Based on my experience, Norton's Personal Firewall is really good at breaking network configurations and needing to be uninstalled to get things working.
I seem to remember that the only computer system ever built on trinary (base-3) logic was produced in the Soviet Union. The name escapes me, but I think something like that is enought to dispell the idea of them not doing any original research (good research, OTOH...).
Even if those 'strict style rules' closely mimic accepted 'best practices' in code formatting?
Indenting code is a lot like punctuatiting English. Yeah, ee cummings was able to get some milage out of creative punctuation, but when most people do it they look retarded. The same goes for code; a few ppl can make their code better by using creative punctuation, but most ppl end up using the freedom to write unreadable, unmaintainable code. It's OK to write in a language that doesn't look like C.
Considering all things that Python does allow you to do, it's a fair trade off. 'Sides, your code never gets confused with line noise.
What is up with this 'Swing is slow' crap? I know Swing is _slower_ than native widgets and even AWT, but it's not *slow*. After the initial JVM load time, I always had more than acceptable performance on my PPro200/64MB, even when using complex Swing-based GUIs, and I highly doubt that any business undertaking a supposedly massive new project is going to be sitting around on anything that gutless.
WTF do you need a portable DVD-RW? All you should need to do on the road is read what you've burned.
Not to mention that you can't cheap your way out of an audio playback unit. Even using lossless compression, a cheap, noisy sound chip will become noticable on a good PA.
Now, here's something that's been bothering me... He said he's got a 500MB/s STR on the drives. Now, I'm not an expert on DB design, but I really doubt he needs 500MB/s of bandwidth to answer queries.
Well at least it's not like the guys at Software Carpentry who got a governement contract and held a 'coding contest' to see who could write their toolset for them.
www.sonicstate.com has some good stuff as well. There are a -lot- of places more well suited to finding out about this. As far as comming out with something usable for $1000, good luck. Music gear is sickly overpriced and depreciates slowly, so buying used doesn't even help -that- much.
While going with the 'leet Protools setup might be overboard (even though they have a basic 8-track version available for free) there are some things that you really can't skimp on. A solid multi-IO soundcard and a -good- MIDI interface are key components. So are good microphones and monitor speakers (trust me. Even the nicest PC speakers just don't compare to a cheap set of reference speakers for mixing). Just those components alone will set you back a grand, and that's before buying the computro.
While I never spent much time looking at the slides, Beazley's book "Python Essential Refference" is definately one of the best dead-tree computro books I've used. In a sense, it's comparable to K It's well written consise and always within reach. It's an indispensable reference guide (hence the name), but not a great book to learn from, as none of the examples actually do anything.
Unfortunately, keeping a programmer on staff that can perform Altivec optimizations for every problem thrown at the cluster will -quickly- eat away at any such bennefits for all but the largest clusters.
I can't entirely comprehend how it happened, but I've seen keyboards that not only have keys on the home-row wiped clean of markings, but they are rippled. There is a discernable ridge of plastic where one should not be.
The problem with looking at friction is that a lot of times you end up rubbing across keys that you aren't actually using.
While it might be a usefull metric for an individual to evaluate their personal use, it becomes meaningless if you want to compare two different users, as I've seen people who literally stab the KB with their fingers while typing.
(this turned into a rant mostly unrelated to the parent comment. sorry)
...or you can drop out of college and become the richest man in the world.
Big deal, you can't reasonably dole out advice based on a few one in a million people who, due
to luck or inate talent get ahead. I had a friend that dropped out in the 8th grade and by the time I started college was already making more money than my parents, but nobody can expect things like that to happen. It's like buying lotto tickets and expecting to win.
Every time these stories come up, there's always a string of "I said FTM, and left the oppressive accademic environment and now I'm working my dream job" but how do we know they wouldn't have done better had they stuck with their education? I'm sure those that dropped out, failed and regreted it aren't so quick to brag about their experiences.
The fact remains that universities haven't yet completely become trade schools. While your young and have no real responsibilities (house payment, family, etc) you should take the time and get the most from the experience. If your interests are in two completely different fields, go ahead and get a double major or multiple minors; you'll never have the freedom to do so again without sacrfices. (IE, the orignal poster, because of the overlap of the subjects, would have no problem majoring in EE & Physics with a CS minor (or you could just pick up an MS in math, making all those fields 'trivial applications) adding an extra 2yr tops to the total time in school). Or maybe an EE/Phyics degree then graduate CS program. If you've got what it takes to be an exception to formal learning, you should have no problem comming up with with a good plan.
Uhh... Why bother with a new signal when you can just write the program to save a checkpoint when recieving one of the normal ones? It's not like handling signals is -that- hard.
I hope it's better than the version of Metrowerks Codewarrior for Linux. The Linux version of Metroworks' product is nothing more than an IDE wrapping the standare GNU tools.
Bandwidth is a bit of a mute point here. If attached to a PCI bus, both Gb ether and firewire can pump more bits than the PCI can handle.
If you want to look at appropriate alternate/existing technologies for this kind of stuff, why not look into a hot swappable SCSI or fibre channel etc.
Anyone spending this kind of money on a cluster is likely to be doing serious number crunching. Now, some types of number crunching don't mind posibly letting a wrong answer slide through somewhere (IE - in a render-farm, you might get a bad pixel). Other environments, where calculations are all interdependant, a single bad number could make all the computation worthless (the old "A butterfly flaps its wings in Tokyo and we get another John Katz article" story).
Besides, no point in spending the time designing & testing this monster OCed box, when you can just buy a few 8-way Xeons, or call Sun/SGI.
I would have picked up an Apple notebook some time ago if ONLY it came with a second mouse button.
I know that MacOS apps don't need that 2nd button, but WTF are you supposed to do if you want to run Linux or run *nix apps on OSX? It's bad enough working with -only- two buttons, going to one button would destroy functionality.
Back in '95 a guy I knew decided to run a small ISP off a single T-1. He mostly did email & web hosting for small busineses, so bandwidth was never a problem for him.
That is until he pissed off the wrong script kiddie on IRC. The irate kiddie made sure that all of his bandwidth was saturated for 2wk straight. It wasn't a true DDOS in the sense that hundreds of computros simultainiously attack the target but the attack was comming from multiple sites with seriously fat pipes, making blocking the attack imposible.
His customers got fed up with the 2wk loss of service and left for greener pastures causing the business to fold.
Go beyond... Think about it. They could run the angle that an AOL subscription will get you -all- the software on your computer both legally AND free. Not to mention that the public would be more receptive to a subscription-based-service becoming an OS/software supplier than a OS/Software supplier trying to become a subscription-based-service. And since the software is still free, there's no 'lost profits' from people 'stealing' your software.
Slashdot Mirror
Of course... 'cuz we know that completely uneducated ppl in the university are the ones that will be using the cluster.
Based on my experience, Norton's Personal Firewall is really good at breaking network configurations and needing to be uninstalled to get things working.
I seem to remember that the only computer system ever built on trinary (base-3) logic was produced in the Soviet Union. The name escapes me, but I think something like that is enought to dispell the idea of them not doing any original research (good research, OTOH...).
Even if those 'strict style rules' closely mimic accepted 'best practices' in code formatting?
Indenting code is a lot like punctuatiting English. Yeah, ee cummings was able to get some milage out of creative punctuation, but when most people do it they look retarded. The same goes for code; a few ppl can make their code better by using creative punctuation, but most ppl end up using the freedom to write unreadable, unmaintainable code. It's OK to write in a language that doesn't look like C.
Considering all things that Python does allow you to do, it's a fair trade off. 'Sides, your code never gets confused with line noise.
What is up with this 'Swing is slow' crap? I know Swing is _slower_ than native widgets and even AWT, but it's not *slow*. After the initial JVM load time, I always had more than acceptable performance on my PPro200/64MB, even when using complex Swing-based GUIs, and I highly doubt that any business undertaking a supposedly massive new project is going to be sitting around on anything that gutless.
WTF do you need a portable DVD-RW? All you should need to do on the road is read what you've burned.
Not to mention that you can't cheap your way out of an audio playback unit. Even using lossless compression, a cheap, noisy sound chip will become noticable on a good PA.
Now, here's something that's been bothering me... He said he's got a 500MB/s STR on the drives. Now, I'm not an expert on DB design, but I really doubt he needs 500MB/s of bandwidth to answer queries.
Well at least it's not like the guys at Software Carpentry who got a governement contract and held a 'coding contest' to see who could write their toolset for them.
What do humans know about Klingons? Well... we -did- invent them.
www.sonicstate.com has some good stuff as well. There are a -lot- of places more well suited to finding out about this. As far as comming out with something usable for $1000, good luck. Music gear is sickly overpriced and depreciates slowly, so buying used doesn't even help -that- much.
While going with the 'leet Protools setup might be overboard (even though they have a basic 8-track version available for free) there are some things that you really can't skimp on. A solid multi-IO soundcard and a -good- MIDI interface are key components. So are good microphones and monitor speakers (trust me. Even the nicest PC speakers just don't compare to a cheap set of reference speakers for mixing). Just those components alone will set you back a grand, and that's before buying the computro.
While I never spent much time looking at the slides, Beazley's book "Python Essential Refference" is definately one of the best dead-tree computro books I've used. In a sense, it's comparable to K It's well written consise and always within reach. It's an indispensable reference guide (hence the name), but not a great book to learn from, as none of the examples actually do anything.
Looking at the website, the first thing I noticed was how much their logo resembles the Sega Dreamcast logo.
Indeed, an ill omen.
They also have official clients (both the Java and Linux ones) that don't have ads.
Unfortunately, keeping a programmer on staff that can perform Altivec optimizations for every problem thrown at the cluster will -quickly- eat away at any such bennefits for all but the largest clusters.
I can't entirely comprehend how it happened, but I've seen keyboards that not only have keys on the home-row wiped clean of markings, but they are rippled. There is a discernable ridge of plastic where one should not be.
The problem with looking at friction is that a lot of times you end up rubbing across keys that you aren't actually using.
While it might be a usefull metric for an individual to evaluate their personal use, it becomes meaningless if you want to compare two different users, as I've seen people who literally stab the KB with their fingers while typing.
(this turned into a rant mostly unrelated to the parent comment. sorry)
...or you can drop out of college and become the richest man in the world.
Big deal, you can't reasonably dole out advice based on a few one in a million people who, due
to luck or inate talent get ahead. I had a friend that dropped out in the 8th grade and by the time I started college was already making more money than my parents, but nobody can expect things like that to happen. It's like buying lotto tickets and expecting to win.
Every time these stories come up, there's always a string of "I said FTM, and left the oppressive accademic environment and now I'm working my dream job" but how do we know they wouldn't have done better had they stuck with their education? I'm sure those that dropped out, failed and regreted it aren't so quick to brag about their experiences.
The fact remains that universities haven't yet completely become trade schools. While your young and have no real responsibilities (house payment, family, etc) you should take the time and get the most from the experience. If your interests are in two completely different fields, go ahead and get a double major or multiple minors; you'll never have the freedom to do so again without sacrfices. (IE, the orignal poster, because of the overlap of the subjects, would have no problem majoring in EE & Physics with a CS minor (or you could just pick up an MS in math, making all those fields 'trivial applications) adding an extra 2yr tops to the total time in school). Or maybe an EE/Phyics degree then graduate CS program. If you've got what it takes to be an exception to formal learning, you should have no problem comming up with with a good plan.
Uhh... Why bother with a new signal when you can just write the program to save a checkpoint when recieving one of the normal ones? It's not like handling signals is -that- hard.
I hope it's better than the version of Metrowerks Codewarrior for Linux. The Linux version of Metroworks' product is nothing more than an IDE wrapping the standare GNU tools.
I expect better from Borland.
Bandwidth is a bit of a mute point here. If attached to a PCI bus, both Gb ether and firewire can pump more bits than the PCI can handle.
If you want to look at appropriate alternate/existing technologies for this kind of stuff, why not look into a hot swappable SCSI or fibre channel etc.
Granted, most laptop pointing devices suck, but still, I don't want a machine that -needs- an external mouse to work correctly.
One catch :
Anyone spending this kind of money on a cluster is likely to be doing serious number crunching. Now, some types of number crunching don't mind posibly letting a wrong answer slide through somewhere (IE - in a render-farm, you might get a bad pixel). Other environments, where calculations are all interdependant, a single bad number could make all the computation worthless (the old "A butterfly flaps its wings in Tokyo and we get another John Katz article" story).
Besides, no point in spending the time designing & testing this monster OCed box, when you can just buy a few 8-way Xeons, or call Sun/SGI.
I would have picked up an Apple notebook some time ago if ONLY it came with a second mouse button.
I know that MacOS apps don't need that 2nd button, but WTF are you supposed to do if you want to run Linux or run *nix apps on OSX? It's bad enough working with -only- two buttons, going to one button would destroy functionality.
Back in '95 a guy I knew decided to run a small ISP off a single T-1. He mostly did email & web hosting for small busineses, so bandwidth was never a problem for him.
That is until he pissed off the wrong script kiddie on IRC. The irate kiddie made sure that all of his bandwidth was saturated for 2wk straight. It wasn't a true DDOS in the sense that hundreds of computros simultainiously attack the target but the attack was comming from multiple sites with seriously fat pipes, making blocking the attack imposible.
His customers got fed up with the 2wk loss of service and left for greener pastures causing the business to fold.
Go beyond... Think about it. They could run the angle that an AOL subscription will get you -all- the software on your computer both legally AND free. Not to mention that the public would be more receptive to a subscription-based-service becoming an OS/software supplier than a OS/Software supplier trying to become a subscription-based-service. And since the software is still free, there's no 'lost profits' from people 'stealing' your software.
Download.com already autodetects your OS from the browser info string, and sends you to an appropriate webzit. Mac, Windows and Linux.