The ACLU is already fighting the "national security letter" issue. And they're winning. Judge Marrero wrote, in his decision: "Democracy abhors undue secrecy. . . . [A]n unlimited government warrant to conceal, effectively a form of secrecy per se, has no place in our open society."
Is there a real court order here, or is this just an FBI letter?
There's a paper linked from the article, but it's so short and weak that it's hard to tell if anything useful has been accomplished. In particular, it's not clear it if scales beyond the 25-line program shown. There are vague claims that it uses some database of "common sense" to help build the programs, but the paper shows no evidence of this.
This isn't a new non-silicon solar cell technology. It's just a metal base under a thin layer of silicon, instead of a thick silicon wafer. This reduces weight, but it doesn't help cost or performance. It may have space applications.
Their solar cells are made in a wafer fab and have no more than 15% efficiency, like everybody else's.
So this isn't the Great Solar Breakthrough. Sorry.
Injecting code - blech.
on
Hacking Mac OS X
·
· Score: 2, Interesting
Microsoft put that in Windows, where it was a bad idea. Now somebody put it on the Mac. This is progress?
The real problem is that interprocess communication under UNIX isn't very good, was added late, isn't portable, and isn't used much. So apps tend to be
monolithic, and intercommunication takes place at a very high level, like CORBA, XSLT, or Java RMI, if at all.
So trying to interpose new features at a lower level tends to involve horrible hacks. In the DOS era, there was "hooking" interrupts (a concept faithfully replicated in all Microsoft's OSs to date.) Then came "injecting DLLs." Now there's this.
One of the sad things about UNIX/Linux is that the original concept of little intercommunicating programs has been lost. Because the original intercommunication mechanism (pipes) was so weak, the concept didn't generalize.
I often wonder how different the history of UNIX might have been if, when you invoked a program, you got results back. You get to pass command line arguments and environment variables into a subprocess, but all you get back is a status code. This one-way model permeates the UNIX world. It's one reason that shell scripts and makefiles tend to be so blind.
What's needed is a sane approach to interprocess subroutine calls. Multics had this. QNX has it. Mach has support for it, but nobody uses it much.
I have a PC (a Shuttle with an Athlon processor) that corrupts files, with single-bit errors, under QNX. This machine ran for two years, then stopped working right after a power failure. I've replaced the hard drive, the CD-ROM drive, and the boot CD. Reformatting the hard drive and reinstalling the OS doesn't fix the problem. Even zeroing the whole drive with a standalone boot utility doesn't help. Setting the BIOS parameters to the default doesn't help. Reinstall of the OS (QNX, again) fails, complaining about some corrupted file, different each time. Sometimes it's a text file, and I can boot from CD-only, go in with an editor, and see a bad character on disk. It's a real hardware problem.
What's annoying is that the machine will pass all easily available diagnostics. The hard drive will pass a 12-hour surface scan. DRAM will pass hours of memtest. Simple CPU diagnostics pass.
There are expensive burn-in test programs, but I don't know which ones are any good.
What's the best way to verify an intermittent hardware problem?
Scanning or copyrighting a document or picture does not create new copyrightable content. (Bridgeman vs. Corel.)
Putting legal documents in a database does not create new copyrightable content, although adding additional information may. But you can remove that information and the result is not copyrightable. (West Publishing.)
Mere records of facts are not copyrightable. (Feist vs. Rural Telephone).
There is no "sweat of the brow" copyright in US law. Originality is required. "The standard of originality is low, but it exists" - U.S. Supreme Court in Feist.
There are animation systems (including Softimage 3D) which support three floats per pixel. This allows huge dynamic range, so you can have full sunlight and shadow in the same frame. The dynamic range is then flattened, logarithmically (like film) for output.
Graphics cards will probably start doing this soon. It's a way out of the "shades of black" problem in games.
Re:The iPod will disappear into the cell phone
on
Re-Imagining Apple
·
· Score: 1
You don't have to carry the washer and dryer around with you all day.
Redirects to a page should be treated as having far less PageRank value than the page itself. That will fix the problem.
It will also break many "click trackers", "portals", "directory sites", "search engine optimizers", and other annoyances, which is probably a plus for Google users. You know, those sites where you click on some phrase in Google and, three redirects later, you're at some irrelevant porno site.
You can dial 911 from Sprint cell phones, but you usually get a recording that says "all operators are busy". If you wait long enough, you get some operator service somewhere, but they're oriented towards highway emergencies.
(3) (a) A service provider may comply with Subsection (1) by:
(i) providing network-level filtering to prevent receipt of material harmful to minors;
or
(ii) providing at the time of a consumer's request under Subsection (1), software for
contemporaneous installation on the consumer's computer that blocks, in an easy-to-enable and
commercially reasonable manner, receipt of material harmful to minors.
(b) (i) Except as provided in Subsection (3)(b)(ii), a service provider may not charge a
consumer for blocking material or providing software under this section, except that a service
provider may increase the cost to all subscribers to the service provider's services to recover the
cost of complying with this section.
So bundling "NetNanny" with ISP service, for those who want it, is sufficient to comply.
If you're in Utah, expect your ISP bill to go up by something under a dollar per month, based on bulk pricing for NetNanny.
(Does entering "~frontdoor" as the password still turn off NetNanny?)
Finished the process of submitting a still image. Never saw the image, though. So that works, sort of. So far, all I get is "Please be patient -- it should appear soon at this media page.
You can watch new items come in on the Ourmedia main page, but you can't see them, because the update process is so slow. I submitted item #902, and half an hour later, the latest one is #907. So the load isn't high. Is the bottleneck a batch job, or does a real person have to look at this stuff before it goes up?
Tried again today. Clicking on the main Ourmedia page popped up an "enter password" dialog. No home page, just a password dialog. It wouldn't accept the password from yesterday. It just prompted for a password over and over. This looks like a broken web server.
Tried again later. No response from the ourmedia.org server at all.
Tried again later. Back to the "enter password" dialog.
I read the IBM article. Sounds like the early days of SpamCop. SpamCop traces headers back to the originator or the first phony header, to validate the source. Mail with tracing problems used to get a challenge from SpamCop, but they gave up on that. Challenge-response effectively does a denial of service attack on joe-job victims. It's also incompatible with too many legitimate autoresponder systems that send mail confirmations of transactions.
I tried to upload our DARPA Grand Challenge video. First, Ourmedia wants me to register. So I fill out the form, and it rejects the registration because I already have an Internet Archive account. Then it changes my password and mails it to me. So I log in with the Internet Archive account and the new password. Ourmedia says I'm logged in. But if I try to upload, it says I need an Internet Archive account, even though I'm logged in with one.
Slashdot Mirror
Is there a real court order here, or is this just an FBI letter?
If it's for the retail channel, it's probably at CES. If it's for business, there are other shows. That's why Comdex is dead.
Political reaction to that may be substantial.
This has been tried before, not successfully.
What bozo at Slashdot let Microsoft run a Flash ad on this page that plays an air horn sound?
Their solar cells are made in a wafer fab and have no more than 15% efficiency, like everybody else's.
So this isn't the Great Solar Breakthrough. Sorry.
The real problem is that interprocess communication under UNIX isn't very good, was added late, isn't portable, and isn't used much. So apps tend to be monolithic, and intercommunication takes place at a very high level, like CORBA, XSLT, or Java RMI, if at all.
So trying to interpose new features at a lower level tends to involve horrible hacks. In the DOS era, there was "hooking" interrupts (a concept faithfully replicated in all Microsoft's OSs to date.) Then came "injecting DLLs." Now there's this.
One of the sad things about UNIX/Linux is that the original concept of little intercommunicating programs has been lost. Because the original intercommunication mechanism (pipes) was so weak, the concept didn't generalize.
I often wonder how different the history of UNIX might have been if, when you invoked a program, you got results back. You get to pass command line arguments and environment variables into a subprocess, but all you get back is a status code. This one-way model permeates the UNIX world. It's one reason that shell scripts and makefiles tend to be so blind.
What's needed is a sane approach to interprocess subroutine calls. Multics had this. QNX has it. Mach has support for it, but nobody uses it much.
What's annoying is that the machine will pass all easily available diagnostics. The hard drive will pass a 12-hour surface scan. DRAM will pass hours of memtest. Simple CPU diagnostics pass. There are expensive burn-in test programs, but I don't know which ones are any good.
What's the best way to verify an intermittent hardware problem?
Putting legal documents in a database does not create new copyrightable content, although adding additional information may. But you can remove that information and the result is not copyrightable. (West Publishing.)
Mere records of facts are not copyrightable. (Feist vs. Rural Telephone).
There is no "sweat of the brow" copyright in US law. Originality is required. "The standard of originality is low, but it exists" - U.S. Supreme Court in Feist.
Graphics cards will probably start doing this soon. It's a way out of the "shades of black" problem in games.
You don't have to carry the washer and dryer around with you all day.
You can't do that any more. Not since the 1996 Telecommunications Act. Municipal regulation of cable companies has been severely limited.
Then again, the consumer electronics industry can't get the TV and the DVD remote to work together.
Vlad says Godot will show tomorrow.
It's been suggested by someone else that redirects within the same domain should be treated more favorably than those outside it.
It will also break many "click trackers", "portals", "directory sites", "search engine optimizers", and other annoyances, which is probably a plus for Google users. You know, those sites where you click on some phrase in Google and, three redirects later, you're at some irrelevant porno site.
You can dial 911 from Sprint cell phones, but you usually get a recording that says "all operators are busy". If you wait long enough, you get some operator service somewhere, but they're oriented towards highway emergencies.
(i) providing network-level filtering to prevent receipt of material harmful to minors; or
(ii) providing at the time of a consumer's request under Subsection (1), software for contemporaneous installation on the consumer's computer that blocks, in an easy-to-enable and commercially reasonable manner, receipt of material harmful to minors.
(b) (i) Except as provided in Subsection (3)(b)(ii), a service provider may not charge a consumer for blocking material or providing software under this section, except that a service provider may increase the cost to all subscribers to the service provider's services to recover the cost of complying with this section.
So bundling "NetNanny" with ISP service, for those who want it, is sufficient to comply.
If you're in Utah, expect your ISP bill to go up by something under a dollar per month, based on bulk pricing for NetNanny.
(Does entering "~frontdoor" as the password still turn off NetNanny?)
Godot didn't show, either.
I'm going to go out and wait with Vlad and Estragon for a while, then try later.
You can watch new items come in on the Ourmedia main page, but you can't see them, because the update process is so slow. I submitted item #902, and half an hour later, the latest one is #907. So the load isn't high. Is the bottleneck a batch job, or does a real person have to look at this stuff before it goes up?
Tried again later. No response from the ourmedia.org server at all.
Tried again later. Back to the "enter password" dialog.
Good concept. Maybe someday it will work.
I read the IBM article. Sounds like the early days of SpamCop. SpamCop traces headers back to the originator or the first phony header, to validate the source. Mail with tracing problems used to get a challenge from SpamCop, but they gave up on that. Challenge-response effectively does a denial of service attack on joe-job victims. It's also incompatible with too many legitimate autoresponder systems that send mail confirmations of transactions.
And the relevant help page is a dead link.
Good concept, needs work.
Plogging is plugging your blog. Like Roland Piquwhatever, the guy who spams for his blog on Slashdot.