One nice thing about the internet are the moments, where you learn a random fact , while searching for something completely different. I don’t even talk about unexpected appearance of problems that are hard to solve. A situation that leaves you a little bit more knowledgable1 after you have actually solved the problem. I think of even more random facts that come along the way in the process. Intellectual
2 by-catch, so to speak. This by-catch came along when i searched for the reason why tests are only performed when you chose never
for the -n
option for the smartctl
.
For example when you search for a translation for a blog title about how a command line option of a tool behaves. The random fact this morning: There are languages with a concept called negative concord. Those langugage know the concepts of affirming a negative instead of canceling each other out. Old Englisho or some dialtects of English nows this concept, the german language no . The modern standard english doesn’t know it as well, but it seems to be in colloquial use. If i say in the undying words of Bob Marley “I shot the sheriff, but I didn’t shoot no deputy” , then the correct translation is “Ich habe den Sheriff erschossen, aber den Hilfssheriff habe ich nicht erschossen” but not “Ich habe den Sheriff erschossen, aber ich habe keinen Hilfssheriff nicht erschossen”. I really didn’t shoot the deputy, not a deputy massacre.
Or Pink Floyds “We don’t need no Education”, which correctly translates to “Wir brauchen keine Bildung”(in the sense of, we really don’t need it) and not to “Wie brauchen nicht keine Bildung” (in the sense of: we want education). There is a long blog entry about double negatives at the Wikipedia. Fits somewhat in overall impression of germans: We don’t have time, for such linguistic concepts … we have to work and to sort our paperclips. We have only the strange concept of adding words into gargantuan single-word constructs enabling us to have a word for everything, like “Rindfleischetikettierungsüberwachungsaufgabenübertragungsgesetz”. Efficient! We don’t like no inefficiency! Did i use this correctly? It’s not native to me ;)
Okay, why i’m writing all this: There is an option in TrueNAS scale to control the behaviour of the S.M.A.R.T service checking the hard disk health. By default it’s set to never. There is a nice question mark behind it explaining that the tests are only performed if “never” is selected. WTF?
At first this makes absolutely no sense whatsoever. It starts to make sense as soon as you look that it’s used for the -n
option of smartctl
. With the command you can prevent the tool from performing tests when the disk is spun down. It really makes total sense, when you look a the full name of the option: --nocheck
. So is option is a double negation. We don’t need no check! Okay, wrongly worded, that would be negative concord. More correct: There is no powermode of the disks where we don’t do a self test of the disk. If you set if for standby for example, the test is only done when the disk is working. If you set it to sleep the test should be only executed when the disk is not spun down.
That said, that never makes only sense when you keep the full name in mind. This morning i learned two things: Why my hard disks woke up without seemingly no reason. And negative concord.
PS: Okay, when i really think about it … shouldn’t --nocheck=never
translate to “I really don’t want your f… check! Ever!” in the light of negative concord?