Humbug, the big difference is: in GC there's perfect knowledge about what you want to throw away (unreferenced or only circularly referenced objects), and the problem is how to find out efficiently.. In caching the whole problem is to figure out which objects to keep, and beyond that, efficiency is no problem.
NEGATIVE NATIONALISM
(i) ANGLOPHOBIA. Within the intelligentsia [slashdotters?], a derisive and mildly hostile attitude towards Britain [United States?] is more or less compulsory,
Now, from dictionary.com:
intelligentsia, n.
The intellectual elite of a society.
Puhh-lease.. slashdotters the intellectual elite of a society? When I think of a bunch of nerdy, pasty-white, geeky young guys, my next thought is not the intellectual elite of a society.. Smarter than average, yes, but that is not the same thing as intellectual, much less elite..
I'm a bit puzzled by GA hoopla as well. Sure it can achieve results, but it's something of a copout. If GA's can do a reasonable job at optimising a (fitness) function (that's, after all, all they really try to do), then surely there's an actual special purpose algorithm that can do a better job.. (Although that goes quite
directly to the heart of the P vs. NP question..)
I think it's just people throwing GA's at a problem because it's quite easy and works slightly..
I, for one, hope that one day someone will
say P=NP, despite all the cryptographic problems we might get into then..
I used to work at an ISP, and we found our ADSL users were using an average of 25kbit per second per session. At the time the connections were a megabit.
So that says something about how much capacity is necessary in total to keep your users happy, but it doesn't mean everyone would be just as happy with 25kbit connections:)
Well, indeed, he said 500 GB "after the hardrive manufacturers stretching of the facts and formatting.", so I presumed he wanted 500*1024*1024*1024 bytes after formatting, and not 500.000.000.000 bytes before formatting..
I need about 500GB and something that is reliable. I'm looking at 3 250GB drives with raid5 which should be close enough to 500GB after the hardrive manufacturers stretching of the facts and formatting.
3*250GB in raid5 is only 2*250gb of raw storage due to a parity disk..
I may be stating the obvious here, but with 5x400GB you'll lose at least one drive for parity if you want to do RAID5.
On the one hand RAID5 isn't ideal because updating the parity is slow, on the other hand that many disks in one unredundant array is asking for trouble, and RAID1 (or 0+1 or 1+0) is a bit wasteful, so it's smart.
Generally the bottleneck with regular use is seeks though, not bulk transfers.
Also the entire array will last only 1/3rd of the time of an individual disk, assuming a poisson failure rate (which is reasonable) because if any of the disks break, the entire array breaks.
You can do parallel seeks though, so the whole set can stand more of a thrashing than a single drive..
"You can put 100% pure H2SO4 in glass forever and not bother it one bit."
I may be stating the obvious here, but 100% pure H2SO4 isn't acid at all in the sense that it corrodes stuff, it needs water to react with in order to form the H3O+ ions (the actual acid that corrodes stuff).
As for sodium hydroxide, again I might be stating the obvious to you, but it isn't an acid, but a base - it forms HO- in water.
Slashdot Mirror
Humbug, the big difference is: in GC there's perfect knowledge about what you want to throw away (unreferenced or only circularly referenced objects), and the problem is how to find out efficiently.. In caching the whole problem is to figure out which objects to keep, and beyond that, efficiency is no problem.
Besides.. the intellectual elite is overrated ;-)
That would be non-chemistry people, IMHO.
There is a protocol (proposed standard). It uses RSA directly on the message, after mashing it, for space-efficiency.
Biometric encryption eh? That doesn't sound very good..
Hear, hear. I agree with you more than I can say without ending up on an NSA/CIA watchlist ;)
Interesting.. what political agenda could/would that be?
That's a good thing IMO :)
I'm a bit puzzled by GA hoopla as well. Sure it can achieve results, but it's something of a copout. If GA's can do a reasonable job at optimising a (fitness) function (that's, after all, all they really try to do), then surely there's an actual special purpose algorithm that can do a better job.. (Although that goes quite directly to the heart of the P vs. NP question..)
I think it's just people throwing GA's at a problem because it's quite easy and works slightly..
I, for one, hope that one day someone will say P=NP, despite all the cryptographic problems we might get into then..
i find the randomness of your suspicions highly useful.
I used to work at an ISP, and we found our ADSL users were using an average of 25kbit per second per session. At the time the connections were a megabit.
So that says something about how much capacity is necessary in total to keep your users happy, but it doesn't mean everyone would be just as happy with 25kbit connections :)
Also a videocard is convenient for installing the thing and for maintenance if there's trouble with it..
Well, nitpicks aside, you're right though.
Would that be.... "The Allen Parsons Project!"
BA-pam! BAA-pam!!
Perhaps?
You run sigs as root eh? Well then I have a sig to sell you ;)
Presuming of course spam is sent legitimately, and not by zombie boxes..
Indeed true, but snapshotting filesystems (or block devices) (storing every block that's used in the live FS or a snapshot only once) can fix that.
Well, indeed, he said 500 GB "after the hardrive manufacturers stretching of the facts and formatting.", so I presumed he wanted 500*1024*1024*1024 bytes after formatting, and not 500.000.000.000 bytes before formatting..
I need about 500GB and something that is reliable. I'm looking at 3 250GB drives with raid5 which should be close enough to 500GB after the hardrive manufacturers stretching of the facts and formatting.
3*250GB in raid5 is only 2*250gb of raw storage due to a parity disk..
On the one hand RAID5 isn't ideal because updating the parity is slow, on the other hand that many disks in one unredundant array is asking for trouble, and RAID1 (or 0+1 or 1+0) is a bit wasteful, so it's smart.
Just thought I'd add that. :)
Also the entire array will last only 1/3rd of the time of an individual disk, assuming a poisson failure rate (which is reasonable) because if any of the disks break, the entire array breaks.
You can do parallel seeks though, so the whole set can stand more of a thrashing than a single drive..
That is only true if the disk doesn't re-order the i/o in a metadata-inconsistent way, which isn't certain on IDE drives.
If it's not, what would the significance be? The factorisation is: 2 2 461 11519 but that doesn't look interesting to me.
Googling for it I only find, as interesting reference:
- An
entry for
something called dipeptidyl anminopeptidase that sounds like a protein or enzyme
But I'm sure that's not itAh, in that case, sorry to explain that basic stuff to you :)
I may be stating the obvious here, but 100% pure H2SO4 isn't acid at all in the sense that it corrodes stuff, it needs water to react with in order to form the H3O+ ions (the actual acid that corrodes stuff).
As for sodium hydroxide, again I might be stating the obvious to you, but it isn't an acid, but a base - it forms HO- in water.