You say their pre-marathon stuff was unremarkable? Curious, have you ever played Pathways into Darkness? I'm not a game tech guru, so it might have been boring from a tech perspective; but, oh man, it had a great plot! I know that sounds kinda chintzy but was the kind of game that really drew you in. Marathon 1 and 2 were even better in this regard.
sorry to nitpick, but I have a problem with you saying that bounds are misleading (and, more so, it seems ribbing "theory"). I mean, it all depends on what you want: bounds are misleading if you want to predict behavior under typical conditions. But medians are just as much misleading if you want to predict things like how the longest/shortest you can go on a single charge. I think knowing these bounds is very important.
I would saying any typical metric is more misleading, because, while bounds are really hard-and-fast, typical behavior depends completely on usage patterns. (e.g. I never use my CD player while on battery, and I dim the screen, and spin down the hard drive, and run the processor at half speed. But do you?)
Actually, isn't the "nobel prize for economics" not one of the original prizes that Nobel wanted. He thought economics was a hack, or something? The original five were part of Nobel's will, in 1895. It wasn't until 1968, in a plan cooked up by the Bank of Sweden to garner some karma, that the foundation added a "memorial prize for economics".
True true! By "algorithms" I meant both. I mean, you can't really have one without the other (when programming; which is what the original conversation was about).
You said: Programming isn't just algorithms, it's mostly heuristics.
I say: Ay caramba! I'm sorry, but it (1) all boils down to algorithms. Remember, an algorithm may be a heuristic algorithm, but it is an algorithm nonetheless. Perhaps you meant to distinguish algorithms from hacks.
(1) "it" being "good programming". good meaning whatever you want it to:)
Actually, there's nothing special about the non-uniformity (NU) of access. You have that feature whenever you throw in caches, for example: the time to access a memory location depends on whether it's in cache or not. A Beowulf cluster is also a non-uniform access architecture: accessing your local memory is cheaper than accessing the memory on a remote machine (via some sort of message passing layer, such as MPI or PVM or RPC, for example). Network file systems also exhibit non-uniform access cost behavior: the speed at which you can access/net/machine/filename depends on whether "machine" is local or remote, and how far away it is.
So what makes a "NUMA machine" so special? It's the hardware cache coherence. That is, a cc-NUMA turns a (cheap) Beowulf cluster into an (relatively expensive) IBM machine.
Whenever access costs are non-uniform, algorithms must be tuned to be latency sensitive. If algorithms are not aware of this nonuniformity, then they'll run really slow!!
Why is hardware cache coherence silly? eheh (my opinion). If we need to make algorithms latency sensitive to take advantage of non-uniform access architectures, then we need explicit control of data movement throughout the system. But a cc-NUMA takes this control away from you.
Notice that your L2 cache effectively has hardware cache coherence. You can't really control what's in L2 cache. And neither do your algorithms need to be conscious of the existence of an L2 cache to perform correctly. To perform well, however, they do.
Not sure if I buy it that "wading in, poking around" is impossible with CLI. That's what "foo --help" or "man foo" or "info foo" are all about. Thought about that way, a GUI is a sort of table-driven CLI, with --help being a default option.
As far as hierarchical organization, many operations are not hierarchical (at least at the top level), nor even sequential: they are parallel, which the piping operation captures. E.g. to count all non-comment lines in a document, you know that it'll involve awk/sed (to strip off comments) and wc (to count lines). To construct a composite command which accomplishes this task then requires filling in the arguments to each command. Isn't this parallel and then hierarchical (i.e. a forest of trees?)
You shouldn't care about choosing the noise-pad size to be large enough. You can always split messages across several communications, right?
Re:Good...weeded out the idiot day traders
on
Tech Stocks Tumble
·
· Score: 1
methinks you don't me to chastise day traders (who were likely out before things got bad; after all they are day traders, no?), but instead you mean foolish people investing their entire retirement stash in internet stocks (invest-and-forget investors)
Now, I agree that the latter were foolish; but really, how many of those do you know? Furthermore, even those "safe" I&F's who wisely invested in the SP500 aren't doing so well this year.
No, it is the day traders who have made off like bandits this year. THey have reaped the profits from early this year, and laughed to the bank on this downtrend.
sorry for the missssspellling. I meant it: the suggested plot would only work if the writers were willing to break up "voyager II" into lots of oddly shaped bits:)
Nah, won't work, for the same reason voyager didn't: you KNOW they are going to succeed, at least in some measure, at least for some time, because if they don't, then the series ends.
Unless the writers are bold enough to deal with the rogue starship being "jerrymandered", etc.
wasn't that about the planet 180 degrees out of phase with us, but in the same orbit, so that we'd never know of it's existence, using ground-based observation?
Lemme see if I got this straight: Compaq buys RAM from Micron, somehow certifies it (i.e. sticks a "Compaq" sticker on it, and runs a few diagnostics), and for this charges a 5x markup? And all for what? To amortize the cost of taking the blame away from Micron, even though Micron is still eventually the one to blame?
(well, it was at least 5x for their DEC workstations).
Ok, this is not just about companies just spinning off companies they've acquired (this post with Compaq+DEC, other posts about USR+3com).
Compaq acquired DEC, not for it's Alpha or research labs, but for it's high-end know-how, and for it's world-class NT support team.
3com acquired USR, not for it's modems, but for Palm!!!
Therefore, it makes sense for 3com and Compaq to spin off (or sell) the unwanted portions of those business they acquired for their (to them, at least) tastier aspects.
Will I be able to use other filesystems, such as FFS or ext2, jfs, etc, on OS-X? I guess what I'm asking is this: is any functionality of OS-X tied to features of HFS+ (e.g., does it depend on case insensitivity?).
drool.. I like your thinking, Aaron. My phd thesis is about a limited form of meta-reasoning for compilers. Whenever I see someone else who groks the power of "meta", I am happy. thanks!
That analogy doesn't seem to quite work either. We can't equate movie tickets with "right to see movie" (which would then somewhat match up with buying a DVD). When you buy a ticket, you have permission to sit in a very well specified set of seats: in this movie screen, and this time.
The difference is that, in the case of movies-theathers, there is, in most cases, an abstraction between "seeing a movie", and "where I see it". THere are some exceptions, like Lucas originally thinking of restricting Phantom Menace to certain "blessed" theaters.
Whereas, with DVDs, "buying a DVD", and "how I view it" are intimately tied. The ones producing the DVDs are the same ones blessing the DVD players. nick
Slashdot Mirror
You say their pre-marathon stuff was unremarkable? Curious, have you ever played Pathways into Darkness? I'm not a game tech guru, so it might have been boring from a tech perspective; but, oh man, it had a great plot! I know that sounds kinda chintzy but was the kind of game that really drew you in. Marathon 1 and 2 were even better in this regard.
sorry to nitpick, but I have a problem with you saying that bounds are misleading (and, more so, it seems ribbing "theory"). I mean, it all depends on what you want: bounds are misleading if you want to predict behavior under typical conditions. But medians are just as much misleading if you want to predict things like how the longest/shortest you can go on a single charge. I think knowing these bounds is very important.
I would saying any typical metric is more misleading, because, while bounds are really hard-and-fast, typical behavior depends completely on usage patterns. (e.g. I never use my CD player while on battery, and I dim the screen, and spin down the hard drive, and run the processor at half speed. But do you?)
off topic???? think F_U_U__U_U_H_N_N_N_YY__Y_Y
Actually, isn't the "nobel prize for economics" not one of the original prizes that Nobel wanted. He thought economics was a hack, or something? The original five were part of Nobel's will, in 1895. It wasn't until 1968, in a plan cooked up by the Bank of Sweden to garner some karma, that the foundation added a "memorial prize for economics".
or use larger fonts, my friend, or use larger fonts. :)
True true! By "algorithms" I meant both. I mean, you can't really have one without the other (when programming; which is what the original conversation was about).
I say: Ay caramba! I'm sorry, but it (1) all boils down to algorithms. Remember, an algorithm may be a heuristic algorithm, but it is an algorithm nonetheless. Perhaps you meant to distinguish algorithms from hacks.
(1) "it" being "good programming". good meaning whatever you want it to :)
Actually, there's nothing special about the non-uniformity (NU) of access. You have that feature whenever you throw in caches, for example: the time to access a memory location depends on whether it's in cache or not. A Beowulf cluster is also a non-uniform access architecture: accessing your local memory is cheaper than accessing the memory on a remote machine (via some sort of message passing layer, such as MPI or PVM or RPC, for example). Network file systems also exhibit non-uniform access cost behavior: the speed at which you can access /net/machine/filename depends on whether "machine" is local or remote, and how far away it is.
So what makes a "NUMA machine" so special? It's the hardware cache coherence. That is, a cc-NUMA turns a (cheap) Beowulf cluster into an (relatively expensive) IBM machine.
Whenever access costs are non-uniform, algorithms must be tuned to be latency sensitive. If algorithms are not aware of this nonuniformity, then they'll run really slow!!
Why is hardware cache coherence silly? eheh (my opinion). If we need to make algorithms latency sensitive to take advantage of non-uniform access architectures, then we need explicit control of data movement throughout the system. But a cc-NUMA takes this control away from you.
Notice that your L2 cache effectively has hardware cache coherence. You can't really control what's in L2 cache. And neither do your algorithms need to be conscious of the existence of an L2 cache to perform correctly. To perform well, however, they do.
Not sure if I buy it that "wading in, poking around" is impossible with CLI. That's what "foo --help" or "man foo" or "info foo" are all about. Thought about that way, a GUI is a sort of table-driven CLI, with --help being a default option.
As far as hierarchical organization, many operations are not hierarchical (at least at the top level), nor even sequential: they are parallel, which the piping operation captures. E.g. to count all non-comment lines in a document, you know that it'll involve awk/sed (to strip off comments) and wc (to count lines). To construct a composite command which accomplishes this task then requires filling in the arguments to each command. Isn't this parallel and then hierarchical (i.e. a forest of trees?)
You shouldn't care about choosing the noise-pad size to be large enough. You can always split messages across several communications, right?
methinks you don't me to chastise day traders (who were likely out before things got bad; after all they are day traders, no?), but instead you mean foolish people investing their entire retirement stash in internet stocks (invest-and-forget investors)
Now, I agree that the latter were foolish; but really, how many of those do you know? Furthermore, even those "safe" I&F's who wisely invested in the SP500 aren't doing so well this year.
No, it is the day traders who have made off like bandits this year. THey have reaped the profits from early this year, and laughed to the bank on this downtrend.
sorry for the missssspellling. I meant it: the suggested plot would only work if the writers were willing to break up "voyager II" into lots of oddly shaped bits :)
uhhh, I didn't say it was real, buddy! I just said that it was rumored to be true, at some point in the dark ages of the 19th century! :)
Nah, won't work, for the same reason voyager didn't: you KNOW they are going to succeed, at least in some measure, at least for some time, because if they don't, then the series ends.
Unless the writers are bold enough to deal with the rogue starship being "jerrymandered", etc.
wasn't that about the planet 180 degrees out of phase with us, but in the same orbit, so that we'd never know of it's existence, using ground-based observation?
nick
hehe, wazza ullage of my oxidizer? thanks
Lemme see if I got this straight: Compaq buys RAM from Micron, somehow certifies it (i.e. sticks a "Compaq" sticker on it, and runs a few diagnostics), and for this charges a 5x markup? And all for what? To amortize the cost of taking the blame away from Micron, even though Micron is still eventually the one to blame?
(well, it was at least 5x for their DEC workstations).
Hasn't LISP (and scheme) been doing byte-code compilation for about 30 years?
Ok, this is not just about companies just spinning off companies they've acquired (this post with Compaq+DEC, other posts about USR+3com).
Compaq acquired DEC, not for it's Alpha or research labs, but for it's high-end know-how, and for it's world-class NT support team.
3com acquired USR, not for it's modems, but for Palm!!!
Therefore, it makes sense for 3com and Compaq to spin off (or sell) the unwanted portions of those business they acquired for their (to them, at least) tastier aspects.
Well, both AT&T and Lucent do basic research, so at least on that point, your statement is false. :)
I've been curious about this for a while: does anyone know the relative amounts of radiation of a CRT versus an LCD (say at the same resolution)?
thanks!
nick
Will I be able to use other filesystems, such as FFS or ext2, jfs, etc, on OS-X? I guess what I'm asking is this: is any functionality of OS-X tied to features of HFS+ (e.g., does it depend on case insensitivity?).
thanks.
nick
drool.. I like your thinking, Aaron. My phd thesis is about a limited form of meta-reasoning for compilers. Whenever I see someone else who groks the power of "meta", I am happy. thanks!
The difference is that, in the case of movies-theathers, there is, in most cases, an abstraction between "seeing a movie", and "where I see it". THere are some exceptions, like Lucas originally thinking of restricting Phantom Menace to certain "blessed" theaters.
Whereas, with DVDs, "buying a DVD", and "how I view it" are intimately tied. The ones producing the DVDs are the same ones blessing the DVD players. nick
They are in a landfill in Utah, supposedly for tax write-off reasons, not cause of failing hard drives :)