Also it lowers the metabolic rate, permanently probably. I went on a near-carbless diet for a while (not to lose weight specifically, different problem) and regret it greatly.
Unfortunately, 0.1 seconds cannot be expressed accurately as a binary number
Is what they're bloody well talking about. As a floating point number. I am well aware how 0.1 could be represented perfectly, but that is not what the post is talking about.
True, to a lesser extent this is already true with files that have large holes in them, or are hardlinks, though.
ZFS makes it a little weirder I admit.
In other news, whenever you format an ext3 partition, remember that 5% is set aside for root use.
True, but you can tune that at mkfs time. The heuristic for how much space to reserve for other metadata (inodes) is probably off too for such large filesystems. (Or are they allocated dynamically nowdays?)
1024 is 2^10 - to be self-consistent, they should have chosen 2^8 or 2^16 for grouping, since 8 = 2^3 and 16 = 2^4, but they chose 2^10 because it happened to be "close to 1000"
Why is it significant what size the bit grouping is w.r.t. the base?
Why would it not be OK to have a grouping of 10 bits, but would it be OK to take a grouping size 2^3? 3 isn't a power of 2.
The decimal system (base 10) is grouped in 3 digits, and 3 is nothing significant base 10. To make it self-consisteny by your logic, it should be grouped in 10 or 100 digits.
3 is just to make it easy to read by humans; it's a good number of significant digits most of the time, when writing down numbers in decimal, and has little to do with the underlying system.
I'm not going to be the one that says a little girl was hot. :)
I meant Christina Ricci now, not then :-)
Alpaca!? I thought he was setting up for a peruvian women joke!
if you're going with the addams family, why in god's name didn't you go with Christina Ricci ;)
The main defense for any aircraft carrier is the billions of dollars of support ships
I'm not saying you'e wrong,but: cost != value :)
No, because the posting un-does the moderation. That was the point of it. As he said in his posting. "Now I'm posting to undo it"
that wasn't THAT obvious :)
big iron != bare metal, which is what GP presumably meant.
queue.
unary would use only zeroes :)
Cue.
asshole of the day award: awarded.
Also it lowers the metabolic rate, permanently probably. I went on a near-carbless diet for a while (not to lose weight specifically, different problem) and regret it greatly.
Unfortunately, 0.1 seconds cannot be expressed accurately as a binary number
Is what they're bloody well talking about. As a floating point number. I am well aware how 0.1 could be represented perfectly, but that is not what the post is talking about.
True, to a lesser extent this is already true with files that have large holes in them, or are hardlinks, though. ZFS makes it a little weirder I admit.
Except most continents are connected via fiber which isn't that high latency.
It's not the medium, it's the distance.
The single block is still stored redundantly, of course. Just not redundantly more than once.
The wording was a little inaccurate. They meant to say that the number 0.1 can't be represented exactly, of course. Which is true, of course.
Drive.
i was going to build myself a HTPC with a 30GB OCZ drive, but i think i'll leave it a bit longer if this is the case..
And I ordered parts for a HTPC with a 30GB OCZ driver. Last night.. Nevertheless I'm not changing my mind :)
From "I dated a robot"? "The Devil's hands are idle playthings" is a better match imho :)
("They're very good hands!")
The problem is that all these companies have set themselves up as fall guys
Nitpick: that's not what a fall guy is.
Good point, I wish everyone realised it's a flawed research strategy :)
http://www.youtube.com/watch?v=CQmoTJFScMY
In other news, whenever you format an ext3 partition, remember that 5% is set aside for root use.
True, but you can tune that at mkfs time. The heuristic for how much space to reserve for other metadata (inodes) is probably off too for such large filesystems. (Or are they allocated dynamically nowdays?)
1024 is 2^10 - to be self-consistent, they should have chosen 2^8 or 2^16 for grouping, since 8 = 2^3 and 16 = 2^4, but they chose 2^10 because it happened to be "close to 1000"
Why is it significant what size the bit grouping is w.r.t. the base?
Why would it not be OK to have a grouping of 10 bits, but would it be OK to take a grouping size 2^3? 3 isn't a power of 2. The decimal system (base 10) is grouped in 3 digits, and 3 is nothing significant base 10. To make it self-consisteny by your logic, it should be grouped in 10 or 100 digits. 3 is just to make it easy to read by humans; it's a good number of significant digits most of the time, when writing down numbers in decimal, and has little to do with the underlying system.