a similar “etc” directory (Someone will surely come along and confirm or deny that)
Anyways, is the “etc” short for “et cetera”? It seems some of the files located here (the host file for example) are actually pretty important and not really worthy of “and such and such” designatio that et cetera entails. Perhaps etc stands for something else?
etc does, in fact, mean et cetera (L., `and others’), and it’s a kind of catchall directory. In *nix-type systems, it’s full of things like configuration files and such. Essential, but too varied to be squirrled away in some endless subdirectory structure. That’s the main reason: If you didn’t have a catchall, you’d have to go through the rather inane process of giving every single file its own directory.
A more descriptive name may be “misc,” but “etc” works. Its amazing how many computer terms come about because of arbitrary decisions of one person a decade ago.
Perhaps “etc” was chosen instead of “misc” because of some archaic 3-character limit (didn’t some systems have 36-bit words and 12-bit characters?). Witness some other common directories in the Unix root: “lib”, “bin”, “var”, and “usr”.
“etc” was chosen because in the early days of UNIX everything was made as short as possible to decrease the likelyhood of typing errors on expensive time-sharing machines. So all your essential commands are two characters (cd, mv, ls, ln, rm and so on) and all your essential directories are three and four letters (/usr, /etc, /opt, /root, /home, /bin, /sbin, /lib, and so on)
Plus, really short (Hebraic, almost) commands and directory names reduce the wear and tear typing imposes on your hands and wrists. Carpal-tunnel syndrome may not have put you on disability back then, but it was certainly real enough to avoid.
The last half is easy… Just copy random noise to the speakers.
And I put in the missing spaces between the strings in the secnd ‘tr’.
Unless you use a different ‘tr’ to the one I have, it requires two strings, in and out, so I tried tr ‘0-9’ ‘a-z’, and plenty of variations thereof. Only gibberish.
You now owe me three quarters of a lunch break. Maybe I can pay it off with this:
The result isn’t a word, really: It’s a common Usenet abbreviation.
And a tr with no second string replaces the characters in the set it does have with nothing: It’s a way to filter out noise characters if they all fit in a relatively easily-defined set. In ed, s/foo//g does the same thing.
For those of you not enlightened enough to use a Real OS who are wondering what the hell the geeks are on about now:
In my initial string, the one following the command echo, all of the lowercase letters and numbers are garbage. The first tr takes care of them. What’s left is four uppercase letters that also resemble garbage. They go through the second tr, which implements the rot13 algorithm to decode them into HAND, a Usenet abbreviation for Have A Nice Day. The part after the semicolon is pure misdirection and simply copies pseudorandom noise (from the *nix device /dev/urand) to the speakers (represented as /dev/audio). The su command is `set user’, which in this case enables the command following it (enclosed in quotes) to be executed as if the superuser (or root) had typed it in on his terminal. I did that because not every system allows a nonroot user to dd garbage into the speakers.
I’d guess 3 characters were chosen because PDP-11 had Radix 50. This was a way to cram 3 characters into a 16 bit word.(Obviously all ASCII characters could not be used.)
When I worked in Bell Labs I always heard it pronounced “etsee”
Alas, I don’t have my old 11/20 manual anymore, and the PDP 11 programming card which is at hand doesn’t have Radix 50 - but it does have the code for the bootstrap loader.
