They could use this for tracking the number of computers the program is installed on, which would work independently of current user, IP, or even reinstalls. Combined with other things this could be a unique and interesting statistic that's hard (impossible) to test by other methods.
Unfortunately, there's no such thing as a "good record deal." The numbers are so stacked against the people making the music that, as recent Atlantic Records signee David Garza noted, "It only works for the artist if more than a million copies are sold. Period."
The obvious problem with that, of course, is that of the approximate 30,000 albums released every year, less than one percent go platinum (certified sales of one million), meaning there are very few recording artists for whom the record deal is actually working.
This problem is further compounded by the fact that very few musicians know what their record contract actually says. Which is quite understandable; the average Egyptologist had a better shot at deciphering hieroglyphics before the discovery of the Rosetta Stone than the average musician today has of making heads or tails out of their recording contract. This turns out to be, perhaps, the worst problem of all since those pages upon pages of tediously rigid terminology, obscure even to the legally trained, hold the financial fate of an artist.
"Once it's installed, the organizers can send alerts to users or update the software with scripts that know how to take particular actions, such as automatically filling in feedback forms on a politician's website. End users can also forward e-mail alerts to their friends, who have the option of installing the software themselves and joining the network."... "By picking a couple of issues that all Americans agree on, we can really rain holy privacy hellfire," Scannell said.
If you simply define spam as "unwanted commentary," a large, disruptive user base that does nothing but repeat itself could easily be placed in there.
Another problem is this: Dr. Smith disagrees with the movement being "addressed" by the Collactive users and wishes to comment. She/He should be able to offer feedback like anyone else, but if 537 near-duplicate comments fly in while she/he responds, then his/her comment is very likely to be either mass-deleted or simply overlooked.
The point is simply this: political debates should be won by the good arguments, and NOT by drowning the opposing side in a flood of automated replies. From where I'm sitting, this just looks like a hack of a piece of software trying to push a hack of an argument.
DRM is far from perfect, but it's encouraging to see that I have a chance of someday releasing a mini-video to the internet without it ending up stripped and hosted on some site like they are the ones that made it. Current options are unreasonably expensive ($100's/year) for some who only releases one short, free e-video per year. I already am _paying_ money to host it, too.
I'm sure you know the sites I'm talking about. I hate seeing someone post a link on the forum and you go look and it's your video with someone else's name on it.
I ran a test using the sort slasrt_() from the CLAPACK library as a comparison. It was last updated September 30, 1994, and I didn't tweak it for this test. On a Pentium 4 3.0GHz, I achieved the following results:
Input Sequence Size: 1024
Qsort() Time: 0.000 sec
Input Sequence Size: 2048
Qsort() Time: 0.000 sec
Input Sequence Size: 4096
Qsort() Time: 0.000 sec
Input Sequence Size: 8192
Qsort() Time: 0.001 sec
Input Sequence Size: 16384
Qsort() Time: 0.002 sec
Input Sequence Size: 32768
Qsort() Time: 0.005 sec
Input Sequence Size: 65536
Qsort() Time: 0.011 sec
Input Sequence Size: 131072
Qsort() Time: 0.028 sec
Input Sequence Size: 262144
Qsort() Time: 0.049 sec
Input Sequence Size: 524288
Qsort() Time: 0.093 sec
Input Sequence Size: 1048576
Qsort() Time: 0.193 sec
Input Sequence Size: 2097152
Qsort() Time: 0.386 sec
Input Sequence Size: 4194304
Qsort() Time: 0.798 sec
Input Sequence Size: 8388608
Qsort() Time: 1.625 sec
I feel this is a legitimate comparison, even though the algorithm varies from the algorithm used on the graphics cards, because the original test already compared different sorting algorithms. So if someone is looking for the result of "how fast can the P4 3._2_ sort a list of numbers?" as a comparison, the answer sure isn't "~6-7 seconds for a list of 8.3million 32-bit floats."
Slashdot Mirror
They could use this for tracking the number of computers the program is installed on, which would work independently of current user, IP, or even reinstalls. Combined with other things this could be a unique and interesting statistic that's hard (impossible) to test by other methods.
Unfortunately, there's no such thing as a "good record deal." The numbers are so stacked against the people making the music that, as recent Atlantic Records signee David Garza noted, "It only works for the artist if more than a million copies are sold. Period."
The obvious problem with that, of course, is that of the approximate 30,000 albums released every year, less than one percent go platinum (certified sales of one million), meaning there are very few recording artists for whom the record deal is actually working.
This problem is further compounded by the fact that very few musicians know what their record contract actually says. Which is quite understandable; the average Egyptologist had a better shot at deciphering hieroglyphics before the discovery of the Rosetta Stone than the average musician today has of making heads or tails out of their recording contract. This turns out to be, perhaps, the worst problem of all since those pages upon pages of tediously rigid terminology, obscure even to the legally trained, hold the financial fate of an artist.
The Cold, Hard Truth About Recording ContractsFrom the article:
...
"Once it's installed, the organizers can send alerts to users or update the software with scripts that know how to take particular actions, such as automatically filling in feedback forms on a politician's website. End users can also forward e-mail alerts to their friends, who have the option of installing the software themselves and joining the network."
"By picking a couple of issues that all Americans agree on, we can really rain holy privacy hellfire," Scannell said.
If you simply define spam as "unwanted commentary," a large, disruptive user base that does nothing but repeat itself could easily be placed in there.
Another problem is this: Dr. Smith disagrees with the movement being "addressed" by the Collactive users and wishes to comment. She/He should be able to offer feedback like anyone else, but if 537 near-duplicate comments fly in while she/he responds, then his/her comment is very likely to be either mass-deleted or simply overlooked.
The point is simply this: political debates should be won by the good arguments, and NOT by drowning the opposing side in a flood of automated replies. From where I'm sitting, this just looks like a hack of a piece of software trying to push a hack of an argument.
DRM is far from perfect, but it's encouraging to see that I have a chance of someday releasing a mini-video to the internet without it ending up stripped and hosted on some site like they are the ones that made it. Current options are unreasonably expensive ($100's/year) for some who only releases one short, free e-video per year. I already am _paying_ money to host it, too.
I'm sure you know the sites I'm talking about. I hate seeing someone post a link on the forum and you go look and it's your video with someone else's name on it.
I ran a test using the sort slasrt_() from the CLAPACK library as a comparison. It was last updated September 30, 1994, and I didn't tweak it for this test. On a Pentium 4 3.0GHz, I achieved the following results: Input Sequence Size: 1024 Qsort() Time: 0.000 sec Input Sequence Size: 2048 Qsort() Time: 0.000 sec Input Sequence Size: 4096 Qsort() Time: 0.000 sec Input Sequence Size: 8192 Qsort() Time: 0.001 sec Input Sequence Size: 16384 Qsort() Time: 0.002 sec Input Sequence Size: 32768 Qsort() Time: 0.005 sec Input Sequence Size: 65536 Qsort() Time: 0.011 sec Input Sequence Size: 131072 Qsort() Time: 0.028 sec Input Sequence Size: 262144 Qsort() Time: 0.049 sec Input Sequence Size: 524288 Qsort() Time: 0.093 sec Input Sequence Size: 1048576 Qsort() Time: 0.193 sec Input Sequence Size: 2097152 Qsort() Time: 0.386 sec Input Sequence Size: 4194304 Qsort() Time: 0.798 sec Input Sequence Size: 8388608 Qsort() Time: 1.625 sec I feel this is a legitimate comparison, even though the algorithm varies from the algorithm used on the graphics cards, because the original test already compared different sorting algorithms. So if someone is looking for the result of "how fast can the P4 3._2_ sort a list of numbers?" as a comparison, the answer sure isn't "~6-7 seconds for a list of 8.3million 32-bit floats."