The title was a little misleading. I was under the impression a spammer had actually filed suit against someone whose abuse@ letter got their service cancelled. Still I'm glad I read. I've had an AOL'er claiming to be an AOL Hometown admin send me an unsolicited legal threat spam, but I still think this Shi*man takes the cake.
Just finished reading the SFGate article on the subject. What particularly struck my interest was the interview with Robert Regular--the name sounded familiar as I got into it with this very same marketing stiff last year, when his company's (Conducent Technologies at that time) TSADBOT spyware somehow got onto my system. (I must admit, as the webmaster of a semi-popular spyware information site, having one go undetected on my own system for nearly a month was rather embarassing.) At any rate, Mr. Regular's answers to my "clueless user" inquiries--not letting on that I had already dissected Conducent's app with a fine-toothed hex editor--led me to almost suggest that he drop the spyware biz in favor of a more lucrative position speechwriting for a certain ex-President.
Rather than redefining "is", it seems that our old friend has found a new home at Cydoor Technologies, makers of another KaZaA-transmitted disease, who are now pushing the ClickTilUWin trojan to spyware-friendly companies.
To quote the article:
Greg Bildson, chief technology officer of Lime Wire LLC, said the company was led to believe the program did no more than link to a game, making the permission request unnecessary.
Robert Regular of Cydoor Technologies Inc., which
distributed the ClickTillUWin software to the file-sharing companies, said the program wasn't supposed to collect information until users activated it -- and had an opportunity to be notified and decline if they so choosed.
Regular said he did not believe deception was intended by any of the parties.
I haven't put on my cost-benefit cap for this, nor do I have any plans to get involved in any hardware decoder projects (these tend to be hairier than Bin Laden's back). But what the decision probably comes down to is the cost a fast, low-power, low-heat DSP chip fast enough to decode OGG vs. the cost of a chip that can decode the less cpu-intensive MP3 format + Fraunhofer royalties (per unit). Currently, Vorbis files take more (about twice as much, IIRC) processing to decode. There are cheap DSPs out there, but the pricing rapidly goes skyward as the MHz, MIPS and MACs/clock increase.
The fact that Vorbis is still an 'evolving standard', meaning that a player made now could be obsoleted (or at least require the user to reflash something) when a newer version of the codec is released, is probably also a deciding factor. Developers would have to hinge their product on the hope that nobody adds a new CPU-intensive enhancement to the codec before it is finalized. (I don't suppose they would, but then again, engineers who suppose also tend to get sacked.)
Using 2 mail accounts for 'noise cancellation'?
on
Crazy Stats on Spam
·
· Score: 1
Also somewhat off-topic, but a common approach to dealing with 'noise' in electrical systems is to place an extra, signalless wire near one carrying a desired signal so that the same noise appears on each, giving the extra wire a pure 'noise signal' that can be subtracted from the other channel (variations of this approach are used with UTP network cable). Can a similar tactic be used to filter spam by placing 2 addresses, one real and one a spam-catcher, on public Web pages where a contact address is required?
It's great to hear Korea's getting less spam...
on
Crazy Stats on Spam
·
· Score: 1
...but, have there yet been studies seeking to find out where the most spam comes from? My mail accounts have always had a minor problem (and one, a not-so-minor) with dot-com spam, but I've recently noticed an upswing in spam from.cn and.jp domains. (Incidentally, how DO you spell 'abuse@' in Kanji?)
Because often, you don't find out the link is irrelevant until after you click on it. There's no way to "un-click" a bad result with a deceptive title.
We colleceted a group together on campus to watch the shower...unfortunately, a *dense* fog came up that afternoon and grew steadily worse as the night wore on, so couldn't see squat. Same problem in Fort Wayne, apparently. We stayed up and watched John Carpenter's "The Fog" instead (natch), then raided Tom's Donuts at about 4:30a.
This is an about-face, no? The last time I used an MS Bookshelf product (BS95), it not only contained every dirty word imaginable, but there was a pronunciation button that would actually recite them aloud. In case any of the young'uns out there was hazy on the enunciation of "blowjob" and "motherfucker".
The point of the changelog censorship is not the ability to "chown riaa" in this (or any kernel), but that it is fixing a vulnerability in the old one where someone may be able to bypass "chown riaa". I think the idea is that publishing more details on the vulnerability could assist others in bypassing file permissions, theoretically making A.C. responsible for assisting circumvention where file permissions are used to protect copyrighted material.
It wasn't clear from the blurb, but the USA Act doesn't open loopholes to hacking at all. It proposes severe penalties for hacking where the combined damage exceeds $5000 USD. It can then be declared a terrorist act.
What the RIAA is trying to do is make an amendment to this bill exempting them from the Act. It means that they (and no one else) can hack where they think there is piracy going on, without facing the penalties anyone else would.
It seems to me, less people with the ability/skill/etc. to rip these CDs means fewer different encodings of the same mp3. In other words, a higher percentage of the available copies being identical. Think of the implications of this in today's multi-source, auto-searching, auto-resuming crop of mp3 swapping clients.
Is it just me, or does this sound like a rehash of interstitals with an easier-to-spell name?
Waaaay back in the day, some companies experimented with a type of ad called an interstitial. When a surfer clicked a link, the page it led to would consist of an ad and a "continue to next page" link. If no action was taken, the ad would loop for n seconds and refresh to the next page.
Web viewers hated it. They found it very annoying, instrusive and a waste of their time. The format died quietly in the midst of then-novel schemes like popups.
(Geocities, one of the earliest non-porno sites to use popups, tried interstitials first and found that people universally hated them. Of course, they'll go on to tell you that popups were met with "overwhelmingly positive feedback":) Go figure.)
Simple solution--just use your own DMCA-grade encrypyion on your circumvention device (Caesar Code comes to mind..."J dbo tvf zpv opx!") and it will be illegal for them to decrypt it to prove it is indeed a circumvention device!
How come it is that it is illegal to perform an in-memory algorithmic transformation X on Copyrighted Work A, but perfectly OK to perform algorithmic transformation Y on Copyrighted Work B?
Where A == DVD Data;
B == our Web sites;
X == deCSS transform;
Y == eZula/GATOR/Surf+ transform
Really, it seems that these adware parasites (Transform Engines) can't be made illegal without dragging other TEs (ad-filters, regexp'ers, translators, Swedish Chef-izers) down with them. But would they be commiting an illegal act by invalidating statements of guarantee that you make on your Web site, e.g. privacy policy? Not to mention the obvious effects on your (commercially-valuable?) goodwill when you state prominently that your site contains no advertising links, no popups, does not set cookies, gathers no information, will never link to doubleclick, etc... and a dozen X10 camera ads pop up.
I see a big whistling teakettle of legal hot water being put on the burners...
...but we need to come up with a new protocol specifically suited to wireless physical networks. TCP/IP is not the best suited to take advantage of a wireless grid, the ultimate of p2p networks.
Today's so-called "p2p" network is not P2P at all. It's more along the lines of:
Person to ISP to Upstream Provider to Upstream Provider [etc.] to Backbone to Downstream Provider [etc.] to ISP to Person. This provides numerous places between one user and another where the data stream can be hijacked or cut off for whatever reason--numerous "choke points" where an unwholesome third party can leverage control over a user's service via threats, lawyers, technical measures, fiber-seeking backhoe, et cetera.
/* Beware -- switches to Long Rant mode */
Problems with TCP/IP:
The majority of users are still using single, slow, intermittent links to an ISP--typically modems running at 53K or lower. These links don't usually last more than a few hours at most; they tie up the phones, and piss off the ISP when users stay on too long.
The redundancy originally intended for the Internet is lost as each path from the backbone to individual users forms a *tree* instead of a grid. As bandwidth is still a limited and metered commodity, each carrier/ISP/reseller/etc. linking end-users to the Internet typically buys "a line" (or at most "a few lines") from a single upstream provider, who in turn has "a line" to another upstream provider, and so on, up to the backbone. If one of these lines is saturated or broken, everyone below the break is cut off from everything on the other side! So much for redundancy.
Those on faster, fatter, more permanent connections are still limited by the tree problem, and quite often impeded by a NAT/firewall acting as the tree branch. This means that even users with "ideal" connections to other users with "ideal" connections are severely limited in what they can transfer, as incoming requests get filtered out or lost by the NAT. NATs are necessary because the pool of available IP addresses is limited, and blocks of addresses still have to be purchased from somewhere. Unfortunately the NAT is still the only way in and out of most high-speed local networks.
Traditional pseudo-P2P networks (Gnutella et al) face additional problems, above and beyond those of the TCP/IP Internet itself. Gnutellalike systems attempt to create a grid of peer-to-peer links, while still using the traditional tree-like Internet structure. It is entirely plausible that packets from a randomly-selected pair of next-door neighbours in Lowell, Indiana travel from Lowell to Gary to Chicago to San Francisco and all the way back (it would be far more efficient in this case, and probably faster, to just slip a diskette under the neighbour's door:). Bearing in mind a typical Gnutella packet propagates through up to (and above) 7 different TCP/IP users between its source and destination, that's an incredible amount of distance for the packet to needlessly travel.
Expanding on the earlier "not P2P" example, this packet goes from person -> ISP -> upstream -> backbone -> downstream -> ISP -> person -> ISP -> upstream -> backbone....you get the idea... -> downstream -> ISP -> person.
What's needed is a geographically-centric protocol which allows a computer system to establish multiple routes of varying speed, duration and quality, over a variety of media (twisted pair, coax, radio, microwave, infra-red, laser, fiber...), and maintain a fairly comprehensive picture of the "quality of service" of each link. /* switches to Mad Scientist Brainstorming mode */
The protocol would take advantage of the physical geography of the networked machines. Machines physically close together, as in an apartment building or even an entire subdivision, would be able to transfer data among themselves at incredibly high speeds (Ethernet, Wireless Ethernet), this becoming less and less the case as the physical distance between the machines increases, as in transfer between two distant subdivisions--interference and line noise, as well as the journey across multiple (possibly saturated) hubs and gateways, would increasingly constrict data flow as machines grew farther apart. This emphasizes the importance of choosing the "closest" of multiple available resources (multiplayer games, shared files) over more distant ones. The QoS information available by querying machines (or embedded in the packets themselves) would make this possible.
(Note: The speed drop between distant machines may be mitigated somewhat by the fact that there will probably be more possible (and roughly equidistant) routes between them. This of course does not help interactive transfers such as games one bit!)
A single any-order transmission (file transer, etc., as opposed to highly interactive processes where packets must come in sequence) could be split up and sent to its destination by several routes, the amount of data being sent through each depending on the QoS for that line. Each packet would be assigned a number. At points during the transfer, the receiving machine would report back which of the numbered packets have been received, so that any that have not shown up in a reasonable time can be retransmitted, preferably over a link that is not dropping so many.
The permanent or semi-permanent nature of most links/routes will permit "always on" connectivity (limited to what's on the wireless/etc. net), not currently possible for most (modem-afflicted) users.
Some form of Asymmetric encryption to persisitently identify a trusted person/server, secure packets against snooping by intermediate parties, and limit packet spoofing would likely be required (especially if the plan is to eventually run persistent WWW servers over this proto). (I'm not up to snuff on this, unfortunately; anyone else care to comment on possibility?)
...some unknown, evil little entity (me^H^HJane Morgandorfer) created a program specifically to hijack Gator's over-ads to display over-over-ads? How long do you suppose until Jane gets a nasty letter from Gator lawyers and have to hire Crocodile Dundee to protect herself?
I had this half-serious idea, not sure if it would be feasible on a global scale though. Currently, communities are setting up their own wireless "freenets"--essentially suburb-wide LANs--that bypass the corporate Internet entirely. Some even have links to the big bad wired 'net, too. How feasible would it be for one suburb's freenet to be linked to another's via wire/laser/microwave, and to another, etc., crisscrossing the country? Naturally, access speeds to other places would depend on their distance from your current location, but I'm sure something could be worked out. (Hint: Multi-routing would be helpful here)
I'm surprised no one brought this up yet (or maybe they did, and I didn't read closely enough, and deserve to be beat with a stick), but what's this bollocks about needing raw sockets in XP to be backwards compatible with 95,98,ME? I thought the big stink about XP (besides the possibility of millions of untraceable no-see-ums hammering your server) is that it will be the first mainstream OS to give built-in full raw socket access to any joe user and the programs he runs. How are they necessary for compatibility with Win9x programs, which never had access to them to beging with?
Slashdot Mirror
Nuts, I think I'm infected. I too am sickened by large screens of advertising...
The title was a little misleading. I was under the impression a spammer had actually filed suit against someone whose abuse@ letter got their service cancelled. Still I'm glad I read. I've had an AOL'er claiming to be an AOL Hometown admin send me an unsolicited legal threat spam, but I still think this Shi*man takes the cake.
Rather than redefining "is", it seems that our old friend has found a new home at Cydoor Technologies, makers of another KaZaA-transmitted disease, who are now pushing the ClickTilUWin trojan to spyware-friendly companies.
To quote the article:
Robert Regular of Cydoor Technologies Inc., which distributed the ClickTillUWin software to the file-sharing companies, said the program wasn't supposed to collect information until users activated it -- and had an opportunity to be notified and decline if they so choosed.
Regular said he did not believe deception was intended by any of the parties.
I guess some things never change.
The fact that Vorbis is still an 'evolving standard', meaning that a player made now could be obsoleted (or at least require the user to reflash something) when a newer version of the codec is released, is probably also a deciding factor. Developers would have to hinge their product on the hope that nobody adds a new CPU-intensive enhancement to the codec before it is finalized. (I don't suppose they would, but then again, engineers who suppose also tend to get sacked.)
Imagine a Beow...nevermind
Also somewhat off-topic, but a common approach to dealing with 'noise' in electrical systems is to place an extra, signalless wire near one carrying a desired signal so that the same noise appears on each, giving the extra wire a pure 'noise signal' that can be subtracted from the other channel (variations of this approach are used with UTP network cable). Can a similar tactic be used to filter spam by placing 2 addresses, one real and one a spam-catcher, on public Web pages where a contact address is required?
...but, have there yet been studies seeking to find out where the most spam comes from? My mail accounts have always had a minor problem (and one, a not-so-minor) with dot-com spam, but I've recently noticed an upswing in spam from .cn and .jp domains. (Incidentally, how DO you spell 'abuse@' in Kanji?)
Because often, you don't find out the link is irrelevant until after you click on it. There's no way to "un-click" a bad result with a deceptive title.
We colleceted a group together on campus to watch the shower...unfortunately, a *dense* fog came up that afternoon and grew steadily worse as the night wore on, so couldn't see squat. Same problem in Fort Wayne, apparently. We stayed up and watched John Carpenter's "The Fog" instead (natch), then raided Tom's Donuts at about 4:30a.
This is an about-face, no? The last time I used an MS Bookshelf product (BS95), it not only contained every dirty word imaginable, but there was a pronunciation button that would actually recite them aloud. In case any of the young'uns out there was hazy on the enunciation of "blowjob" and "motherfucker".
I'm surprised I didn't see this post already.
SafeSurf, an Internet blocking company, is complaining... about... censorship...
(Post ends abruptly as Mr. Obvious departs to address the burning smell eminating from his irony meter.)
The point of the changelog censorship is not the ability to "chown riaa" in this (or any kernel), but that it is fixing a vulnerability in the old one where someone may be able to bypass "chown riaa". I think the idea is that publishing more details on the vulnerability could assist others in bypassing file permissions, theoretically making A.C. responsible for assisting circumvention where file permissions are used to protect copyrighted material.
What the RIAA is trying to do is make an amendment to this bill exempting them from the Act. It means that they (and no one else) can hack where they think there is piracy going on, without facing the penalties anyone else would.
404...
It seems to me, less people with the ability/skill/etc. to rip these CDs means fewer different encodings of the same mp3. In other words, a higher percentage of the available copies being identical. Think of the implications of this in today's multi-source, auto-searching, auto-resuming crop of mp3 swapping clients.
Waaaay back in the day, some companies experimented with a type of ad called an interstitial. When a surfer clicked a link, the page it led to would consist of an ad and a "continue to next page" link. If no action was taken, the ad would loop for n seconds and refresh to the next page.
Web viewers hated it. They found it very annoying, instrusive and a waste of their time. The format died quietly in the midst of then-novel schemes like popups.
(Geocities, one of the earliest non-porno sites to use popups, tried interstitials first and found that people universally hated them. Of course, they'll go on to tell you that popups were met with "overwhelmingly positive feedback" :) Go figure.)
Who needs strong crypto anyway? We've got the DMCA...
Simple solution--just use your own DMCA-grade encrypyion on your circumvention device (Caesar Code comes to mind..."J dbo tvf zpv opx!") and it will be illegal for them to decrypt it to prove it is indeed a circumvention device!
Where A == DVD Data;
B == our Web sites;
X == deCSS transform;
Y == eZula/GATOR/Surf+ transform
Really, it seems that these adware parasites (Transform Engines) can't be made illegal without dragging other TEs (ad-filters, regexp'ers, translators, Swedish Chef-izers) down with them. But would they be commiting an illegal act by invalidating statements of guarantee that you make on your Web site, e.g. privacy policy? Not to mention the obvious effects on your (commercially-valuable?) goodwill when you state prominently that your site contains no advertising links, no popups, does not set cookies, gathers no information, will never link to doubleclick, etc... and a dozen X10 camera ads pop up.
I see a big whistling teakettle of legal hot water being put on the burners...
Today's so-called "p2p" network is not P2P at all. It's more along the lines of: Person to ISP to Upstream Provider to Upstream Provider [etc.] to Backbone to Downstream Provider [etc.] to ISP to Person. This provides numerous places between one user and another where the data stream can be hijacked or cut off for whatever reason--numerous "choke points" where an unwholesome third party can leverage control over a user's service via threats, lawyers, technical measures, fiber-seeking backhoe, et cetera.
Problems with TCP/IP:
What's needed is a geographically-centric protocol which allows a computer system to establish multiple routes of varying speed, duration and quality, over a variety of media (twisted pair, coax, radio, microwave, infra-red, laser, fiber...), and maintain a fairly comprehensive picture of the "quality of service" of each link.
/* switches to Mad Scientist Brainstorming mode */
The protocol would take advantage of the physical geography of the networked machines. Machines physically close together, as in an apartment building or even an entire subdivision, would be able to transfer data among themselves at incredibly high speeds (Ethernet, Wireless Ethernet), this becoming less and less the case as the physical distance between the machines increases, as in transfer between two distant subdivisions--interference and line noise, as well as the journey across multiple (possibly saturated) hubs and gateways, would increasingly constrict data flow as machines grew farther apart. This emphasizes the importance of choosing the "closest" of multiple available resources (multiplayer games, shared files) over more distant ones. The QoS information available by querying machines (or embedded in the packets themselves) would make this possible.
(Note: The speed drop between distant machines may be mitigated somewhat by the fact that there will probably be more possible (and roughly equidistant) routes between them. This of course does not help interactive transfers such as games one bit!)
A single any-order transmission (file transer, etc., as opposed to highly interactive processes where packets must come in sequence) could be split up and sent to its destination by several routes, the amount of data being sent through each depending on the QoS for that line. Each packet would be assigned a number. At points during the transfer, the receiving machine would report back which of the numbered packets have been received, so that any that have not shown up in a reasonable time can be retransmitted, preferably over a link that is not dropping so many.
The permanent or semi-permanent nature of most links/routes will permit "always on" connectivity (limited to what's on the wireless/etc. net), not currently possible for most (modem-afflicted) users.
Some form of Asymmetric encryption to persisitently identify a trusted person/server, secure packets against snooping by intermediate parties, and limit packet spoofing would likely be required (especially if the plan is to eventually run persistent WWW servers over this proto). (I'm not up to snuff on this, unfortunately; anyone else care to comment on possibility?)
...some unknown, evil little entity (me^H^HJane Morgandorfer) created a program specifically to hijack Gator's over-ads to display over-over-ads? How long do you suppose until Jane gets a nasty letter from Gator lawyers and have to hire Crocodile Dundee to protect herself?
I had this half-serious idea, not sure if it would be feasible on a global scale though. Currently, communities are setting up their own wireless "freenets"--essentially suburb-wide LANs--that bypass the corporate Internet entirely. Some even have links to the big bad wired 'net, too. How feasible would it be for one suburb's freenet to be linked to another's via wire/laser/microwave, and to another, etc., crisscrossing the country? Naturally, access speeds to other places would depend on their distance from your current location, but I'm sure something could be worked out. (Hint: Multi-routing would be helpful here)
/. does get slashdotted, or at least amazingly slow.
I'm surprised no one brought this up yet (or maybe they did, and I didn't read closely enough, and deserve to be beat with a stick), but what's this bollocks about needing raw sockets in XP to be backwards compatible with 95,98,ME? I thought the big stink about XP (besides the possibility of millions of untraceable no-see-ums hammering your server) is that it will be the first mainstream OS to give built-in full raw socket access to any joe user and the programs he runs. How are they necessary for compatibility with Win9x programs, which never had access to them to beging with?