Did anyone notice that the articles on their site are formated really similarly to slashdot's articles? Replies are missing, but the headers and the ownership comment at the bottom are basically the same.
Unlike Microsoft Word, being charged one cent every time you opened a Google application (word processor or whatever) would not be such a bad deal. After all, it wouldn't crash every other minute.
> But seriously... Too much of anything is not a good thing. Too much sex, you'd either be bored or exhausted, then the high would go away, you'd get depressed and take the plasma rifle to your head.
Sounds like someone not speaking from experience... what site is this again?
Maybe not (my calling card was always $0.09/minute at night) but Sprint is charging $0.40/minute if you go over your cell phone minutes now, unless you get that silly pay as you go plan. When I first signed up, it was only $0.12/minute, but of course I had to cancel my service for six months when they were being especially retarded. Sigh.
All these posts about brute force and "multiple possible matches" are missing the point, and didn't RTFA. I've done research with Bloom filters.
The fact that they give false positives (aka multiple possible matches) means that you can take all billion or so VALID email addresses which actually accept emails and have a live human being reading them, test every single one of them, and have say, 10 million matches in the Bloom filter. That's hardly cracking the scheme. Sure, you can start using outside knowledge, e.g. billg@microsoft.com is probably not really on my buddy list, but that's a different matter, and hardly a crack of the scheme.
That said, Bloom filters are very neat, older than most slashdot posters, and would be good dupe detectors. (Just kidding on the last one, editor mispellings would confuse the hash functions.)
To make this as easy as possible for insurance company representatives (or any other representatives of big business and government) to understand: Stay the f**k out of my life.
Is now a good time to say, "Who is forcing you to use these devices? This plan doesn't even force to report results after you check them."
Ignoring slippery slopes for a moment, the insurance company is trying to "prove" that you aren't a problem case waiting to happen. And why wouldn't you want to let the crazier drivers pay for the risk? Do you really like subsidizing their rates? Of course, reading the posts in other articles here, slashdot readers are pretty crazy drivers (e.g. passing at 100mph driving on the wrong side of the street) so maybe I'll be mobbed in a minute.
Also, they aren't trying to collect much information at the moment, but I imagine it would be a lot harder to justify the increased benefits of full tracking logs vs just speed logs. That, and for the non-tinfoil crowd, the detail to really recreate an accident would probably take way to much storage unless it was only the most recent data. The tinfoil crowd isn't reading this anyway.
P.S. For those who worry about it being sub-poenaed and self-incrimination, I agree it shouldn't be but it probably will. I still don't feel sorry for those who actually cause accidents by being deliberately reckless time and time again and try to hide it though.
So if they're getting laid for their work, they're complaining about license rights on the video produced?
digitizing consciousness
on
Broken Angels
·
· Score: 1
I've only read Altered Carbon so far, but he did talk about some of the questions with digitizing the human mind. For example, the Catholics were very much against being brought back in another sleeve if your original body died. Also, Takeshi encountered an assassin which illegally duplicated himself into two different sleeves. (The assassin didn't trust anyone but "himself" as a partner. The second sleeve was a woman though, a theme which came up later during a virtual torture scene.)
So why doesn't someone make a DNSBL taking names of the form ip.ip.ip.ip.fromdomain.myfunkydnsbl.net? If I recall correctly, it isn't too hard to add such a rule to sendmail.
If you were ever a sendmail admin before they added the most recent set of macros, its probably the same amount of effort. Basically, you were saying "do a check on $4.$3.$2.$1.rbl.net", so if you can do the same check adding from domain, then you could leverage most of the same code.
Wouldn't you love for your girlfriend (mythical or otherwise) to pose naked for this?
"They can take a twelve-inch piece and in less than a week we'll have it twelve feet," said Jon Lash, director of the institute's Digital Atelier division. "Then they'll come in and do the final manipulation of the surface."
and then again, you could pose and give her something to keep her company while you are away... perhaps with some editing, but you don't want her to prefer it to the real thing, do you?
So, after RTFA,we can secure our network by checking for viruses and making sure wireless chips are off? Are they forcing everyone to install special software to check these things? Would all the conventionners really want to deal with that hassle? And who make sures the software provider is really trusted and hasn't been compromised themselves?
Shouldn't the main thrust of the security efforts to protect computers from each other be off of those computers? There are plenty of other ways to compromise computers them without a virus or WiFi, same as always, and they won't necessarily show up on a virus scan. Isolating each machine from the others will do a ton more than running a program on what you think is the actual machine. Cisco can surely do a good job at making sure they can only reach the outside world and not their "LAN" neighbors. This is not to say that the conventionners will all be running VMWare or Virtual PC, but if their machine was already compromised, do you really trust the results of a program running on their machine? It's not like this a big geek-fest with a LAN party every night.
MBTA != Metro Boston Transit Authority
on
1984 Comes To Boston
·
· Score: 3, Informative
Good point... The fingerprinting service already has a legit use, as proposed by Shawn Fanning, so it's ok, but may not be separately available in his setup. A program for scanning through an audio stream an identifying what songs played when would be good for auditing purposes, those it wouldn't really need the same ~second accuracy. Maybe an indie band/label could use it to verify that a station was paying them the right amount though. Saving and splitting the audio stream up already have legitimate uses, though the format used might have issues. Of course, the little wrapper that combines them all would definitely be a candidate for lawsuits about "contributory copyright infringement".
So what we really need is for an indie label to provide such a program so you can split up a webcast of just their stuff. That way, they get paid when you first listen to the webcast, and you get to keep mp3s or oggs and listen to them later and skip the pay-per-play part.
This brings to mind a question about audio fingerprinting... Clearly, the fingerprints are meant (regardless of reality) to be robust to small changes to the content that don't affect the quality substantially. Otherwise, ripping or re-encoding them destroys the fingerprints. Are any of them capable of matching a different recording of the same song by the same band without a high false positive rate? The other recording could just be from a different studio session (for a different CD) or maybe from a live performance (which would probably harder). Here, even an answer of "looks like this song by this band but definitely not the same recording" would be useful.
I haven't read the actual text of the act in a while, but these do give the gist of it... Basically, you're not supposed to let the user pick content or tell them when their favorite stuff is coming up, or make it easier for them to buffer the last few hours and pick out the stuff they like. However, what if you had an audio fingerprint database, like the one Shawn Fanning is working on. It shouldn't be too hard to automatically divide an audio stream into separate songs (or a moderate number of candidate intervals for songs) and then you could do lookups in the fingerprint database to divide the stream into actual labeled songs. From there, the rest is easy:)
they said the functionality of the "computer system". maybe "delivery" is working, but the "system" isn't. certainly the system's users are seeing issues with its performance since they have to deal with all the spam that shouldn't have reached them (and I don't mean "shouldn't" in an anal SMTP RFC way).
(FWIW, IANAL) Web site content is copyrighted. Therefore, you have a right to make your own personal copy, and backup copies, but it is not legal to redistribute those copies without the site owner's permission. I cannot imagine that the Wayback machine or the Google cache is legal. They are blatantly disregarding the site owners' copyright.
This confuses fair use on purchased items that you own with what you are allowed to do with temporary copies for viewing. By the same logic, you could legally take a video camera to see Terminator 3 in a couple days and make a copy for personal use to watch at home later all by yourself (let's pretend you don't have friends over to watch it too or post it on the internet afterwards). And you could save it "just in case" the MPAA lost every single copy they had.
However, the latest HTTP protocol has provisions for caching, and headers for controlling it, so if you get a file with headers allowing caching, then you should be able to cache it. I don't remember if HTTP 1.0 had them, but HTTP 1.1 certainly does. If you get the file through HTTP 1.1 and they don't include the headers to prevent caching, then the protocol with which they chose to provide you the file is stating that you can cache it. So, if you play by those rules, you should be able to check if you can cache content legally. If they don't want you to cache, then they can say so when they send it to you. (I'm ignoring the messy rules regarding stale content for simplicity.)
IANAL, but could one argue that going for high profile traders was selective according to some other criteria? For example, if they are downloading everything, it's probably (very) slightly less likely that they're really rich but they probably aren't too poor if they have a good connection at home.
Slashdot Mirror
If a 256GB SSD drive is only 256 million bytes, I'm a lot more concerned about the missing factor of ~1000 than the binary/decimal notation.
Did anyone notice that the articles on their site are formated really similarly to slashdot's articles? Replies are missing, but the headers and the ownership comment at the bottom are basically the same.
Right at the bottom (links stripped):
Hot Products at Yahoo! Shopping:
Apple iBook G4 Laptop Computer
Windows XP
Mac OS
Unlike Microsoft Word, being charged one cent every time you opened a Google application (word processor or whatever) would not be such a bad deal. After all, it wouldn't crash every other minute.
> But seriously... Too much of anything is not a good thing. Too much sex, you'd either be bored or exhausted, then the high would go away, you'd get depressed and take the plasma rifle to your head.
Sounds like someone not speaking from experience... what site is this again?
Maybe not (my calling card was always $0.09/minute at night) but Sprint is charging $0.40/minute if you go over your cell phone minutes now, unless you get that silly pay as you go plan. When I first signed up, it was only $0.12/minute, but of course I had to cancel my service for six months when they were being especially retarded. Sigh.
All these posts about brute force and "multiple possible matches" are missing the point, and didn't RTFA. I've done research with Bloom filters.
The fact that they give false positives (aka multiple possible matches) means that you can take all billion or so VALID email addresses which actually accept emails and have a live human being reading them, test every single one of them, and have say, 10 million matches in the Bloom filter. That's hardly cracking the scheme. Sure, you can start using outside knowledge, e.g. billg@microsoft.com is probably not really on my buddy list, but that's a different matter, and hardly a crack of the scheme.
That said, Bloom filters are very neat, older than most slashdot posters, and would be good dupe detectors. (Just kidding on the last one, editor mispellings would confuse the hash functions.)
To make this as easy as possible for insurance company representatives (or any other representatives of big business and government) to understand: Stay the f**k out of my life.
Is now a good time to say, "Who is forcing you to use these devices? This plan doesn't even force to report results after you check them."
Ignoring slippery slopes for a moment, the insurance company is trying to "prove" that you aren't a problem case waiting to happen. And why wouldn't you want to let the crazier drivers pay for the risk? Do you really like subsidizing their rates? Of course, reading the posts in other articles here, slashdot readers are pretty crazy drivers (e.g. passing at 100mph driving on the wrong side of the street) so maybe I'll be mobbed in a minute.
Also, they aren't trying to collect much information at the moment, but I imagine it would be a lot harder to justify the increased benefits of full tracking logs vs just speed logs. That, and for the non-tinfoil crowd, the detail to really recreate an accident would probably take way to much storage unless it was only the most recent data. The tinfoil crowd isn't reading this anyway.
P.S. For those who worry about it being sub-poenaed and self-incrimination, I agree it shouldn't be but it probably will. I still don't feel sorry for those who actually cause accidents by being deliberately reckless time and time again and try to hide it though.
So if they're getting laid for their work, they're complaining about license rights on the video produced?
I've only read Altered Carbon so far, but he did talk about some of the questions with digitizing the human mind. For example, the Catholics were very much against being brought back in another sleeve if your original body died. Also, Takeshi encountered an assassin which illegally duplicated himself into two different sleeves. (The assassin didn't trust anyone but "himself" as a partner. The second sleeve was a woman though, a theme which came up later during a virtual torture scene.)
Not having played Street Fighter since 2 was good, I knew nothing about parrying and read this as "the video of the amazing editing"...
I suppose that means that the submitter wasn't copying and pasting from another source (for a change).
So why doesn't someone make a DNSBL taking names of the form ip.ip.ip.ip.fromdomain.myfunkydnsbl.net? If I recall correctly, it isn't too hard to add such a rule to sendmail.
If you were ever a sendmail admin before they added the most recent set of macros, its probably the same amount of effort. Basically, you were saying "do a check on $4.$3.$2.$1.rbl.net", so if you can do the same check adding from domain, then you could leverage most of the same code.
Wouldn't you love for your girlfriend (mythical or otherwise) to pose naked for this?
and then again, you could pose and give her something to keep her company while you are away... perhaps with some editing, but you don't want her to prefer it to the real thing, do you?
So, after RTFA,we can secure our network by checking for viruses and making sure wireless chips are off? Are they forcing everyone to install special software to check these things? Would all the conventionners really want to deal with that hassle? And who make sures the software provider is really trusted and hasn't been compromised themselves?
Shouldn't the main thrust of the security efforts to protect computers from each other be off of those computers? There are plenty of other ways to compromise computers them without a virus or WiFi, same as always, and they won't necessarily show up on a virus scan. Isolating each machine from the others will do a ton more than running a program on what you think is the actual machine. Cisco can surely do a good job at making sure they can only reach the outside world and not their "LAN" neighbors. This is not to say that the conventionners will all be running VMWare or Virtual PC, but if their machine was already compromised, do you really trust the results of a program running on their machine? It's not like this a big geek-fest with a LAN party every night.
Actually, it's Massachusetts Bay Transit Authority. Almost noone gets it right, even native Bostonians...
so can someone look up their corporate charter and have them broken up for violating it? sadly, it is probably to broad for that
So, for everyone complaining about changes happening without you knowing, would you rather it be Microsoft or some virus writer making these changes?
heck, it could be a guy with a confused heterosexual boyfriend
Don't you mean everyone with a Kazaa username starting with a letter greater than B?
Good point... The fingerprinting service already has a legit use, as proposed by Shawn Fanning, so it's ok, but may not be separately available in his setup. A program for scanning through an audio stream an identifying what songs played when would be good for auditing purposes, those it wouldn't really need the same ~second accuracy. Maybe an indie band/label could use it to verify that a station was paying them the right amount though. Saving and splitting the audio stream up already have legitimate uses, though the format used might have issues. Of course, the little wrapper that combines them all would definitely be a candidate for lawsuits about "contributory copyright infringement".
So what we really need is for an indie label to provide such a program so you can split up a webcast of just their stuff. That way, they get paid when you first listen to the webcast, and you get to keep mp3s or oggs and listen to them later and skip the pay-per-play part.
This brings to mind a question about audio fingerprinting... Clearly, the fingerprints are meant (regardless of reality) to be robust to small changes to the content that don't affect the quality substantially. Otherwise, ripping or re-encoding them destroys the fingerprints. Are any of them capable of matching a different recording of the same song by the same band without a high false positive rate? The other recording could just be from a different studio session (for a different CD) or maybe from a live performance (which would probably harder). Here, even an answer of "looks like this song by this band but definitely not the same recording" would be useful.
I haven't read the actual text of the act in a while, but these do give the gist of it... Basically, you're not supposed to let the user pick content or tell them when their favorite stuff is coming up, or make it easier for them to buffer the last few hours and pick out the stuff they like. However, what if you had an audio fingerprint database, like the one Shawn Fanning is working on. It shouldn't be too hard to automatically divide an audio stream into separate songs (or a moderate number of candidate intervals for songs) and then you could do lookups in the fingerprint database to divide the stream into actual labeled songs. From there, the rest is easy :)
they said the functionality of the "computer system". maybe "delivery" is working, but the "system" isn't. certainly the system's users are seeing issues with its performance since they have to deal with all the spam that shouldn't have reached them (and I don't mean "shouldn't" in an anal SMTP RFC way).
(FWIW, IANAL) Web site content is copyrighted. Therefore, you have a right to make your own personal copy, and backup copies, but it is not legal to redistribute those copies without the site owner's permission. I cannot imagine that the Wayback machine or the Google cache is legal. They are blatantly disregarding the site owners' copyright.
This confuses fair use on purchased items that you own with what you are allowed to do with temporary copies for viewing. By the same logic, you could legally take a video camera to see Terminator 3 in a couple days and make a copy for personal use to watch at home later all by yourself (let's pretend you don't have friends over to watch it too or post it on the internet afterwards). And you could save it "just in case" the MPAA lost every single copy they had.
However, the latest HTTP protocol has provisions for caching, and headers for controlling it, so if you get a file with headers allowing caching, then you should be able to cache it. I don't remember if HTTP 1.0 had them, but HTTP 1.1 certainly does. If you get the file through HTTP 1.1 and they don't include the headers to prevent caching, then the protocol with which they chose to provide you the file is stating that you can cache it. So, if you play by those rules, you should be able to check if you can cache content legally. If they don't want you to cache, then they can say so when they send it to you. (I'm ignoring the messy rules regarding stale content for simplicity.)
IANAL, but could one argue that going for high profile traders was selective according to some other criteria? For example, if they are downloading everything, it's probably (very) slightly less likely that they're really rich but they probably aren't too poor if they have a good connection at home.