I disagree. Refactoring sounds like a much better solution. And I highly doubt this is a serious bug that affects the whole design - why would the introduction of a new CPU flag require the complete reconstruction of an entire OS kernel? Or more?
I don't see how you could possibly suggest a full redesign and rewrite and then, in the same post, complain that fast fixes are rarely 100% correct. As if the rewrite won't be a thousand times worse!
Taking your analogy to its extreme, it seems like you're suggesting rewriting the entire piece of software every time a bug is found.
I think I'd prefer a different analogy.
It's much harder to get through a barrier that's been tested and reinforced appropriately than one that's never been tried out - no matter what the skill is of the person who constructed it.
Software has bugs. Software gets fixed. Software is stronger for it. This is true of Windows, Linux, OpenSSL, and any other piece of software you care to name. Microsoft is fully aware that rewriting Windows would be a ludicrously stupid move on all fronts, and so they're not going to do it - and they shouldn't, either.
I've worked at precisely one game company so far. The hours were flexible, the co-workers were dedicated and brilliant, and I worked approximately 40 hours a week every week.
I realize this isn't much of a sample. However, at least part of the industry is much better than EA.
(I left to go back to college, btw, which turned out to be a mistake. Right now I'm working at Google, and soon I'll be working towards founding my own game studio.)
Google. Yes, of course we'd love to hire people who are geniuses in their field and hard-working. I have no doubt that we'd also hire people who are extremely experienced.
(Yes, I work at Google.)
Don't judge a company just by the hiring group's ad copy - saying "we hire experienced people" would kick out *all* the college graduates, whereas saying "we hire supremely competent people" should include everyone.:)
It's possible the answer to the question is something completely unrelated to the actual question. All of my answers are like this - I don't want someone to be able to get into my account with something as similar as my mother's maiden name (especially since that's the name she uses right now!) so I always type something utterly unrelated.
Usually I also write down what it is, or use one of my standard responses.:P
Yes, that's entirely true. On the other hand, they will have cost you half as much as the equivalent capacity in high-end systems, they won't have any downtime, and you won't need a full time employee just to pick up the pieces when they collapse.
And so what if it's different home built hardware? Even ignoring the fact that there's no reason to make them different, it flat-out doesn't matter - PCs have the convenient property that you can swap out hardware with different specs and it keeps working.
Seriously. I don't care how good of a system you buy, someday something *will* go wrong and it'll go bottom up. If that's the only server you bought you're now down for the count.
On the other hand, if you bought three much cheaper commodity servers, then even if two of them go down you can probably still keep *something* going. Same basis as RAID.
Anybody who makes the assumption that good quality components means they won't melt down is setting themselves up for disappointment - and if they're lucky, it won't lead to severe financial problems.
Redundancy is king.
(Of course, in some cases, it's not practical - but I'd always choose it over individual part quality if possible. And as your scale goes up, it gets more and more practical.)
I didn't say "Baldrson claims X, therefore X is not true" - I said "I'd want more proof". For every example of a person people were skeptical of that turned out to be right I can give you a dozen examples where the opposite is true. Whichever part you want to make the opposite.
I went and looked at that link. To be honest, I agree with them. If the letter is genuine it should exist other places than your website, and judging from what else is on your website, I wouldn't necessarily trust you if you said 2+2=4.
Sorry.
I'd ask if you can prove it's from the person you claim it is - to be honest I haven't been able to look at it though (your geocities site is out of bandwidth) and so maybe it's a lot more genuine than I'm imagining it. But in any case, I'd still look for a more authoritative site that corroborates it.
Otherwise we might as well start worshiping Gene Ray as a prophet.
The problem with that is light speed. Transmitting a lightspeed signal across one centimeter takes about 3.3*10^-11 seconds - which sounds like a lot, until you realize that a single CPU cycle now takes about 3.3*10^-10 seconds. And I don't even know if electricity travels at true lightspeed or at something below that.
Another problem, of course, is heat - if your 1cm^2 CPU outputs 100w of heat, a 10cm^2 CPU is going to dump 1000w of heat. That's a hell of a lot of heat.
A third problem is reliability. Yields are bad enough with the current core sizes, tripling the core sizes will drop yield even further.
And a fourth problem is what exactly to *do* with the extra space.:) Yes, you could just fill it with cache, but that still won't give you a computer twice as fast for every twice as much cache - MHz has nothing to do with how many transistors you can pile on a chip. (Of course, you could just put a second CPU on the same chip . ..)
Same here. IHOP wouldn't let us plug in a laptop, Denny's did. Denny's now gets my business whenever I'm wandering about with a laptop looking for something to eat.
I hope the fraction of a cent IHOP saved was worth it to them.
They've tried to crack single encrypted messages that exist for the sole purpose of being cracked. I'm sorry, but that just plain doesn't excite me anymore.
OGR's more interesting, but I got bored of RC5 years ago.
Tor is great. I've been playing with it for a while - the sheer simplicity of setup makes it fantastic, and it's highly amusing to go to whatismyip.com half a dozen times and get different IPs.
Once I get the firewall box I want set up I plan to make one port link directly into Tor, so that anything plugged into that port is shunted 100% into the Tor network. Right now you've sort of got to trust that your program really is punching everything through the SOCKS proxy - not all programs are really reliable about that, plus the program can still see your IP if you're not behind a firewall.
Depends what you plan to do with it. Personally, I'm planning to shift to watercooling with my next computer - not so I can overclock, but so I can quiet my computer down.
(Speed is important enough that I want to use high-end CPUs, also, so I can't just "switch to mini-ITX" or "buy a slow CPU".)
What he was doing wasn't just storing movies. It was storing movies in a place where others could download them. Having movies gives you points, but having movies on a really big internet connection and letting lots of other people download them gives you major points. Especially if you set it up as a trade system, so you *get* lots of movies and such at the same time.
NASA (like most companies) has lots of space on a big internet connection. There ya go.
It's that the posted link, to the article that this is a dupe of, is a link into the admin interface. For the curious, right now it's https://slashdot.org/admin.pl?op=edit&sid=04/12/15/1936218 - I imagine this will be changed once the admins notice . . . well, probably.
is actually rather solvable, especially in this situation.
Most people decide to use Euclidean distance, or distance-squared. It's possible to do some statistical tests comparing it to Manhattan distance, or distance-added, and you end up with Manhattan distance often being a "better" indication. So why not exaggerate?
Take the general formula d=sum(abs(x_n^v),n=1..nmax)^(1/v). Euclidean distance is this formula with v=2, Manhattan distance with v=1. Lower v below 1 - 0.5, 0.3, or lower - and you get a distance metric that works quite well with high numbers of dimensions.
Meanwhile, back in reality, the meaning of this distance metric is something along the lines of "it's okay if there are a few major differences, as long as mostly we're a good match", as opposed to "avoid major differences at all costs" . . . so instead of getting someone who's marginally different from you in all ways, you get someone who's very similar to you except for one major difference.
Which can be interesting.
Sometimes, the kind of "interesting" that involves handcuffs . . . either in the good way or the bad way.
I don't know if any online dating sites do this or not. But they should.
Slashdot Mirror
I disagree. Refactoring sounds like a much better solution. And I highly doubt this is a serious bug that affects the whole design - why would the introduction of a new CPU flag require the complete reconstruction of an entire OS kernel? Or more?
I don't see how you could possibly suggest a full redesign and rewrite and then, in the same post, complain that fast fixes are rarely 100% correct. As if the rewrite won't be a thousand times worse!
Taking your analogy to its extreme, it seems like you're suggesting rewriting the entire piece of software every time a bug is found.
I think I'd prefer a different analogy.
It's much harder to get through a barrier that's been tested and reinforced appropriately than one that's never been tried out - no matter what the skill is of the person who constructed it.
Software has bugs. Software gets fixed. Software is stronger for it. This is true of Windows, Linux, OpenSSL, and any other piece of software you care to name. Microsoft is fully aware that rewriting Windows would be a ludicrously stupid move on all fronts, and so they're not going to do it - and they shouldn't, either.
I've worked at precisely one game company so far. The hours were flexible, the co-workers were dedicated and brilliant, and I worked approximately 40 hours a week every week.
I realize this isn't much of a sample. However, at least part of the industry is much better than EA.
(I left to go back to college, btw, which turned out to be a mistake. Right now I'm working at Google, and soon I'll be working towards founding my own game studio.)
It would have been even better to pick a nearby town, then do all your hacking in a 10-mile circle from that town, somewhat evenly distributed. :)
Google. Yes, of course we'd love to hire people who are geniuses in their field and hard-working. I have no doubt that we'd also hire people who are extremely experienced.
:)
(Yes, I work at Google.)
Don't judge a company just by the hiring group's ad copy - saying "we hire experienced people" would kick out *all* the college graduates, whereas saying "we hire supremely competent people" should include everyone.
Personally, I don't even write it up - I just start playing with something. Once it's cool enough to be useful, that's the time to write it up. :)
.)
(Which has the added side effect that nobody knows about the ideas that turned out to be completely useless . .
Oh, yeah. That'd be great. Instead of having to squeeze the public services I want behind a single IP, I'd just be screwed. That's a real step up.
Encouraged? Sure. Forced? I like having my open static IP, thanks.
It's possible the answer to the question is something completely unrelated to the actual question. All of my answers are like this - I don't want someone to be able to get into my account with something as similar as my mother's maiden name (especially since that's the name she uses right now!) so I always type something utterly unrelated.
:P
Usually I also write down what it is, or use one of my standard responses.
Then you find a fix for it, or acknowledge that you made a terrible hardware decision. :P
In theory, you're running a system where "crashes" doesn't always imply "hard locks". In that case you just ssh in and reboot.
Better yet, you whip up a quick watchdog program to autoreboot it.
Better yet, you either fix it or junk it - even worse is a computer that subtly corrupts data.
Yes, that's entirely true. On the other hand, they will have cost you half as much as the equivalent capacity in high-end systems, they won't have any downtime, and you won't need a full time employee just to pick up the pieces when they collapse.
And so what if it's different home built hardware? Even ignoring the fact that there's no reason to make them different, it flat-out doesn't matter - PCs have the convenient property that you can swap out hardware with different specs and it keeps working.
. . . build lots.
Seriously. I don't care how good of a system you buy, someday something *will* go wrong and it'll go bottom up. If that's the only server you bought you're now down for the count.
On the other hand, if you bought three much cheaper commodity servers, then even if two of them go down you can probably still keep *something* going. Same basis as RAID.
Anybody who makes the assumption that good quality components means they won't melt down is setting themselves up for disappointment - and if they're lucky, it won't lead to severe financial problems.
Redundancy is king.
(Of course, in some cases, it's not practical - but I'd always choose it over individual part quality if possible. And as your scale goes up, it gets more and more practical.)
I didn't say "Baldrson claims X, therefore X is not true" - I said "I'd want more proof". For every example of a person people were skeptical of that turned out to be right I can give you a dozen examples where the opposite is true. Whichever part you want to make the opposite.
I went and looked at that link. To be honest, I agree with them. If the letter is genuine it should exist other places than your website, and judging from what else is on your website, I wouldn't necessarily trust you if you said 2+2=4.
Sorry.
I'd ask if you can prove it's from the person you claim it is - to be honest I haven't been able to look at it though (your geocities site is out of bandwidth) and so maybe it's a lot more genuine than I'm imagining it. But in any case, I'd still look for a more authoritative site that corroborates it.
Otherwise we might as well start worshiping Gene Ray as a prophet.
I don't buy it. Look at AMD, they've been pushing innovation better than Intel has lately.
The *real* problem is humanity's love of backwards compatibility. And that one you're not going to get rid of, ever.
Thanks for a good answer to this on a subject I wasn't sure of - I'd mod you up if I could. :)
I'd entirely forgotten about the capacitance issue.
The problem with that is light speed. Transmitting a lightspeed signal across one centimeter takes about 3.3*10^-11 seconds - which sounds like a lot, until you realize that a single CPU cycle now takes about 3.3*10^-10 seconds. And I don't even know if electricity travels at true lightspeed or at something below that.
:) Yes, you could just fill it with cache, but that still won't give you a computer twice as fast for every twice as much cache - MHz has nothing to do with how many transistors you can pile on a chip. (Of course, you could just put a second CPU on the same chip . . .)
Another problem, of course, is heat - if your 1cm^2 CPU outputs 100w of heat, a 10cm^2 CPU is going to dump 1000w of heat. That's a hell of a lot of heat.
A third problem is reliability. Yields are bad enough with the current core sizes, tripling the core sizes will drop yield even further.
And a fourth problem is what exactly to *do* with the extra space.
Same here. IHOP wouldn't let us plug in a laptop, Denny's did. Denny's now gets my business whenever I'm wandering about with a laptop looking for something to eat.
I hope the fraction of a cent IHOP saved was worth it to them.
"Far more impressive work"?
They've tried to crack single encrypted messages that exist for the sole purpose of being cracked. I'm sorry, but that just plain doesn't excite me anymore.
OGR's more interesting, but I got bored of RC5 years ago.
Tor is great. I've been playing with it for a while - the sheer simplicity of setup makes it fantastic, and it's highly amusing to go to whatismyip.com half a dozen times and get different IPs.
Once I get the firewall box I want set up I plan to make one port link directly into Tor, so that anything plugged into that port is shunted 100% into the Tor network. Right now you've sort of got to trust that your program really is punching everything through the SOCKS proxy - not all programs are really reliable about that, plus the program can still see your IP if you're not behind a firewall.
Depends what you plan to do with it. Personally, I'm planning to shift to watercooling with my next computer - not so I can overclock, but so I can quiet my computer down.
(Speed is important enough that I want to use high-end CPUs, also, so I can't just "switch to mini-ITX" or "buy a slow CPU".)
No, it makes sense.
What he was doing wasn't just storing movies. It was storing movies in a place where others could download them. Having movies gives you points, but having movies on a really big internet connection and letting lots of other people download them gives you major points. Especially if you set it up as a trade system, so you *get* lots of movies and such at the same time.
NASA (like most companies) has lots of space on a big internet connection. There ya go.
isn't just the fact that it's a dupe.
5 /1936218 - I imagine this will be changed once the admins notice . . . well, probably.
It's that the posted link, to the article that this is a dupe of, is a link into the admin interface. For the curious, right now it's https://slashdot.org/admin.pl?op=edit&sid=04/12/1
is actually rather solvable, especially in this situation.
Most people decide to use Euclidean distance, or distance-squared. It's possible to do some statistical tests comparing it to Manhattan distance, or distance-added, and you end up with Manhattan distance often being a "better" indication. So why not exaggerate?
Take the general formula d=sum(abs(x_n^v),n=1..nmax)^(1/v). Euclidean distance is this formula with v=2, Manhattan distance with v=1. Lower v below 1 - 0.5, 0.3, or lower - and you get a distance metric that works quite well with high numbers of dimensions.
Meanwhile, back in reality, the meaning of this distance metric is something along the lines of "it's okay if there are a few major differences, as long as mostly we're a good match", as opposed to "avoid major differences at all costs" . . . so instead of getting someone who's marginally different from you in all ways, you get someone who's very similar to you except for one major difference.
Which can be interesting.
Sometimes, the kind of "interesting" that involves handcuffs . . . either in the good way or the bad way.
I don't know if any online dating sites do this or not. But they should.
(For the curious: On the surprising behavior of distance metrics in high dimensional space)
Well, read the article - apparently they were actually in the house, and ran off, and yes, later got arrested.