Re:What do you say...
on
Today's SCO News
·
· Score: 0, Redundant
" I'll have a Big Mac, large fries and a Coke."
Wouldn't that be too complicated for a poor guy to understand?
IBM should buy Novell
on
Today's SCO News
·
· Score: 2, Interesting
May be IBM should buy Novell. That way they get 95% of all royalties collected by SCO too. They can then sue SCO for distributing Linux containing SVR-V code. IBM can then force all of SCO customers to move to AIX and MS will need to get license from IBM.
IBM supports wide variety of systems and supporting Novell wouldn't be that hard since Novell is business oriented only. Many people don't buy Novell products because they don't have faith that Novell would be around for long. With IBM that won't be the issue. IBM has a history of buying second rank companies: Lotus, Informix etc.
i believe in strong antropological principle: the universe is such that intelligent life must exist over a cosimic time scale.
how does nature guarantees it? few mad men can easily destroy intelligent life on earth. theoretically, even nearby stellar systems are not safe, since we can always send virii there. the only way for nature to ensure that intelligent life can exist over cosmic time, is to distribute life over cosmic distances. these means that intelligent life exists throughout the universe. further, to ensure success, the nature must have tried to create life wherever possible. thus, "if life can exist, then most likely it does exist or existed" is the final outcome of this principle.
so if mars environment does allow life to exist and flourish, then perhaps, life still exists there or atleast existed earlier.
"I often wonder why the heck credit card purchases don?t require a PIN at the very least."
do you mean credit card should require PIN + signature or just the PIN?
you use your ATM, typically at certified places. credit cards on the other hand can be used at lot more places where possibility of cheating is higher. so PIN is secure enough for ATM, but is not secure for credit card. by requiring a signature, they (credit card companies) make sure, shopkeeper cannot cheat. Note that, credit cards, when used for cash advances at ATM, you only need PIN.
Requiring signature AND PIN would be totally redundant.
Why not sell IP address blocks also, just like domain names? this will bring all unused addresses back to the market. First assign number of IP addresses to each country based on their population taken some time in the past. Then allow them to auction these addresses whichever way they want. If some country needs more IP addresses, they can just buy from another country.
The problem with IPv6 is that it is totally different from IPv4. The change is not just from 32 bit to 128 addressing. If it were this simple, the migration could have been easier. Doing address reallocation in IPv4 is hard. But if we had designed something which is like IPv4 but with 64 bit or 128 bit addressing and done reallocation, we would have been able to resolve this issue somewhat more easily. I guess, in this area, the 100 year old system of telephony has worked better.
the header tries to give impression as if the whole third world is greatly impacted by malaria. this is not true. it is true that most cases of malaria are in third world, but most of the third world is not in mess because of malaria. nearly 80-90 % of the estimated 1.5 to 2.7 million deaths occur in sub-sahara africa where about 90% of the infected people leave (www.idrc.ca/books/reports/1996/01-07e.html). this means 10-20% of 1.5-2.7 million deaths occurs in rest of the third world. taking middle figures, this gives 315K/yr deaths. this figure is not too much differnt than common flu.
since Bismuth decay rate is the slowest that has been observed, how do they know that other stable isotopes are not so stable? if they don't, then how do they claim that Bismuth is not the heaviest stable element. I guess, the only sure heaviest element would then turn out to be Hydrogen-1 isotope.
who wants to touch MIT women?
on
Shocking Clothing
·
· Score: 0, Flamebait
until a theory has any way to disprove it, it is largely a philosophic piece rather than scientific. multiverse is one such thing. I can easily say that if i had not typed this letter, everyone on this earth would have beed dead. Go prove me wrong!
other major factor affecting sales data of RIAA is that CDs are digital and it is the only digital product with no upgrade in 21 years! My 1985 CD is as good as new. So while in the past, people used to buy a new copy of the album to replace used one, it is no more necessary.
Also, many people who had vinyl, tape etc, replaced such things with CD. The replacement is largely complete. During the replacement period, people not only bought albums they didn't have, but also bought albums they had. Now, people only buy what they don't have.
To analyze the above points, the RIAA should publish data of sales of new CD albums only and see if there is any decline. My guess is that it is actually increasing. By means of new, I mean never published before.
The third major factor is legal copying. IANAL, but I think it is allowed by law to make duplicate copies of album for personal use. It was hard to make such copies for tape and impossible for vinyl, but this is trivial for CDs.
So, it is doubtful that piracy is the cause of declining RIAA sales.
Same genus is not same as same species. Species is the most scientifically and strictly defined term than any other classification terms. All others are mostly defined on the basis of some characteristics which does not necessarily tell us the distance between one type to another. Also, species is the only definition which can be determined experimentally. To determine if two group of animals are same species or no, you need to interbreed them. If they can produce male and female fertile offsprings, then they are same species. No such experimental definition can be made for genus. Read the comment in the article, it says, '...chimps and humans split six to 10 million years ago. "That's an awful long time to be in the same genus,"...'. You see, this is how they argue about genus.
So there is no big deal, when some scientist determines humans are mostly chimps. All that s/he says is that the distance between human and chimp species is less than we thought. And mind you, statistically, there was a 50% chance that this study would have said this!
Not true. The SGI didn't have and still doesn't have IRIX on X86, so they couldn't migrate their customers successfully. If I am installing Linux or Windows on X86, why should SGI matter? On the other hand, if you need Solaris X86, it does matter. Thanks to SCO, Sun's importance is even higher. This is number one difference between DEC, SGI, HP migration to PC and Sun's.
SGI also didn't have SGI proprietary software installed free on their x86 boxes. OTOH, Sun includes StarOffice, JDK, App Server etc.
SGI was more expensive than Dell, HP etc... I just compared Sun offering and found that they are cheaper than even Dell.
SGI x86 hardware (initial) was proprietary. I remember stock Linux would not install on them. Sun hardware is same as rest of X86.
for the first time, apple to apple comparison shows sun cheaper than dell. i selected sun v65 and tried identical system at dell. dell doesn't give 3.06 GHz in 2U rack, so i selected 2.8 GHz. This is 600 cheaper. However, Dell charges $600 for upgrade from 2.6 to 2.8, so their upgrade from 2.8 to 3.06 would have been higher than 600 (upgrade from 2.4 to 2.6 is 200, 2.6 to 2.8 is 600). dell comes with customer installed RedHat Advanced Server while Sun comes with Solaris 9 and both are atleast comparable system (to be frank, RH profession is cheaper. but i am aware of several server apps which require RH AS patches and won't be certified on RH Pro).
the idea is that no one knows if they are on the list. so the david nelson has no idea that he is on the list and he may try to fly on his own name and will be arrested. police do these kind of things all the time. you may be suspected in some crime but you have no idea. when you are stopped by a traffic cop, he/she will check your license and see if you are wanted for any investigation. it is just that sometimes, the names are too common, search is too wide, inconvenience is very high, name remains on the list for long etc... which makes it special.
It is doubly misleading. In a single exposure, the SNR increases as the exposure time. On the other hand, when you combine multiple images, you get noise from each image. Thus signal increases linearly and noise increases as a square root of the number of images combined. If all (250) images had identical exposure time, the resulting image has 16 times less SNR than a single image of 3.5 days exposure. This is the worst case and in this case, it is equivalent of about 5 hours exposure. So effectively, they wasted 3 days of hubble time!
Caldera used to distribute Linux. Implicitly, they licensed all trade secrets, patents, copyright on whatever source went in that. So the only code which they can claim to be infringing is that which is not included in any versions of Caldera Linux.
This means, we only need to check the code which is not part of Caldera and account for it. Also, we only need to include the type of code that SCO is developing. Thus any code related to porting to different architecture, device driver, font, X, KDE, GNOME, file systems etc is safe. Any code SCO wants to lay claims, they will have to prove that they never distributed it.
If I were a commercial Linux user, I would feel totally safe.
It is also likely, that accidently or deliberately some of the Linux code has gone in SCO Unix. Linux vendors should file counter claim against SCO to check for this. By doing this, you can ask all recipient of SCO Unix to stop using it and there by get all of them to use Linux!
the disk storage has reduced cost faster than moore's law in recent times. however, the overall rate is not significantly higher than microchips. in mid 80's, 8 kbytes ram used to cost $50. today, you can get 512 MB for the same price. that is 64K tims price reduction which is same as for hard drive. also, the ram price reduced faster than harddrive from 1956 (hard drive introduction year) till mid-80s. in 1956, 5 Meg hard drives were available. in mid-80s, this figure barely jumped to couple of gigs. that is about several hundred times. on the other hand, in 1956, a single byte RAM required a whole circuit board and in 1985, Toshiba had announced Mega bit ram chips. so overall, microchips have made faster progress than harddrives.
Paperless toilets are far more likely than paperless office. in most civilized nations, it is customery to wash your ass with water than just wipe it with paper.
if you are ready to take chances on software format like MPEG-4 or some other highly compressive codec, then you can convert your collection to DVD quite cheaply. 650 video cassettes at 2 hours each is 1300 hours of video. Assuming 1 GBytes/hour (good enough for VHS quality), you need 1.3 TB space or about 300 DVD disks. At $1 a disk (in bulk market), the cost is $300. Add to it your labor and equipment. If you want to be more flexible, use standard DVD encoding, and you will need 650 disks. As for recording equipment, I would recommend stand alone equipment (costs 400-500), since they are easier to handle. Once done copying, buy a 300 disc DVD changer (costs 300-500) and enjoy.
the guys are as much guilty as the guy who laid pipe connecting dorms to the campus computers. the search engine is a generic process. google, yahoo etc have vested interest and sooner or later, they too would have joined the fight in their favor. by settling the case, they have given wrong signal to the world. if a case like this goes to jury, i doudt, any jury would consider them guilty. it would be hard to find a juror in US who hasn't recorded a song from somewhere (either friend's borrowed media, or doenloaded or from public library etc). the most euphemistic term that I can use for these guys is "practical coward".
4 tracks a second means 120 million tracks a year or roughly about 3% of all retail tracks sold by the music companies which have licenses with apple. my guess is that for every 1 apple owner, there are 100 CD player owners. So an average Apple owners are buying three times more tracks than an average CD player owner is buying in retail market! Indeed amazing if the trend continues.
one of the greatest strength of Unix is symlinks. Unfortunately that is one of its weakness also. Imagine a newbie who sees following two files:/home/photos/my_most_precious_photo.jpg
and/user/photos/my_most_precious_photo.jpg. Well,
he or she thinks this is duplicate and deletes one of the files which happens to be a real one! The default version of "ls" with no option makes no disntinction between the symlink and a real file.
Few possible solutions, none perfect:
Warn user during delete if any symlinks are pointing there. Requires kernel filesystem rewrite.
Default "ls" should be modified to warn user that the file is a symlink. This may break many shell scripts.
Shells should have "ls" built-in. In interactive mode, it should warn users. Requires users to use one of these shells.
Second problem with symlink is that you can't move up and down in the hierarchy in intuitive way. If you do "cd x/y" followed by "cd..". You should in directory "x", right? Not necessarily, if you are using symlinks. Since "cd" command is a shell built-in, the shells should be able to keep track of directory navigation and should be able to keep track of this, so that user would in directory "x" even if there are symlinks. This may break some unknown things.
In short, I believe that for home user, symlinks should not be requirement (no executible or scripts should use) and user must get visible signs that they are dealing with symlinks whenever they encounter one.
The two beasts are quite different. First check the quality of photo taken with 1.3 MP camera vs 1.3 MP camcorder. The camera photo is far superior. The camera CCDs are designed for better dynamic range, color, long exposure etc.
Next check the size of CCD. The camera CCD size is big. So to create a reasonable zoom lens for video, you need really big (physically) lens. CCD size in most camcorder is 1/4.7" (this is related to physical size, but is not the physical size). Camera CCD on the other hand are about 2/3" (Sony 717, Nikon 5000 etc).
As you increase more pixels, you need corresponding bigger lens (else light falling on each pixel with reduce and will reduce low light sensitivity). Again this is a challenging thing to do. This is one of the reason, why we are seeing saturation in digital camera too. It was more than 2 years ago, 5 MP non-interchangeable lens digital cameras were announced. We haven't had any pixel increase since then. We can scale pixel count, but without scaling sensitivity, lens size etc, it is of no use.
Camcorder CCDs also have to read pictures really fast. This increases noise. In digital camera, you are reading CCD fast only in preview mode. During the actual picture taking, the CCD read is really slow (compared to camcorder). Another reason, why it is harder to increase camcorder pixels.
Slashdot Mirror
" I'll have a Big Mac, large fries and a Coke."
Wouldn't that be too complicated for a poor guy to understand?
May be IBM should buy Novell. That way they get 95% of all royalties collected by SCO too. They can then sue SCO for distributing Linux containing SVR-V code. IBM can then force all of SCO customers to move to AIX and MS will need to get license from IBM.
IBM supports wide variety of systems and supporting Novell wouldn't be that hard since Novell is business oriented only. Many people don't buy Novell products because they don't have faith that Novell would be around for long. With IBM that won't be the issue. IBM has a history of buying second rank companies: Lotus, Informix etc.
i believe in strong antropological principle: the universe is such that intelligent life must exist over a cosimic time scale.
how does nature guarantees it? few mad men can easily destroy intelligent life on earth. theoretically, even nearby stellar systems are not safe, since we can always send virii there. the only way for nature to ensure that intelligent life can exist over cosmic time, is to distribute life over cosmic distances. these means that intelligent life exists throughout the universe. further, to ensure success, the nature must have tried to create life wherever possible. thus, "if life can exist, then most likely it does exist or existed" is the final outcome of this principle.
so if mars environment does allow life to exist and flourish, then perhaps, life still exists there or atleast existed earlier.
"I often wonder why the heck credit card purchases don?t require a PIN at the very least."
do you mean credit card should require PIN + signature or just the PIN?
you use your ATM, typically at certified places. credit cards on the other hand can be used at lot more places where possibility of cheating is higher. so PIN is secure enough for ATM, but is not secure for credit card. by requiring a signature, they (credit card companies) make sure, shopkeeper cannot cheat. Note that, credit cards, when used for cash advances at ATM, you only need PIN.
Requiring signature AND PIN would be totally redundant.
'Ok... so define "fair". Sure,'...
Why not sell IP address blocks also, just like domain names? this will bring all unused addresses back to the market. First assign number of IP addresses to each country based on their population taken some time in the past. Then allow them to auction these addresses whichever way they want. If some country needs more IP addresses, they can just buy from another country.
The problem with IPv6 is that it is totally different from IPv4. The change is not just from 32 bit to 128 addressing. If it were this simple, the migration could have been easier. Doing address reallocation in IPv4 is hard. But if we had designed something which is like IPv4 but with 64 bit or 128 bit addressing and done reallocation, we would have been able to resolve this issue somewhat more easily. I guess, in this area, the 100 year old system of telephony has worked better.
the header tries to give impression as if the whole third world is greatly impacted by malaria. this is not true. it is true that most cases of malaria are in third world, but most of the third world is not in mess because of malaria. nearly 80-90 % of the estimated 1.5 to 2.7 million deaths occur in sub-sahara africa where about 90% of the infected people leave (www.idrc.ca/books/reports/1996/01-07e.html). this means 10-20% of 1.5-2.7 million deaths occurs in rest of the third world. taking middle figures, this gives 315K/yr deaths. this figure is not too much differnt than common flu.
since Bismuth decay rate is the slowest that has been observed, how do they know that other stable isotopes are not so stable? if they don't, then how do they claim that Bismuth is not the heaviest stable element. I guess, the only sure heaviest element would then turn out to be Hydrogen-1 isotope.
enough said!
there is an old saying, "it is hard to understand other's pain". don't read BBC to find out the pain of buddishts. ask buddhists.
until a theory has any way to disprove it, it is largely a philosophic piece rather than scientific. multiverse is one such thing. I can easily say that if i had not typed this letter, everyone on this earth would have beed dead. Go prove me wrong!
other major factor affecting sales data of RIAA is that CDs are digital and it is the only digital product with no upgrade in 21 years! My 1985 CD is as good as new. So while in the past, people used to buy a new copy of the album to replace used one, it is no more necessary.
Also, many people who had vinyl, tape etc, replaced such things with CD. The replacement is largely complete. During the replacement period, people not only bought albums they didn't have, but also bought albums they had. Now, people only buy what they don't have.
To analyze the above points, the RIAA should publish data of sales of new CD albums only and see if there is any decline. My guess is that it is actually increasing. By means of new, I mean never published before.
The third major factor is legal copying. IANAL, but I think it is allowed by law to make duplicate copies of album for personal use. It was hard to make such copies for tape and impossible for vinyl, but this is trivial for CDs.
So, it is doubtful that piracy is the cause of declining RIAA sales.
Same genus is not same as same species. Species is the most scientifically and strictly defined term than any other classification terms. All others are mostly defined on the basis of some characteristics which does not necessarily tell us the distance between one type to another. Also, species is the only definition which can be determined experimentally. To determine if two group of animals are same species or no, you need to interbreed them. If they can produce male and female fertile offsprings, then they are same species. No such experimental definition can be made for genus. Read the comment in the article, it says, '...chimps and humans split six to 10 million years ago. "That's an awful long time to be in the same genus,"...'. You see, this is how they argue about genus.
So there is no big deal, when some scientist determines humans are mostly chimps. All that s/he says is that the distance between human and chimp species is less than we thought. And mind you, statistically, there was a 50% chance that this study would have said this!
Not true. The SGI didn't have and still doesn't have IRIX on X86, so they couldn't migrate their customers successfully. If I am installing Linux or Windows on X86, why should SGI matter? On the other hand, if you need Solaris X86, it does matter. Thanks to SCO, Sun's importance is even higher. This is number one difference between DEC, SGI, HP migration to PC and Sun's.
SGI also didn't have SGI proprietary software installed free on their x86 boxes. OTOH, Sun includes StarOffice, JDK, App Server etc.
SGI was more expensive than Dell, HP etc... I just compared Sun offering and found that they are cheaper than even Dell.
SGI x86 hardware (initial) was proprietary. I remember stock Linux would not install on them. Sun hardware is same as rest of X86.
for the first time, apple to apple comparison shows sun cheaper than dell. i selected sun v65 and tried identical system at dell. dell doesn't give 3.06 GHz in 2U rack, so i selected 2.8 GHz. This is 600 cheaper. However, Dell charges $600 for upgrade from 2.6 to 2.8, so their upgrade from 2.8 to 3.06 would have been higher than 600 (upgrade from 2.4 to 2.6 is 200, 2.6 to 2.8 is 600). dell comes with customer installed RedHat Advanced Server while Sun comes with Solaris 9 and both are atleast comparable system (to be frank, RH profession is cheaper. but i am aware of several server apps which require RH AS patches and won't be certified on RH Pro).
the idea is that no one knows if they are on the list. so the david nelson has no idea that he is on the list and he may try to fly on his own name and will be arrested. police do these kind of things all the time. you may be suspected in some crime but you have no idea. when you are stopped by a traffic cop, he/she will check your license and see if you are wanted for any investigation. it is just that sometimes, the names are too common, search is too wide, inconvenience is very high, name remains on the list for long etc... which makes it special.
It is doubly misleading. In a single exposure, the SNR increases as the exposure time. On the other hand, when you combine multiple images, you get noise from each image. Thus signal increases linearly and noise increases as a square root of the number of images combined. If all (250) images had identical exposure time, the resulting image has 16 times less SNR than a single image of 3.5 days exposure. This is the worst case and in this case, it is equivalent of about 5 hours exposure. So effectively, they wasted 3 days of hubble time!
Caldera used to distribute Linux. Implicitly, they licensed all trade secrets, patents, copyright on whatever source went in that. So the only code which they can claim to be infringing is that which is not included in any versions of Caldera Linux.
This means, we only need to check the code which is not part of Caldera and account for it. Also, we only need to include the type of code that SCO is developing. Thus any code related to porting to different architecture, device driver, font, X, KDE, GNOME, file systems etc is safe. Any code SCO wants to lay claims, they will have to prove that they never distributed it.
If I were a commercial Linux user, I would feel totally safe.
It is also likely, that accidently or deliberately some of the Linux code has gone in SCO Unix. Linux vendors should file counter claim against SCO to check for this. By doing this, you can ask all recipient of SCO Unix to stop using it and there by get all of them to use Linux!
the disk storage has reduced cost faster than moore's law in recent times. however, the overall rate is not significantly higher than microchips. in mid 80's, 8 kbytes ram used to cost $50. today, you can get 512 MB for the same price. that is 64K tims price reduction which is same as for hard drive. also, the ram price reduced faster than harddrive from 1956 (hard drive introduction year) till mid-80s. in 1956, 5 Meg hard drives were available. in mid-80s, this figure barely jumped to couple of gigs. that is about several hundred times. on the other hand, in 1956, a single byte RAM required a whole circuit board and in 1985, Toshiba had announced Mega bit ram chips. so overall, microchips have made faster progress than harddrives.
if yes, then it is a crime and must be banned!
Paperless toilets are far more likely than paperless office. in most civilized nations, it is customery to wash your ass with water than just wipe it with paper.
if you are ready to take chances on software format like MPEG-4 or some other highly compressive codec, then you can convert your collection to DVD quite cheaply. 650 video cassettes at 2 hours each is 1300 hours of video. Assuming 1 GBytes/hour (good enough for VHS quality), you need 1.3 TB space or about 300 DVD disks. At $1 a disk (in bulk market), the cost is $300. Add to it your labor and equipment. If you want to be more flexible, use standard DVD encoding, and you will need 650 disks. As for recording equipment, I would recommend stand alone equipment (costs 400-500), since they are easier to handle. Once done copying, buy a 300 disc DVD changer (costs 300-500) and enjoy.
the guys are as much guilty as the guy who laid pipe connecting dorms to the campus computers. the search engine is a generic process. google, yahoo etc have vested interest and sooner or later, they too would have joined the fight in their favor. by settling the case, they have given wrong signal to the world. if a case like this goes to jury, i doudt, any jury would consider them guilty. it would be hard to find a juror in US who hasn't recorded a song from somewhere (either friend's borrowed media, or doenloaded or from public library etc). the most euphemistic term that I can use for these guys is "practical coward".
4 tracks a second means 120 million tracks a year or roughly about 3% of all retail tracks sold by the music companies which have licenses with apple. my guess is that for every 1 apple owner, there are 100 CD player owners. So an average Apple owners are buying three times more tracks than an average CD player owner is buying in retail market! Indeed amazing if the trend continues.
Few possible solutions, none perfect:
Warn user during delete if any symlinks are pointing there. Requires kernel filesystem rewrite.
Default "ls" should be modified to warn user that the file is a symlink. This may break many shell scripts.
Shells should have "ls" built-in. In interactive mode, it should warn users. Requires users to use one of these shells.
Second problem with symlink is that you can't move up and down in the hierarchy in intuitive way. If you do "cd x/y" followed by "cd ..". You should in directory "x", right? Not necessarily, if you are using symlinks. Since "cd" command is a shell built-in, the shells should be able to keep track of directory navigation and should be able to keep track of this, so that user would in directory "x" even if there are symlinks. This may break some unknown things.
In short, I believe that for home user, symlinks should not be requirement (no executible or scripts should use) and user must get visible signs that they are dealing with symlinks whenever they encounter one.
Next check the size of CCD. The camera CCD size is big. So to create a reasonable zoom lens for video, you need really big (physically) lens. CCD size in most camcorder is 1/4.7" (this is related to physical size, but is not the physical size). Camera CCD on the other hand are about 2/3" (Sony 717, Nikon 5000 etc).
As you increase more pixels, you need corresponding bigger lens (else light falling on each pixel with reduce and will reduce low light sensitivity). Again this is a challenging thing to do. This is one of the reason, why we are seeing saturation in digital camera too. It was more than 2 years ago, 5 MP non-interchangeable lens digital cameras were announced. We haven't had any pixel increase since then. We can scale pixel count, but without scaling sensitivity, lens size etc, it is of no use.
Camcorder CCDs also have to read pictures really fast. This increases noise. In digital camera, you are reading CCD fast only in preview mode. During the actual picture taking, the CCD read is really slow (compared to camcorder). Another reason, why it is harder to increase camcorder pixels.