You know it's time to get some sleep when you read that `AOL porno discs'. I thought perhaps that was one of the things AOL sent people when they sign up so that when they get busy signals they can use the CD instead...
I washed my keyboard this summer. It was kind of grimey, so I unplugged it (the computer was still up), took off the keys, unscrewed the bottom, and carefully washed everything -- even the circuit board-like thing. I've done similar things to network cards, hard drives, even motherboards. I don't know what the big deal is with electronics and water. I would like to some day set up an old computer, and completely submerge it while it's running. I suppose that would be a bad idea for someone with hard water, but there aren't many minerals in the water where I live.
Well, I'm at an american college. When I cleaned my keyboard this summer, I found mostly hair (I shed alot; I suppose I must be going bald), some kind of grease (perhaps from coffee?), and the occasional crumb. It really wasn't too dirty inside; it was the tops of the keys that needed the cleaning. The grime on the keys was a fairly good indication of how often they were pressed. The backspace was among the grimer amusingly enough. Anyway, no beer at all (but I'll accept the coffee proposal).
I've had keyboard problems where I had to reboot using only the mouse in X. Not the easiest thing to do, especially when the only program I can call with the mouse is an xterm. A bit of left click/middle click, and I had enough letters to get a 'shutdown -h now', but I think I should have just saved myself the trouble and just pressed the reset button...
...it's really stupid to use a program which is not a programming langauge since you loose the full power of the computer.
I just noticed that I'm tending toward that ideal of only using compilers for doing productive things. I use TeX for making reports (it compiles to the `machine code' of a postscript printer). I use Fortran instead of a spread sheet. Bash for playing with the OS and scripting, metapost for making vector graphics, and the gimp, which can use scheme and perl for scripting, for bitmap graphics. Now I just need to find a slashdot-reading programing language...
It looks like this would be about comparable to your average O2000 SGI. These systems go for a lot more than 35k. The problem is that research grants take much longer than 5 days to go through, and even if 100k was allocated for an SGI, the department can't just change the wording of the grant. Also, research departments like to get thier hardware from official sources. Getting something on Ebay would be considered a gamble, and one doesn't gamble with 35k of thier research group's money. I don't think ebay is the right medium to sell this system. It will probably be bought privately for much less than it's worth.
That means only 8 G4s to equal the computer power of the Cray
Apples and Oranges (sorry about the pun). The bandwidth, bus-speed, memory size/type, etc. of the Crey will destroy any workstation you could throw at it. If all you're doing is SETI, go for the G4's. If you are doing scientific high powered computing, there is no comparison.
And the graphics from linux.com are much better looking that the other graphics on his site. They work nicely with the color scheme of the layout, but the graphics the kid got elsewhere clash (and is it just me, or is that dancing penguin really annoying and overused?).
I had an account on some linux machines at my college. They came up with gnome because the proff just did a default install and didn't change much of anything. I make a quick little script that I ran after I logged in: kill `ps ax | grep gnome | awk '{print $1}'`
fvwm &
This normally did the job of getting a non-windows/non-user-friendy environment going so that I could get some real work done.
BTW, I just reinstalled one of my systems, and made the big jump from fvwm to fvmw2. Of course, I configured it exactly the same (well, the pager is a little different, and I switched to rclock). All that gnome and KDE stuff is too bloated for me. I would rather cache libraries with my memory than display eye-candy.
I don't know how you drink your coffee, but I bring water to a boil before adding coffee grounds and decanting. I even pour some of the boiling water into my coffee cup and the decanter first to warm them up. You like your coffee luke warm? I once took a sauna where the temperature read 210F (I was sitting on the top bench too). I didn't die or anything (although when I jumped in the snow bank afterwards, I'm sure my heart missed a few beats). If I spilled 180F coffee on myself, I would say `ouch' and call myself an idiot. Now I'm going to go get a lab thermometer and see how hot exactly my coffee is when I drink it. I bet it's more than 180F.
You're result is that 36=36, and this proof requires that the hypothesis is true to be valid (yes, I am with the math police). Furthermore, mutating equalities until you get a reult is poor proof technique, and lead to things like multiplying both sides by zero to get 0=0. Here's my attempt at a more correct version (the math is in TeX for all you computer geeks; go ahead, render it...):
Define: $$ 0.999\dots =\sum_{i=1}^\infty 9*10^{-i} $$ Corollory
This is a geometric series, so we get: $$ 0.999\dots =\sum_{i=1}^\infty 9*10^{-i}={9\over 1-10}={9\over 9}=1 $$ Proof
$$ 3.999\dots =3+0.999\dots=3+1=4 $$ Quod Erat Demonstrandum
The only problem is that black holes are relatively rare. Take the number of binary star systems compared with the number of stars. Compare that ratio with the number of black holes and you'll see how likely a binary black hole system would be to occur. Black holes can be regarded as the same thing as stars only that they have so much mass that the escape velocity from their surface is greater than the speed of light. Some black holes also create a singularity, but I believe most are just big stars, and can be treated as such.
Use a simple vector format like metapost which should be fairly easy to decypher. A gif would probably be dificult to figure out, and a jpg nearly impossable. For bitmaps, pnm is rather simple (asci art would probably be meaningless because they wouldn't know what a 00100101 would look like).
I am sitting happily right now on my internet2 connection. It's nice downloading from metalab.unc.edu 500KB/s. What I really like is that I've been doing research with some supercomputers over in another university. I can send the X programs over to my display with almost no latency. I can even rotate molecules and do complex visualizations. I don't even keep a computer in my apartment because I would never be able to stand the dial-up connection. Sorry for bragging...
Interesting that you tell a bunch of computer geeks that one of their favorite star wars characters will be computer generated, and they all get angry. Don't we like computers? Perhaps we have a human side to us after all, and realize that for some things computers are not the answer. Could it be that was have an artistic side after all? (okay, programming is an art form, but...). Just some food for thought.
So what would happen if I decided to set up a DNS server, alias as many websites I can find to.smart and.stupid domain names (unofficial, of course), and suggest that people add my DNS server to thier resolve.conf file. Would this be illeagal? It would sure be fun (although lots of work). I wonder what sort of powers exist that keep underground domain name servers from sprining up.
I've successfully gotten through college without purchasing too many text books. In the field of mathematics, the textbooks are extremely small, and so I would just photocopy the 200 pages (reducing slightly to fit two pages on a sheet), and use that. 100 x 5cents = $5. These books generally go for $80-$100 each.
I would also check books out from the library. I checked out a physical chemistry textbook for p-chem, chemical energetics, spectroscopy, and a year of research. I saved over a hundred dollars (and somehow avoided library fines). I've never had any literature courses, but one could be sure that I would get those books from the Gutenberg Project if not from the library.
Of course, there were even a few courses that I just plain went without a book in. I would pay attention during the lecture (well...), and somehow would get by without failing. Homework sets were a problem in some classes, but I rarely turned in homework anyway. I purchased my Organic Chem textbook my freshman year. After realizing by the end of the two semesters that I had never read it, I decided that I would make sure I need a text book before purchasing one.
I suppose since I'm working on a doctorate and want to stay in education that I will at some point end up being a professor. I vow now that I will never ask any student to purchase one of these digital books. I will teach my courses in such a manner that a student could get by with any general textbook in the subject (hence be able to use the library), and make any problem sets available on the course web site (as well as in printouts for those who don't like computers). Education should not discriminate against people who don't have limitless financial resources (or cheap people like me).
Not quite as easy as that for me. Galleon doesn't have an rpm for alpha linux, so I have to compile it on my own. Not too bad until one realizes that it requires header files from mozilla, so the source for mozilla has to be downloaded (which is over 200MB unpacked). Now I see that all the header files aren't in the uncompiled source, so it looks like I'm going to have to compile mozilla. If the source is over 200MB, it's probably going to need several GB's to compile. So now I have to make room on a partition, and let it compile all day (it probably won't compile right either given my luck). Glad it's easy for some people, though.
Another big plus to having lots of little SCSI drives is that if one goes bad, it can be replaced with only a few gigs of data lost (unless if you keep backups on different drives like I do). If your 80GB IDE drive goes bad, you've got a major problem. If it's just bad sectors, you can partition in so that the bad cylender isn't being used (I've done it, and it's a pain), but if it stops spinning up, that's a major investment gone. I've noticed that SCSI drives tend to be better made, and I've never seen bad sectors show up (which means they probably allocate more space to fixing them). I still use my IDE drives to store mp3's, but 4GB SCSI3 is plenty for me to play with.
Sure, methanol is poisonous, but so is gasoline. People probably don't drink gasoline because it smells bad. I beleive that this smell is added to prevent accidental drinking (at least I know it is for methane). Ethanol would be safer (our bodies don't convert it to fomaldehyde and formic acid as they do with methanol), but the temptation to drink it would probably be too great for some people. Regardless, ethanol is corrently used to cut down gasoline in the corn belt, and can be used at about 30% or so without modifications to the engine (it runs cleaner and soaks up water in the gas tank too). The reason why we aren't switching fuels is because the oil industry is too big and powerful to let that happen (conspiracy!).
HCl isn't that bad, particularly when only the skin is exposed. H2SO4 can be nasty, and will turn a person's skin black. This is not due to the acidity, but rather affinity for water. If it is being rinsed with water, this shouldn't be a problem. A drop of thionyl chloride thus isn't too much of a danger. If it isn't rinsed off quickly, it will react with the water in one's skin to make those lovely acids, which would destroy lipids, etc. If one's eyes got splashed with thionyl chloride, it would react with the water there, giving off lots of heat, and blinding acid (much the same as tear gas only worse). Instead of quickly walking to the eyewash station to rinse, the victim would most likely claw thier eyes out because they would hurt so much. That's why one never works in a lab alone...
Well, you may be right (I don't know C++, so I can't in good faith argue), but I am doing research in quantum chemistry using lots of computational tools. Everything is written in fortran. Take a look at virtually any piece of software written for this subject and you'll see that it's written in fortran. Sometimes they'll use C for a user interface (if there is one), but all the major code is in fortran. Let's say I want to multiply two matricies. How do I do that in C++? Compare f90's way of doing that, and you'll see why fortran is still used so much. I've run a few C++ programs, and have noticed that they always tend to be a bit sluggish. Fortran is written with one view in mind: to crunch numbers. C++ has other interests. There are specialized languages for math, but I've never seen anyone use them. There is a standard in the sciences (math, chemistry, physics, geology, etc. (ok, not CS)), and it's fortran. How well C++ can be used for MPI and PVM? Perhaps just fine (I don't know), but it's quite easy in fortran. I will admit that fortran isn't always portable (I've had troubles with g77), but if a good compiler is used (sorry g77 developers) like SGI's, then things compile and run without a problem. I don't want to get involved in any compiler wars, but will have to go along with fortran as the best for a scientist to learn and use.
Slashdot Mirror
I don't know what you're talking about; it works fine in lynx...
You know it's time to get some sleep when you read that `AOL porno discs'. I thought perhaps that was one of the things AOL sent people when they sign up so that when they get busy signals they can use the CD instead...
I washed my keyboard this summer. It was kind of grimey, so I unplugged it (the computer was still up), took off the keys, unscrewed the bottom, and carefully washed everything -- even the circuit board-like thing. I've done similar things to network cards, hard drives, even motherboards. I don't know what the big deal is with electronics and water. I would like to some day set up an old computer, and completely submerge it while it's running. I suppose that would be a bad idea for someone with hard water, but there aren't many minerals in the water where I live.
Well, I'm at an american college. When I cleaned my keyboard this summer, I found mostly hair (I shed alot; I suppose I must be going bald), some kind of grease (perhaps from coffee?), and the occasional crumb. It really wasn't too dirty inside; it was the tops of the keys that needed the cleaning. The grime on the keys was a fairly good indication of how often they were pressed. The backspace was among the grimer amusingly enough. Anyway, no beer at all (but I'll accept the coffee proposal).
I've had keyboard problems where I had to reboot using only the mouse in X. Not the easiest thing to do, especially when the only program I can call with the mouse is an xterm. A bit of left click/middle click, and I had enough letters to get a 'shutdown -h now', but I think I should have just saved myself the trouble and just pressed the reset button...
I just noticed that I'm tending toward that ideal of only using compilers for doing productive things. I use TeX for making reports (it compiles to the `machine code' of a postscript printer). I use Fortran instead of a spread sheet. Bash for playing with the OS and scripting, metapost for making vector graphics, and the gimp, which can use scheme and perl for scripting, for bitmap graphics. Now I just need to find a slashdot-reading programing language...
It looks like this would be about comparable to your average O2000 SGI. These systems go for a lot more than 35k. The problem is that research grants take much longer than 5 days to go through, and even if 100k was allocated for an SGI, the department can't just change the wording of the grant. Also, research departments like to get thier hardware from official sources. Getting something on Ebay would be considered a gamble, and one doesn't gamble with 35k of thier research group's money. I don't think ebay is the right medium to sell this system. It will probably be bought privately for much less than it's worth.
That means only 8 G4s to equal the computer power of the Cray
Apples and Oranges (sorry about the pun). The bandwidth, bus-speed, memory size/type, etc. of the Crey will destroy any workstation you could throw at it. If all you're doing is SETI, go for the G4's. If you are doing scientific high powered computing, there is no comparison.
And the graphics from linux.com are much better looking that the other graphics on his site. They work nicely with the color scheme of the layout, but the graphics the kid got elsewhere clash (and is it just me, or is that dancing penguin really annoying and overused?).
I had an account on some linux machines at my college. They came up with gnome because the proff just did a default install and didn't change much of anything. I make a quick little script that I ran after I logged in:
kill `ps ax | grep gnome | awk '{print $1}'`
fvwm &
This normally did the job of getting a non-windows/non-user-friendy environment going so that I could get some real work done.
BTW, I just reinstalled one of my systems, and made the big jump from fvwm to fvmw2. Of course, I configured it exactly the same (well, the pager is a little different, and I switched to rclock). All that gnome and KDE stuff is too bloated for me. I would rather cache libraries with my memory than display eye-candy.
I don't know how you drink your coffee, but I bring water to a boil before adding coffee grounds and decanting. I even pour some of the boiling water into my coffee cup and the decanter first to warm them up. You like your coffee luke warm? I once took a sauna where the temperature read 210F (I was sitting on the top bench too). I didn't die or anything (although when I jumped in the snow bank afterwards, I'm sure my heart missed a few beats). If I spilled 180F coffee on myself, I would say `ouch' and call myself an idiot. Now I'm going to go get a lab thermometer and see how hot exactly my coffee is when I drink it. I bet it's more than 180F.
You're result is that 36=36, and this proof requires that the hypothesis is true to be valid (yes, I am with the math police). Furthermore, mutating equalities until you get a reult is poor proof technique, and lead to things like multiplying both sides by zero to get 0=0. Here's my attempt at a more correct version (the math is in TeX for all you computer geeks; go ahead, render it...):
Define:
$$ 0.999\dots =\sum_{i=1}^\infty 9*10^{-i} $$
Corollory
This is a geometric series, so we get:
$$ 0.999\dots =\sum_{i=1}^\infty 9*10^{-i}={9\over 1-10}={9\over 9}=1 $$
Proof
$$ 3.999\dots =3+0.999\dots=3+1=4 $$
Quod Erat Demonstrandum
The only problem is that black holes are relatively rare. Take the number of binary star systems compared with the number of stars. Compare that ratio with the number of black holes and you'll see how likely a binary black hole system would be to occur. Black holes can be regarded as the same thing as stars only that they have so much mass that the escape velocity from their surface is greater than the speed of light. Some black holes also create a singularity, but I believe most are just big stars, and can be treated as such.
What about non text data? ... line drawings...
Use a simple vector format like metapost which should be fairly easy to decypher. A gif would probably be dificult to figure out, and a jpg nearly impossable. For bitmaps, pnm is rather simple (asci art would probably be meaningless because they wouldn't know what a 00100101 would look like).
I am sitting happily right now on my internet2 connection. It's nice downloading from metalab.unc.edu 500KB/s. What I really like is that I've been doing research with some supercomputers over in another university. I can send the X programs over to my display with almost no latency. I can even rotate molecules and do complex visualizations. I don't even keep a computer in my apartment because I would never be able to stand the dial-up connection. Sorry for bragging...
Interesting that you tell a bunch of computer geeks that one of their favorite star wars characters will be computer generated, and they all get angry. Don't we like computers? Perhaps we have a human side to us after all, and realize that for some things computers are not the answer. Could it be that was have an artistic side after all? (okay, programming is an art form, but...). Just some food for thought.
So what would happen if I decided to set up a DNS server, alias as many websites I can find to .smart and .stupid domain names (unofficial, of course), and suggest that people add my DNS server to thier resolve.conf file. Would this be illeagal? It would sure be fun (although lots of work). I wonder what sort of powers exist that keep underground domain name servers from sprining up.
I've successfully gotten through college without purchasing too many text books. In the field of mathematics, the textbooks are extremely small, and so I would just photocopy the 200 pages (reducing slightly to fit two pages on a sheet), and use that. 100 x 5cents = $5. These books generally go for $80-$100 each.
I would also check books out from the library. I checked out a physical chemistry textbook for p-chem, chemical energetics, spectroscopy, and a year of research. I saved over a hundred dollars (and somehow avoided library fines). I've never had any literature courses, but one could be sure that I would get those books from the Gutenberg Project if not from the library.
Of course, there were even a few courses that I just plain went without a book in. I would pay attention during the lecture (well...), and somehow would get by without failing. Homework sets were a problem in some classes, but I rarely turned in homework anyway. I purchased my Organic Chem textbook my freshman year. After realizing by the end of the two semesters that I had never read it, I decided that I would make sure I need a text book before purchasing one.
I suppose since I'm working on a doctorate and want to stay in education that I will at some point end up being a professor. I vow now that I will never ask any student to purchase one of these digital books. I will teach my courses in such a manner that a student could get by with any general textbook in the subject (hence be able to use the library), and make any problem sets available on the course web site (as well as in printouts for those who don't like computers). Education should not discriminate against people who don't have limitless financial resources (or cheap people like me).
I'm sure everbody can do that!!
Not quite as easy as that for me. Galleon doesn't have an rpm for alpha linux, so I have to compile it on my own. Not too bad until one realizes that it requires header files from mozilla, so the source for mozilla has to be downloaded (which is over 200MB unpacked). Now I see that all the header files aren't in the uncompiled source, so it looks like I'm going to have to compile mozilla. If the source is over 200MB, it's probably going to need several GB's to compile. So now I have to make room on a partition, and let it compile all day (it probably won't compile right either given my luck). Glad it's easy for some people, though.
I don't see fortran anywhere. I don't see a hex editor either. And what if I wanted to code in assembly for a R1000 CPU? At least they have vi.
...and then the NeXT community sued Apple.
Another big plus to having lots of little SCSI drives is that if one goes bad, it can be replaced with only a few gigs of data lost (unless if you keep backups on different drives like I do). If your 80GB IDE drive goes bad, you've got a major problem. If it's just bad sectors, you can partition in so that the bad cylender isn't being used (I've done it, and it's a pain), but if it stops spinning up, that's a major investment gone. I've noticed that SCSI drives tend to be better made, and I've never seen bad sectors show up (which means they probably allocate more space to fixing them). I still use my IDE drives to store mp3's, but 4GB SCSI3 is plenty for me to play with.
Sure, methanol is poisonous, but so is gasoline. People probably don't drink gasoline because it smells bad. I beleive that this smell is added to prevent accidental drinking (at least I know it is for methane). Ethanol would be safer (our bodies don't convert it to fomaldehyde and formic acid as they do with methanol), but the temptation to drink it would probably be too great for some people. Regardless, ethanol is corrently used to cut down gasoline in the corn belt, and can be used at about 30% or so without modifications to the engine (it runs cleaner and soaks up water in the gas tank too). The reason why we aren't switching fuels is because the oil industry is too big and powerful to let that happen (conspiracy!).
HCl isn't that bad, particularly when only the skin is exposed. H2SO4 can be nasty, and will turn a person's skin black. This is not due to the acidity, but rather affinity for water. If it is being rinsed with water, this shouldn't be a problem. A drop of thionyl chloride thus isn't too much of a danger. If it isn't rinsed off quickly, it will react with the water in one's skin to make those lovely acids, which would destroy lipids, etc. If one's eyes got splashed with thionyl chloride, it would react with the water there, giving off lots of heat, and blinding acid (much the same as tear gas only worse). Instead of quickly walking to the eyewash station to rinse, the victim would most likely claw thier eyes out because they would hurt so much. That's why one never works in a lab alone...
Well, you may be right (I don't know C++, so I can't in good faith argue), but I am doing research in quantum chemistry using lots of computational tools. Everything is written in fortran. Take a look at virtually any piece of software written for this subject and you'll see that it's written in fortran. Sometimes they'll use C for a user interface (if there is one), but all the major code is in fortran. Let's say I want to multiply two matricies. How do I do that in C++? Compare f90's way of doing that, and you'll see why fortran is still used so much. I've run a few C++ programs, and have noticed that they always tend to be a bit sluggish. Fortran is written with one view in mind: to crunch numbers. C++ has other interests. There are specialized languages for math, but I've never seen anyone use them. There is a standard in the sciences (math, chemistry, physics, geology, etc. (ok, not CS)), and it's fortran. How well C++ can be used for MPI and PVM? Perhaps just fine (I don't know), but it's quite easy in fortran. I will admit that fortran isn't always portable (I've had troubles with g77), but if a good compiler is used (sorry g77 developers) like SGI's, then things compile and run without a problem. I don't want to get involved in any compiler wars, but will have to go along with fortran as the best for a scientist to learn and use.