What's likewise crazy about online fraud to me is the following scenario.
As an online merchant, we see online orders that are clearly fraudulent. But the credit card still goes through (we 'authorize' first which just deducts from your credit limit). We decide not to take the order; thus we don't do a 'capture' on the card that would deduct the money from the poor guy's credit card account. That way we avoid getting charge-backs that would ruin our merchant rating and that would cost us in the end anyway (if caught). But we do log that credit card # in our database. Sometimes SIX MONTHS LATER the fraudster will use that same credit card # on our site again and it is *still* being accepted by Visa/Mastercard!
This is a broken system. As a merchant, we have no way (that I know of) to warn Visa/MasterCard or the issuing bank or the card holder that the number is being used for fraud! (Besides just going ahead and charging the card, knowing its fraud.) Certainly not an automated way to do so in the same way that we connect to payment gateways. It's just not in Visa's/Mastercard's interest to put a system in place because at the end of the day, the merchant is liable.
I'm interested if anyone knows of a place where merchants can swap info about fraudulent cards or other fraud data.
I use tcpdump UNIX-side, and Ethereal Windows-side. Personally? I find Ethereal hard to use, but it gets the job done. I've traced down bugs in OpenBSD TCP stacks with it on my production servers. I've tried half a dozen other packages but they didn't add enough value to make them worth trying to hit my boss up for cash.
To install Ethereal, you will need to download and install the low-level WinPcap driver.
And you may find the Ethereal packet analysis plug-in Packetyzer helpful; sometimes reading raw logs gets a bit annoying.
My supposition is this. Sun had just proved that it could hound/"assist" the global legal system into fining Microsoft 600 million.
The $2 billion valuation figure for leaving Microsoft alone wasn't arrived at until it was clear what financial penalties Sun could (indirectly) cause to Microsoft if they persisted in pursuing them legally.
By agreeing to shell out $2 billion, Microsoft is pragmatically admitting that it would be subject to at least that many fines going forward if Sun kept pursuing the matter over the next decade.
(Microsoft *did* eviscerate the Java platform by tying IE to windows and trying to change the behavior of Java base clases rather than just adding easily recognized com.ms.* classes as its original contract clearly encouraged. All in all, a $2 billion settlement to kill off the biggest platform competitor to threaten them in a decade isn't *that* bad for someone of MS's size.)
On this subject, I always liked Max Tegmark's speculations on the topic, which includes some assessment of why we have only one, and not zero or two or more temporal dimensions.
There's lots of other cool stuff on Max Tegmark's site too if you want to procrastinate on whatever else you're doing. (He's a physics (astrophysics?) professor at U.Penn.)
Color me cynical but I think greater precision in terms of bits of color (beyond the 8-12 bits per channel level) these days is a marketing game designed to sell product. It's like getting 300 fps at Quake. Particularly once you start saying 32-bits per channel is not enough, you need 64-bits per channel. It's there because the tech is there, not because the demand for that result is there.
The real problem is that our output devices are markedly inferior to reality (in terms of brightness, not to mention the color gamut issues which you alude to), and our lighting algorithms still leave a lot to be desired. It's much easier economically to ride Moore's Law to achieve some minimal degree of color precision improvement than to fix those.
I don't follow anime that closely nor Japanese cinema at all, but I watched the trailer and the eye candy looked pretty good. I have a bunch of questions:
1) Will there be a release in English?
2) If so, will it have subtitles or overdubbing?
3) Where was the movie shot?
4) The actors and target audience are clearly Japanese. Was the rest of the production (directing, crew, special effects) all done by Japanese?
5) Most of the writing in the trailer was Japanese (no surprise), but the title (CASSHERN) was in English. Is there any significance to this? Is this a common practice? Does the movie have both a Japanese and an English title? (I couldn't tell on the Japanese side.)
6) The production values and trailer looked pretty darn good (and I didn't even understand what was going on...) Is there any particular reason why Japanese movies don't get released in the US the way US movies get released in Japan? Don't most US movies get released in Japanese when shown in Japan?
I've never had a story accepted either, and on a number of occassions I've submitted stories hours, days or weeks before the topic appeared on Slashdot. It's pretty common; I wouldn't make anything out of it. It's quite possible that someone submitted the story before you did even earlier in the morning and the editors put that one in the queue to go up at 2:43PM. They pre-scheduled the various stories that go up hours (and sometimes even days?) in advance. Or perhaps they decided it was a worthy story after they saw the 27th submission of it.
I realized one day that we could essentially have a user-contributed, user-moderated article queue of sorts using the journaling system here. I've dedicated my journal to it. I haven't figured out how to draw larger traffic to it without making this a part-time job, but you're welcome to contribute to it and I welcome suggestions.
I'm pretty sure the term avatar (for VR) predates 1990.
My first memory of the term "Avatar" being used to represent an online persona was on the online service Q-Link aka Quantum Link, a nationwide BBS system for the popular Commodore 64. (The parent company later became AOL.) They had a 2D graphics chat world called "Club Caribe" which I remember using the term "Avatar". (At the time, I thought it was a bit odd, since I was used to the term Avatar being used for the main character of Ultima IV (1985).) This would have been around 1988-1989 or so, which is earlier than the OED citation, although I do not have a printed source backup for this. (Check a C-64 magazine of that time period? Old copies of Compute Gazette, anyone?)
I've found a post from a MUD-Dev mailing list discussion thread held in 2001 on the same topic (what's the earliest use of the term avatar) that supports this recollection, and adds to it that the term might have been used by the predecessor of Club Caribe, Lucasfilm's Habitat (1984-1988), or possibly even earlier by Jaron Lanier. Again, no paper-based backup on this.
Regarding the term "morph", 1993 doesn't sound too far off; it might be a year or two earlier though. I ran across the term in late 1993 when trying to replicate the morphing process used by Michael Jackson's "Black or White" music video for a computer graphics class (based on a white paper by Pacific Data Images). Both that video and Terminator 2: Judgement Day which used morphing came out in 1992. The CG morphing technique was known as morphing when I took the class in 1993. I'm not sure the PDI white paper used the term morphing though, so maybe the term's name caught on some time after the video came out. So it might be 1993, but I wouldn't be surprised if the term was used in 1992.
My question is this: Has it been decided by any courts or legislation whether domain names are property, or merely a contractual service? Particularly as this pertains to bankruptcy law?
I'm interested in a domain name, but it seems to be in legal limbo and I'd be interested in understanding the principles involved.
Backround: The company that holds the domain name filed for bankruptcy about 2 months after domain name was first requested. This was back in 1998. There is no evidence that the company ever used the domain name (e.g. internet archive, trademark registration, etc). The company was trying to restructure and then filed to liquidate back in 2000.
I looked up the court docket for the bankruptcy case but it looks like there have been no judicial or trustee actions for ~3 years.
However, the domain name, which appears to be in some sort of Registrar Lock and also a Legal Lock at Network Solutions, appears to be renewing at 1-year intervals somehow.
There are two other sub-questions that this scenario raises that I presume you don't have time to answer but I'll mention them in case someone else knows: 1) Does a bankruptcy trustee typically have the ability to renew contracts such as that for a domain name? 2) What sort of court papers would signal to a registrar that the entity registering the name has been bankrupt and the name should be removed from whatever legal lock and auto-renewal process is in place? (We've tried faxing stuff to their legal department but it seems to go into a black hole and we don't really know what should be sent.)
Color me cynical, but I suspect we won't get true broadband (10Mbs to 100Mbs) to the home any time soon (by 2010 years, for under $50/month, in any reasonable US geographic region) for the following reasons:
1) The cable guys don't want to cannibalize or lose control of the distribution channel for TV/HDTV video which requires such bandwidths.
2) The telco guys don't want to cannibalize their business T1 sales.
3) The satellite guys can't provide that bandwidth on a bi-directional, many-to-many basis.
4) The wireless phone guys may get there someday, but it'll take a while to improve their network bandwidth 1000x to do this.
I tried creating very small (~16x16?) GIF icons 4 years ago using Paint Shop Pro (the 30-day trial version) and I noticed that whenever I saved an image, it kept adding some sort of watermark to the image, shifting the color of a handful of non-adjacent pixels within what had previously been a solid band of color to a slightly different color in a way that was barely noticeable to the eye, but very noticable to me when trying to hand-edit the GIFs while zoomed in.
I kept trying to change the pixels back and re-save the image, and whenever I saved the image, the mysterious watermark pixels would re-appear.
I think I switched to something more primitive like MS Paint (eep) to workaround the problem.
I wrote a couple haikus, tweaked them a bit, and then figured I'd look at the options again before submitting them. Ah well, too late. I'll submit them to you the Slashdot readership instead. Maybe I can get a Tshirt from CmdrTaco or another 1000 ad-free page-views? Lord knows I don't need the karma..
.
As words from my lips,
so Perl flies from my fingers -
Ecstatic coding! .
.
From God to Larry,
"Be fruitiful and multiply!"
Perl coders went forth.
AT LAST! The secret to beating Solitaire... This could perhaps be the most significant event of our times!
Shh, don't tell anybody but the solutions for all 32,000 Microsoft FreeCell deals (except for number 11982 which can't be won) have already been posted!!!
I feel pathetic for knowing this.
At least I didn't try playing them all myself, right?
I'm aware of Veritas and it's fair to mention it. I tried to avoid complicating my argument by explicitly bringing it up, since it's not relevant.
What I meant by the sentence I said, "If any version of AT&T Unix/Unixware that shipped to people like IBM included journal filesystems or volume management"... was that the AT&T System V Release 3/4/5 licensees (like IBM) didn't get (journalling/volume management/etc) as part of their SVR license.
I should have said "SVR licensees" rather than "people like IBM." (In that sense "every version of UnixWare has " *not* "shipped with Veritas," right?) Anyway, I recognize what you mean, and I hope you recognize what I meant, at least now that I've clarified it.
Related to this, think about what SCO is claiming... if IBM's JFS is a derivative work, then so is Veritas... and if SCO owns or controls the Veritas copyright, why is SCO paying Veritas for the privelege?
What's so ironic about the suit to me is that when IBM entered the Unix market with AIX in 1990, they were the *first* Unix vendor to introduce their own journal filesystems and volume management tools integrated into the OS.
These now-common approaches to improving the reliability and flexibility of Unix were part of IBM's value-add to Unix... a bit of heritage from their mainframe and minicomputer perspective. It wasn't enough in the marketplace to overcome IBM's late-ness to Unix and the odd uniquenesses of its registry-based configuration, but it did help somewhat in enterprise environments.
Anyway, after 10 years in the Unix market, IBM decides that having had minimal/modest success in the commercial Unix marketplace, perhaps they would have better luck in the free Unix marketplace (making money selling services,) particularly if they can catch this wave early rather than spending a decade worrying about cannibalization of their own product line. So they take the AIX 'crown jewels' and share them with the free Unix community.
And SCO claims that they are derivatives of SCO's original Unix work?!
If any version of AT&T Unix/Unixware that shipped to people like IBM included journal filesystems or volume management or NUMA SMP, then maybe I could buy it, but given the dictionary definitions of "derived" I just can't.
(dictionary.com entry for "derive": "to obtain or receive from a source". "derivative": "copied or adapted from others")
This is not a new story, although the mention of specific names (and the Mitterand connection) involved is new to me. I read this story 3-4 years ago in the book Victory: The Reagan Administration's Secret Strategy That Hastened the Collapse of the Soviet Union by Peter Schweizer, which is fairly extensively footnoted and on the record. The interviews with the Reagan officials may of course be self-serving, but it seems reasonably well sourced. The tech details, like this Safire piece, are not that informative, but I wouldn't expect them to be really. Used hardback copies are $2 on Amazon.
Right, but if Yahoo/etc serves up the image via SSL, while they could be getting a forged referrer, they should have a valid IP # that they could then block or prosecute.
Unless the porn purveyor had compromised zombie PCs making the SSL requests and relaying the image to the porn webserver over IRC or something.
The other problem with captcha-relaying is that if your captcha's have a distinct look about them, it's easy to tell if some porn site is using yours. So...
Think I could convince Yahoo or Paypal to give me a job looking at porn in an attempt to find captchas that look just like theirs?
You're right. But. A) you're repeating what the editor already said, and B) you are overstating your case a bit for the following reasons:
In fairness, the poster on the blog was Cory Doctorow, who is a long time, well-known net-citizen and isn't exactly some random guy, although you may not know him. For a sample of his work, see this piece in Salon which mentions that he won the John W. Campbell Award for best new science fiction writer at the 2000 Hugo Awards. He's not a journalist, he's a blogger, but it's an interesting tidbit nonetheless...
And even if he was a random blogger, his credentials are much less important than the core concept he's disclosing: that someone seeking to generate email accounts (or open bank accounts or whatever) could have porn-seeking humans workaround the turing-ish test security measures. The story is less that someone is doing it, than that someone could be doing it. At least to me.
Plus this is a hacker-type story... I wouldn't expect Reuters, etc. to carry it first.
I actually was glad to see the Slashdot editor point out the "someone told me" caveat... it's a sign to me that the editors here are getting better. They're warning us about the weaknesses in the story, not just slapping stuff up here without a care.
If the image ...has been inlined from Yahoo or Hotmail... as the article says, couldn't Yahoo/etc have their image generation scripts setup dynamically to check the referrer (or should I say referer?;-)).
I seem to recall this approach being used by online comic strips trying to prevent inline linking from elsewhere...
The expansion of the Universe is an expansion of space itself.
That concept I understand. Your simple restatement of it though makes me wonder... has there been any thinking/investigation of whether the Plank constant could have (gradually) changed during some stage of the Big Bang and/or what the consequences of such a change would entail?
So you're saying we should invade and 'liberate' every country whose citizens oppose the government?
No. Just every country intent on gaining WMD while sitting on $20B/year of oil revenue that can easily be siphoned to enable them to get it in so many ways that we won't be able to notice it. Every country which had a grudge match against our country (applauding 9/11) and then went around giving payouts to suicide bomber families in other countries. Every country that we gave low-grade WMD to (!) and then actually showed they were dumb/dangerous enough to use them. Every country that invaded its neighbors without military provocation, etc. Hint: unlike Iran or North Korea or China or Libya or Syria, Saddam actually started two wars, against Iran and Kuwait. Every country willing to let 500,000+ of its citizens die rather than comply with cease-fire or UN terms. Every country whether the dictator taught his sons to practice torture so foreign governments would be nervous about his successors enough not to kill papa.
Power=interests*capability*will. Besides the base of economic capabilities Iraq's oil revenue gave Saddam (even when under UN sanction "containment"), Saddam had will in spades and that could not be ignored.
Trust me, it takes a great director not to f*ck up a great book.
With a good book, you do have more to work with... but also the stakes are higher, and your choices are more tested, more scrutinized. It's double or nothing. If you do well, you are more worthy. With a great book, it's quadruple or nothing.
In general, most Unixes and Linux (as you say) have adopted the LP64 model where longs/pointers are 64-bits and ints are 32 bits (some gory details here. (Cray's Unix is an exception; it's ILP64).
Windows OSes however have adopted the LLP64 model where ints and longs are 32-bits still, but long longs and pointers are 64-bits (gory Windows details here and here.)
Both 32-bit Windows and Unix traditionally used ILP32, so the porting characteristics moving to 64-bit code are slightly different across the two platforms.
Slashdot Mirror
What's likewise crazy about online fraud to me is the following scenario.
As an online merchant, we see online orders that are clearly fraudulent. But the credit card still goes through (we 'authorize' first which just deducts from your credit limit). We decide not to take the order; thus we don't do a 'capture' on the card that would deduct the money from the poor guy's credit card account. That way we avoid getting charge-backs that would ruin our merchant rating and that would cost us in the end anyway (if caught). But we do log that credit card # in our database. Sometimes SIX MONTHS LATER the fraudster will use that same credit card # on our site again and it is *still* being accepted by Visa/Mastercard!
This is a broken system. As a merchant, we have no way (that I know of) to warn Visa/MasterCard or the issuing bank or the card holder that the number is being used for fraud! (Besides just going ahead and charging the card, knowing its fraud.) Certainly not an automated way to do so in the same way that we connect to payment gateways. It's just not in Visa's/Mastercard's interest to put a system in place because at the end of the day, the merchant is liable.
I'm interested if anyone knows of a place where merchants can swap info about fraudulent cards or other fraud data.
--LP
I use tcpdump UNIX-side, and Ethereal Windows-side. Personally? I find Ethereal hard to use, but it gets the job done. I've traced down bugs in OpenBSD TCP stacks with it on my production servers. I've tried half a dozen other packages but they didn't add enough value to make them worth trying to hit my boss up for cash.
To install Ethereal, you will need to download and install the low-level WinPcap driver.
And you may find the Ethereal packet analysis plug-in Packetyzer helpful; sometimes reading raw logs gets a bit annoying.
--LP
I noticed the EU fine timing also.
My supposition is this. Sun had just proved that it could hound/"assist" the global legal system into fining Microsoft 600 million.
The $2 billion valuation figure for leaving Microsoft alone wasn't arrived at until it was clear what financial penalties Sun could (indirectly) cause to Microsoft if they persisted in pursuing them legally.
By agreeing to shell out $2 billion, Microsoft is pragmatically admitting that it would be subject to at least that many fines going forward if Sun kept pursuing the matter over the next decade.
(Microsoft *did* eviscerate the Java platform by tying IE to windows and trying to change the behavior of Java base clases rather than just adding easily recognized com.ms.* classes as its original contract clearly encouraged. All in all, a $2 billion settlement to kill off the biggest platform competitor to threaten them in a decade isn't *that* bad for someone of MS's size.)
--LP
On this subject, I always liked Max Tegmark's speculations on the topic, which includes some assessment of why we have only one, and not zero or two or more temporal dimensions.
There's lots of other cool stuff on Max Tegmark's site too if you want to procrastinate on whatever else you're doing. (He's a physics (astrophysics?) professor at U.Penn.)
--LP
Color me cynical but I think greater precision in terms of bits of color (beyond the 8-12 bits per channel level) these days is a marketing game designed to sell product. It's like getting 300 fps at Quake. Particularly once you start saying 32-bits per channel is not enough, you need 64-bits per channel. It's there because the tech is there, not because the demand for that result is there.
The real problem is that our output devices are markedly inferior to reality (in terms of brightness, not to mention the color gamut issues which you alude to), and our lighting algorithms still leave a lot to be desired. It's much easier economically to ride Moore's Law to achieve some minimal degree of color precision improvement than to fix those.
--LP
I don't follow anime that closely nor Japanese cinema at all, but I watched the trailer and the eye candy looked pretty good. I have a bunch of questions:
1) Will there be a release in English?
2) If so, will it have subtitles or overdubbing?
3) Where was the movie shot?
4) The actors and target audience are clearly Japanese. Was the rest of the production (directing, crew, special effects) all done by Japanese?
5) Most of the writing in the trailer was Japanese (no surprise), but the title (CASSHERN) was in English. Is there any significance to this? Is this a common practice? Does the movie have both a Japanese and an English title? (I couldn't tell on the Japanese side.)
6) The production values and trailer looked pretty darn good (and I didn't even understand what was going on...) Is there any particular reason why Japanese movies don't get released in the US the way US movies get released in Japan? Don't most US movies get released in Japanese when shown in Japan?
--LP
I've never had a story accepted either, and on a number of occassions I've submitted stories hours, days or weeks before the topic appeared on Slashdot. It's pretty common; I wouldn't make anything out of it. It's quite possible that someone submitted the story before you did even earlier in the morning and the editors put that one in the queue to go up at 2:43PM. They pre-scheduled the various stories that go up hours (and sometimes even days?) in advance. Or perhaps they decided it was a worthy story after they saw the 27th submission of it.
I realized one day that we could essentially have a user-contributed, user-moderated article queue of sorts using the journaling system here. I've dedicated my journal to it. I haven't figured out how to draw larger traffic to it without making this a part-time job, but you're welcome to contribute to it and I welcome suggestions.
--LP
I'm pretty sure the term avatar (for VR) predates 1990.
My first memory of the term "Avatar" being used to represent an online persona was on the online service Q-Link aka Quantum Link, a nationwide BBS system for the popular Commodore 64. (The parent company later became AOL.) They had a 2D graphics chat world called "Club Caribe" which I remember using the term "Avatar". (At the time, I thought it was a bit odd, since I was used to the term Avatar being used for the main character of Ultima IV (1985).) This would have been around 1988-1989 or so, which is earlier than the OED citation, although I do not have a printed source backup for this. (Check a C-64 magazine of that time period? Old copies of Compute Gazette, anyone?)
I've found a post from a MUD-Dev mailing list discussion thread held in 2001 on the same topic (what's the earliest use of the term avatar) that supports this recollection, and adds to it that the term might have been used by the predecessor of Club Caribe, Lucasfilm's Habitat (1984-1988), or possibly even earlier by Jaron Lanier. Again, no paper-based backup on this.
Regarding the term "morph", 1993 doesn't sound too far off; it might be a year or two earlier though. I ran across the term in late 1993 when trying to replicate the morphing process used by Michael Jackson's "Black or White" music video for a computer graphics class (based on a white paper by Pacific Data Images). Both that video and Terminator 2: Judgement Day which used morphing came out in 1992. The CG morphing technique was known as morphing when I took the class in 1993. I'm not sure the PDI white paper used the term morphing though, so maybe the term's name caught on some time after the video came out. So it might be 1993, but I wouldn't be surprised if the term was used in 1992.
--LP
My question is this: Has it been decided by any courts or legislation whether domain names are property, or merely a contractual service? Particularly as this pertains to bankruptcy law?
I'm interested in a domain name, but it seems to be in legal limbo and I'd be interested in understanding the principles involved.
Backround:
The company that holds the domain name filed for bankruptcy about 2 months after domain name was first requested. This was back in 1998. There is no evidence that the company ever used the domain name (e.g. internet archive, trademark registration, etc). The company was trying to restructure and then filed to liquidate back in 2000.
I looked up the court docket for the bankruptcy case but it looks like there have been no judicial or trustee actions for ~3 years.
However, the domain name, which appears to be in some sort of Registrar Lock and also a Legal Lock at Network Solutions, appears to be renewing at 1-year intervals somehow.
There are two other sub-questions that this scenario raises that I presume you don't have time to answer but I'll mention them in case someone else knows:
1) Does a bankruptcy trustee typically have the ability to renew contracts such as that for a domain name?
2) What sort of court papers would signal to a registrar that the entity registering the name has been bankrupt and the name should be removed from whatever legal lock and auto-renewal process is in place? (We've tried faxing stuff to their legal department but it seems to go into a black hole and we don't really know what should be sent.)
--LP
Color me cynical, but I suspect we won't get true broadband (10Mbs to 100Mbs) to the home any time soon (by 2010 years, for under $50/month, in any reasonable US geographic region) for the following reasons:
1) The cable guys don't want to cannibalize or lose control of the distribution channel for TV/HDTV video which requires such bandwidths.
2) The telco guys don't want to cannibalize their business T1 sales.
3) The satellite guys can't provide that bandwidth on a bi-directional, many-to-many basis.
4) The wireless phone guys may get there someday, but it'll take a while to improve their network bandwidth 1000x to do this.
--LP
I tried creating very small (~16x16?) GIF icons 4 years ago using Paint Shop Pro (the 30-day trial version) and I noticed that whenever I saved an image, it kept adding some sort of watermark to the image, shifting the color of a handful of non-adjacent pixels within what had previously been a solid band of color to a slightly different color in a way that was barely noticeable to the eye, but very noticable to me when trying to hand-edit the GIFs while zoomed in.
I kept trying to change the pixels back and re-save the image, and whenever I saved the image, the mysterious watermark pixels would re-appear.
I think I switched to something more primitive like MS Paint (eep) to workaround the problem.
--LP
As words from my lips,
.
so Perl flies from my fingers -
Ecstatic coding!
From God to Larry,
"Be fruitiful and multiply!"
Perl coders went forth.
--LP
AT LAST! The secret to beating Solitaire... This could perhaps be the most significant event of our times!
Shh, don't tell anybody but the solutions for all 32,000 Microsoft FreeCell deals (except for number 11982 which can't be won) have already been posted!!!
I feel pathetic for knowing this.
At least I didn't try playing them all myself, right?
--LP
I'm aware of Veritas and it's fair to mention it. I tried to avoid complicating my argument by explicitly bringing it up, since it's not relevant.
What I meant by the sentence I said, "If any version of AT&T Unix/Unixware that shipped to people like IBM included journal filesystems or volume management"... was that the AT&T System V Release 3/4/5 licensees (like IBM) didn't get (journalling/volume management/etc) as part of their SVR license.
I should have said "SVR licensees" rather than "people like IBM." (In that sense "every version of UnixWare has " *not* "shipped with Veritas," right?) Anyway, I recognize what you mean, and I hope you recognize what I meant, at least now that I've clarified it.
Related to this, think about what SCO is claiming... if IBM's JFS is a derivative work, then so is Veritas... and if SCO owns or controls the Veritas copyright, why is SCO paying Veritas for the privelege?
It's crazy logic.
--LP
What's so ironic about the suit to me is that when IBM entered the Unix market with AIX in 1990, they were the *first* Unix vendor to introduce their own journal filesystems and volume management tools integrated into the OS.
These now-common approaches to improving the reliability and flexibility of Unix were part of IBM's value-add to Unix... a bit of heritage from their mainframe and minicomputer perspective. It wasn't enough in the marketplace to overcome IBM's late-ness to Unix and the odd uniquenesses of its registry-based configuration, but it did help somewhat in enterprise environments.
Anyway, after 10 years in the Unix market, IBM decides that having had minimal/modest success in the commercial Unix marketplace, perhaps they would have better luck in the free Unix marketplace (making money selling services,) particularly if they can catch this wave early rather than spending a decade worrying about cannibalization of their own product line. So they take the AIX 'crown jewels' and share them with the free Unix community.
And SCO claims that they are derivatives of SCO's original Unix work?!
If any version of AT&T Unix/Unixware that shipped to people like IBM included journal filesystems or volume management or NUMA SMP, then maybe I could buy it, but given the dictionary definitions of "derived" I just can't.
(dictionary.com entry for
"derive": "to obtain or receive from a source".
"derivative": "copied or adapted from others")
--LP
This is not a new story, although the mention of specific names (and the Mitterand connection) involved is new to me. I read this story 3-4 years ago in the book Victory: The Reagan Administration's Secret Strategy That Hastened the Collapse of the Soviet Union by Peter Schweizer, which is fairly extensively footnoted and on the record. The interviews with the Reagan officials may of course be self-serving, but it seems reasonably well sourced. The tech details, like this Safire piece, are not that informative, but I wouldn't expect them to be really. Used hardback copies are $2 on Amazon.
--LP
Right, but if Yahoo/etc serves up the image via SSL, while they could be getting a forged referrer, they should have a valid IP # that they could then block or prosecute.
Unless the porn purveyor had compromised zombie PCs making the SSL requests and relaying the image to the porn webserver over IRC or something.
--LP
Relevant. But not determinative.
Evidence. But not proof.
--LP
The other problem with captcha-relaying is that if your captcha's have a distinct look about them, it's easy to tell if some porn site is using yours. So...
Think I could convince Yahoo or Paypal to give me a job looking at porn in an attempt to find captchas that look just like theirs?
(j/k)
You're right. But. A) you're repeating what the editor already said, and B) you are overstating your case a bit for the following reasons:
In fairness, the poster on the blog was Cory Doctorow, who is a long time, well-known net-citizen and isn't exactly some random guy, although you may not know him. For a sample of his work, see this piece in Salon which mentions that he won the John W. Campbell Award for best new science fiction writer at the 2000 Hugo Awards. He's not a journalist, he's a blogger, but it's an interesting tidbit nonetheless...
And even if he was a random blogger, his credentials are much less important than the core concept he's disclosing: that someone seeking to generate email accounts (or open bank accounts or whatever) could have porn-seeking humans workaround the turing-ish test security measures. The story is less that someone is doing it, than that someone could be doing it. At least to me.
Plus this is a hacker-type story... I wouldn't expect Reuters, etc. to carry it first.
I actually was glad to see the Slashdot editor point out the "someone told me" caveat... it's a sign to me that the editors here are getting better. They're warning us about the weaknesses in the story, not just slapping stuff up here without a care.
--LP
If the image ...has been inlined from Yahoo or Hotmail... as the article says, couldn't Yahoo/etc have their image generation scripts setup dynamically to check the referrer (or should I say referer? ;-)).
I seem to recall this approach being used by online comic strips trying to prevent inline linking from elsewhere...
--LP
The expansion of the Universe is an expansion of space itself.
That concept I understand. Your simple restatement of it though makes me wonder... has there been any thinking/investigation of whether the Plank constant could have (gradually) changed during some stage of the Big Bang and/or what the consequences of such a change would entail?
--LP
So you're saying we should invade and 'liberate' every country whose citizens oppose the government?
No. Just every country intent on gaining WMD while sitting on $20B/year of oil revenue that can easily be siphoned to enable them to get it in so many ways that we won't be able to notice it. Every country which had a grudge match against our country (applauding 9/11) and then went around giving payouts to suicide bomber families in other countries. Every country that we gave low-grade WMD to (!) and then actually showed they were dumb/dangerous enough to use them. Every country that invaded its neighbors without military provocation, etc. Hint: unlike Iran or North Korea or China or Libya or Syria, Saddam actually started two wars, against Iran and Kuwait. Every country willing to let 500,000+ of its citizens die rather than comply with cease-fire or UN terms. Every country whether the dictator taught his sons to practice torture so foreign governments would be nervous about his successors enough not to kill papa.
Power=interests*capability*will. Besides the base of economic capabilities Iraq's oil revenue gave Saddam (even when under UN sanction "containment"), Saddam had will in spades and that could not be ignored.
That's my guess anyway.
--LP
Why should Peter Jackson get the Oscar?
Trust me, it takes a great director not to f*ck up a great book.
With a good book, you do have more to work with... but also the stakes are higher, and your choices are more tested, more scrutinized. It's double or nothing. If you do well, you are more worthy. With a great book, it's quadruple or nothing.
"Artistically incorrect?" Perhaps.
--LP
In general, most Unixes and Linux (as you say) have adopted the LP64 model where longs/pointers are 64-bits and ints are 32 bits (some gory details here. (Cray's Unix is an exception; it's ILP64).
Windows OSes however have adopted the LLP64 model where ints and longs are 32-bits still, but long longs and pointers are 64-bits (gory Windows details here and here.)
Both 32-bit Windows and Unix traditionally used ILP32, so the porting characteristics moving to 64-bit code are slightly different across the two platforms.
--LinuxParanoid