I agree, maximum human life span has not increased as much as average (or median) life span.
But there's been a lot of individual and social adaptation, from a world where most people wouldn't see their grandchildren hit puberty, to a world where 60 year olds run 10K's.
I don't claim that the next 100 years are going to be similar to the past 100 years. I really claim that the human race has absorbed massive change already, and that it has been super beneficial.
I hate the "oooh, things as they are in 2000 are normal, any change will be scary" attitude of the New York Times. Demographics in 2000 are way abnormal (or supernormal) compared to the yardstick of 1900. That was a big change and I love it.
In the USA, life expectancy increased 60% from 1900 to 2000. In Italy, 80%. In Japan, 80%. In Mexico, 120%.
We are already living in an age of radical life extension compared to previous generations. A much higher percentage of the population lives to 60, 80, or 100 than used to. And I don't see a lot of people clamoring to roll back life extancy from 75+ years to 45.
75 is a lot better than 45. 120 will be better than 75. And 200 will be better than 120.
You claim that the scope of this lawsuit is a dispute with IBM.
sCO has sent thousands of letters to Linux end users warning them of legal liability. SCO publicly stated that Linux cannot possibly work on enterprise systems without illegal code theft from SCO. And Darl McBride said last week: "What is at issue is more than SCO and Red Hat. What is at issue is intellectual property rights in the age of the Internet." (Conference Call, 2003-08-05).
So don't even try copping that "this is about IBM, why is the community so upset?" line. SCO says that it is about the community and attacks the community repeatedly in their conference calls and legal filings.
"80. Any software licensed under the GPL (including Linux) must, by its terms, not be held proprietary or confidential, and may not be claimed by any party as a trade secret or copyright property."
SCO denies that any GPL software is the copyrighted property of anybody. This means that SCO denies that the Free Software Foundation owns the copyright to gcc.
That's SCO's interpretation of copyright law. You don't agree with it, and I don't agree with it, but in the hands of an expensive lawyer such as David Boies, it could cause a great deal of grief to the Free Software Foundation.
... the FSF holds the copyright to my work.
You think so, and I think so. SCO thinks that nobody holds this copyright. Which would leave the status of a copyright assignment in limbo.
Can you cite any recent public statement from a SCO officer that says otherwise?
As far as wanting help goes: my copyright assignment with the FSF says that I indemnify the FSF in case I contribute any code that contains other people's intellectual property.
Developer... will indemnify FSF for all losses if the claim [of adverse ownership] is not spurious...
I'm curious -- is that clause in your copyright assignment?
Which means, given SCO's litigious behavior, that I won't even be reading any contributions from any SCO employees in the future. I don't want to be the target of an SCO lawsuit.
SCO pays at least one employee to maintain gcc and gdb for SCO operating systems.
SCO's supported version of gcc is gcc 2. They are working on upgrading to gcc 3 but are not planning to support gcc until gcc 3.4.
SCO's gcc generates dwarf-1 debugging format (not dwarf-2). I've researched this, and the only dwarf-1 compilers I sighted were proprietary compilers from Diab and Absoft and the SCO version of gcc. All other versions of gcc in the field use other debugging formats now (dwarf-2 and stabs+, mostly).
My opinion: disengaging from SCO would hurt SCO's version of the gnu toolchain materially. Which would be good.
The company said "we are going to disengage with you" has a worldwide reputation for "adult behavior" in the computer industry. Indeed, it's a cliche that no one ever got fired for buying products from them.
During one of the conference calls, a reporter asked Darl McBride: "Is SCO turning into an IP licensing shop"?
McBride didn't like that question. Eventually he pulled some answer out of an orifice to the effect that SCO was simply defending their products and services from unfair competition.
The more that SCO'S potential "SCO License for Linux" customers perceive SCO as litigous bastards, the less likely they are to roll over and pay out $$$.
That's why it's valuable to remove support for SCO operating systems from as much software as we can.
The Regents of the University of California copyrighted their code.
The Regents licensed their code under the BSD license (obviously) including the advertising clause.
USL, a company related to AT&T, sued the University of California for distributing their proprietary code.
UC sued back, on the grounds that AT&T was distributing files developed at UC without honoring UC's license.
Ray Noorda of USL swiftly settled the suit after that, essentially abandoning almost all the original claims. The BSD license did not get tested by a judge, as far as I know.
Ray Noorda gets around. He went on to found the Canopy Group.
The rumors have some foundation. For one particular C program, on one particular machine, at the particular optimization level of -O2:
gcc 3.0.4 takes 28% more time than gcc 2.95.3
gcc 3.1.1 takes 24% more time than gcc 3.0.4
gcc 3.2.3 takes 7% more time than gcc 3.1.1
gcc 3.3 takes 5% more time than gcc 3.2.3
gcc 3.4* takes 6% more time than gcc 3.3
gcc 3.5* takes 9% more time than gcc 3.4*
The "3.4*" and "3.5*" are cvs versions as of a certain date, as these versions are far from release.
The next gcc will be gcc 3.3.2 and it is estimated for October 1. If it meets that date, and if it continues to have the same performance as gcc 3.3 and gcc 3.3.1, then that would be: 4 months, 5%.
If you use Moore's Law to estimate processor speed then your CPU is getting 100% faster every 18 months, or 4% faster per month. So in the period from 2.95.3 to 3.1.1 gcc was getting slower about the same rate as processors were getting faster. Since 3.1.1, gcc is getting slower at just 1% a month or so, and processors are getting faster at 4% a month.
Refinements to my model welcomed.
As far the trade-off goes: "compile speed" is one dimension and "new and cool features" is another dimension and "object code speed" is yet another dimension. There is no universal answer about trade-offs between dimensions, you just have to make the decision yourself.
"Hash cash" works like this: the recipient forces the sender to burn some configurable amount of CPU time before accepting a message for receipt. The recipient does this by giving the sender a problem which is hard to solve, but easy to verify the solution for.
This fight-back filter sounds similar. The fight-back filter imposes some configurable amount of load on sites extracted from the message. At least, that's what it does when everything works properly.
I'm not an expert at these things, but I like hash cash better. With hash cash, the recipient imposes the cost on the actual machine that's trying to send mail. With a fight-back filter, the recipient visits some other machines which are selected by the spammer and under the control of the spammer. That opens the door for security holes.
The disadvantage of hash cash is that it may require a SMTP protocol change to be effective.
Sun and Microsoft pay SCO $13 million ($8 million already, $5 million more per contract).
Canopy Group buys the IP, picks the lawyers, picks the executives, and changes the name of the company from "Caldera" to "The SCO Group".
SCO goes berserk with the FUD and the lawsuits.
If it doesn't work, Canopy regroups and tries something else. They will be 1/2 with this business model (succeeded with Caldera International versus Microsoft, failed with SCO versus IBM).
At the worst, Microsoft and Sun will lose their $13 million and they will move on and try the next thing. Canopy is not taking much of a hit.
The SCO Group is simply a cut-out for Canopy, and a puppet for Microsoft and Sun.
Potentially telltale mail was identified using masses of computer data recorded as each letter entering the highly automated sorting centers is scanned for an address, given identifying bar codes recording its time and place of posting, and sent on its way.
The data include digital images of almost every hand-addressed envelope, which optical scanners cannot easily read, postal officials said.
The big question is: will the post office stop delivering mail that doesn't have a valid return address?
In the time of the Unabomer, the PO stopped delivering mail that weighs over one pound and came from a collection box. Mail that weighs over one pound has to be brought in person to a post office.
The post office OCR's the mail and keeps the scans. They also apply a unique bar code to each piece of first class mail. Then they record all this information and save it for a while.
I also remember reading that they save the sender's information as well. It was in an anthrax story that said they went to all the curbside mailboxes where all the pieces that were close to an anthrax-related piece had been sent.
Monday: Red Hat sues SCO. Thursday: IBM sues SCO. Friday: Oracle announces commitment to Linux.
This is a good way to do PR: a rolling wave from different sources.
Objectively, it's the same total amount of commitment to Linux whether everybody does it in one day or they do it three weeks apart. But this timing feeds the news cycle better.
I'm hoping that Google will issue a press release soon. And then a behemoth retailer, like Home Depot. And then a brokerage firm, like Merrill Lynch.
Sun's spokesman also didn't mention that Sun's deal with SCO includes a warrant to purchase 210,000 shares of SCOX at $1.83 per share. Which is quite a bargain given the current price of SCOX.
Naturally, Sun does whatever is good for Sun. Sometimes that hurts Linux (they finance SCO and get ownership of 2% of the comany) and sometimes that helps Linux (Open Office).
SCO are pissing off their Unixware and Open Server customers, and their resellers, and their ISV's. Who wants to be a business partner of SCO? Heard much about customer wins for SCO/X Web Services lately? Me neither.
Next question: why would SCO do such a thing? Answer: because 40% of their revenue, and all of their net profit, comes from "SCO Source" contracts with Microsoft and Sun.
SCO gets paid precisely *for* this FUD activity. They don't even need to sell any of these $699 licenses, they just need to get every IT director in the world to link "Linux" and "IP lawsuit" together.
The SCO Group is a weird aberration of a company. It's actually a sock puppet of The Canopy Group, where all the money is flowing (that's where all the Caldera International lawsuit money from Microsoft went, too).
IBM is a more normal company. They sell products and services. Unlike SCO, they make a profit doing so. If they started screwing with Linux the way that SCO is, IBM would lose profitable business.
You bet that IBM is in it for the money. It so happens that supporting Linux, in a big way, makes IBM lots of money.
What ownership of the company does 210K shares represent?
SCO has 12,262,003 shares outstanding as of March 28, 2003. So 210,000 shares is 1.7%.
It was Sun, rather than Microsoft, who got the warrant.
BTW, the guy who was talking about stock that receives dividends but doesn't vote -- that would be preferred stock. Preferred stock doesn't vote, but they are guaranteed to get their dividends paid before common stock gets any dividends paid. They are "preferred" in that sense.
This is a warrant to buy common stock. Common stock does vote, and it is last in line to get paid when the company pays out money. If I recall correctly, the order is: taxes, employees, creditors, senior bonds, junior bonds, preferred stock, common stock. That generally only matters when the company is insolvent and has to choose which people to pay because they can't pay everyone.
Kean Johnston said that SCO's current working version of gcc is gcc 2. That is more than two years old now. SCO have not upgraded their customers to gcc 3 yet and Kean is aiming for gcc 3.4 to do that.
Which means their is a window of opportunity where disengaging from SCO starts hurting them right now.
It will be relatively easy to maintain...
I disagree with your assessment about how easy "relatively easy" will be.
Slashdot Mirror
The Free Software Foundation is not about IT. It's about development. They are different things.
That is, they know all about how to write software, and not much about operating a public Internet site.
You're new here, aren't you?
Obligatory question which Kean can't discuss: has everybody in SCO's engineering sold 100% of their stock yet, or just Opinder Bawa?
Exhibit F
Exhibit G
I agree with the meat of your analysis about Sequent. I don't know enough contract law to comment on SCO's thesis about Sequent.
I agree, maximum human life span has not increased as much as average (or median) life span.
But there's been a lot of individual and social adaptation, from a world where most people wouldn't see their grandchildren hit puberty, to a world where 60 year olds run 10K's.
I don't claim that the next 100 years are going to be similar to the past 100 years. I really claim that the human race has absorbed massive change already, and that it has been super beneficial.
I hate the "oooh, things as they are in 2000 are normal, any change will be scary" attitude of the New York Times. Demographics in 2000 are way abnormal (or supernormal) compared to the yardstick of 1900. That was a big change and I love it.
Dunno if it's in good taste to link to my own comment in a previous slashdot article ... I'll let the mods decide.
Get your info straight from the SEC
Before considering the future, let's have a look at the past.
TIME 100: 1900 vs. Now
In the USA, life expectancy increased 60% from 1900 to 2000. In Italy, 80%. In Japan, 80%. In Mexico, 120%.
We are already living in an age of radical life extension compared to previous generations. A much higher percentage of the population lives to 60, 80, or 100 than used to. And I don't see a lot of people clamoring to roll back life extancy from 75+ years to 45.
75 is a lot better than 45. 120 will be better than 75. And 200 will be better than 120.
You claim that the scope of this lawsuit is a dispute with IBM.
sCO has sent thousands of letters to Linux end users warning them of legal liability. SCO publicly stated that Linux cannot possibly work on enterprise systems without illegal code theft from SCO. And Darl McBride said last week: "What is at issue is more than SCO and Red Hat. What is at issue is intellectual property rights in the age of the Internet." (Conference Call, 2003-08-05).
So don't even try copping that "this is about IBM, why is the community so upset?" line. SCO says that it is about the community and attacks the community repeatedly in their conference calls and legal filings.
SCO's opinion of the GPL:
... the FSF holds the copyright to my work.
... will indemnify FSF for all losses if the claim [of adverse ownership] is not spurious ...
From their filing of 2003-03-06:
"80. Any software licensed under the GPL (including Linux) must, by its terms, not be held proprietary or confidential, and may not be claimed by any party as a trade secret or copyright property."
SCO denies that any GPL software is the copyrighted property of anybody. This means that SCO denies that the Free Software Foundation owns the copyright to gcc.
That's SCO's interpretation of copyright law. You don't agree with it, and I don't agree with it, but in the hands of an expensive lawyer such as David Boies, it could cause a great deal of grief to the Free Software Foundation.
You think so, and I think so. SCO thinks that nobody holds this copyright. Which would leave the status of a copyright assignment in limbo.
Can you cite any recent public statement from a SCO officer that says otherwise?
As far as wanting help goes: my copyright assignment with the FSF says that I indemnify the FSF in case I contribute any code that contains other people's intellectual property.
Developer
I'm curious -- is that clause in your copyright assignment?
Which means, given SCO's litigious behavior, that I won't even be reading any contributions from any SCO employees in the future. I don't want to be the target of an SCO lawsuit.
Re: Deprecate dwarf and mdebug support, delete nlm?
Some facts:
SCO pays at least one employee to maintain gcc and gdb for SCO operating systems.
SCO's supported version of gcc is gcc 2. They are working on upgrading to gcc 3 but are not planning to support gcc until gcc 3.4.
SCO's gcc generates dwarf-1 debugging format (not dwarf-2). I've researched this, and the only dwarf-1 compilers I sighted were proprietary compilers from Diab and Absoft and the SCO version of gcc. All other versions of gcc in the field use other debugging formats now (dwarf-2 and stabs+, mostly).
My opinion: disengaging from SCO would hurt SCO's version of the gnu toolchain materially. Which would be good.
The company said "we are going to disengage with you" has a worldwide reputation for "adult behavior" in the computer industry. Indeed, it's a cliche that no one ever got fired
for buying products from them.
During one of the conference calls, a reporter asked Darl McBride: "Is SCO turning into an IP licensing shop"?
McBride didn't like that question. Eventually he pulled some answer out of an orifice to the effect that SCO was simply defending their products and services from unfair competition.
The more that SCO'S potential "SCO License for Linux" customers perceive SCO as litigous bastards, the less likely they are to roll over and pay out $$$.
That's why it's valuable to remove support for SCO operating systems from as much software as we can.
... and I think it's either the web server at myownlittleworld.com or my own Mozilla browser.
I see seriously different numbers with mozilla and Lynx. The Mozilla numbers are as I described, and the Lynx numbers are as you described.
The page in Mozilla says "All snapshots were from 20030614" and the page in Lynx says "All snapshots were from 20030725".
And the numbers for 3.1.1 and 3.2.3 are significantly different between the versions of the page.
Ouch!
That was part of the USL - BSDI lawsuits.
The Regents of the University of California copyrighted their code.
The Regents licensed their code under the BSD license (obviously) including the advertising clause.
USL, a company related to AT&T, sued the University of California for distributing their proprietary code.
UC sued back, on the grounds that AT&T was distributing files developed at UC without honoring UC's license.
Ray Noorda of USL swiftly settled the suit after that, essentially abandoning almost all the original claims. The BSD license did not get tested by a judge, as far as I know.
Ray Noorda gets around. He went on to found the Canopy Group.
Wha?
These are run times, in seconds
Also, to help developers, an effort has been made to speed up GCC 3.3.x at -O0 where it is faster than GCC 3.1.x and 3.2.x
That's true.
A little google whoring turns this up:
GCC Compilation Comparison
The rumors have some foundation. For one particular C program, on one particular machine, at the particular optimization level of -O2:
gcc 3.0.4 takes 28% more time than gcc 2.95.3
gcc 3.1.1 takes 24% more time than gcc 3.0.4
gcc 3.2.3 takes 7% more time than gcc 3.1.1
gcc 3.3 takes 5% more time than gcc 3.2.3
gcc 3.4* takes 6% more time than gcc 3.3
gcc 3.5* takes 9% more time than gcc 3.4*
The "3.4*" and "3.5*" are cvs versions as of a certain date, as these versions are far from release.
Here are some release dates:
2001-03-22 gcc 2.95.3
2002-02-21 gcc 3.0.4
2002-07-26 gcc 3.1.1
2003-04-23 gcc 3.2.3
2003-05-14 gcc 3.3
Correlating these:
gcc 3.0.4, 11 months, 28%
gcc 3.1.1, 5 months, 24%
gcc 3.2.3, 9 months, 5%
The next gcc will be gcc 3.3.2 and it is estimated for October 1. If it meets that date, and if it continues to have the same performance as gcc 3.3 and gcc 3.3.1, then that would be: 4 months, 5%.
If you use Moore's Law to estimate processor speed then your CPU is getting 100% faster every 18 months, or 4% faster per month. So in the period from 2.95.3 to 3.1.1 gcc was getting slower about the same rate as processors were getting faster. Since 3.1.1, gcc is getting slower at just 1% a month or so, and processors are getting faster at 4% a month.
Refinements to my model welcomed.
As far the trade-off goes: "compile speed" is one dimension and "new and cool features" is another dimension and "object code speed" is yet another dimension. There is no universal answer about trade-offs between dimensions, you just have to make the decision yourself.
"Hash cash" works like this: the recipient forces the sender to burn some configurable amount of CPU time before accepting a message for receipt. The recipient does this by giving the sender a problem which is hard to solve, but easy to verify the solution for.
This fight-back filter sounds similar. The fight-back filter imposes some configurable amount of load on sites extracted from the message. At least, that's what it does when everything works properly.
I'm not an expert at these things, but I like hash cash better. With hash cash, the recipient imposes the cost on the actual machine that's trying to send mail. With a fight-back filter, the recipient visits some other machines which are selected by the spammer and under the control of the spammer. That opens the door for security holes.
The disadvantage of hash cash is that it may require a SMTP protocol change to be effective.
Sun and Microsoft pay SCO $13 million ($8 million already, $5 million more per contract).
Canopy Group buys the IP, picks the lawyers, picks the executives, and changes the name of the company from "Caldera" to "The SCO Group".
SCO goes berserk with the FUD and the lawsuits.
If it doesn't work, Canopy regroups and tries something else. They will be 1/2 with this business model (succeeded with Caldera International versus Microsoft, failed with SCO versus IBM).
At the worst, Microsoft and Sun will lose their $13 million and they will move on and try the next thing. Canopy is not taking much of a hit.
The SCO Group is simply a cut-out for Canopy, and a puppet for Microsoft and Sun.
Some more link whoring ...
Postal Theory: Mail Sorter Acted as Mill for Anthrax
Read down towards the bottom:
Potentially telltale mail was identified using masses of computer data recorded as each letter entering the highly automated sorting centers is scanned for an address, given identifying bar codes recording its time and place of posting, and sent on its way.
The data include digital images of almost every hand-addressed envelope, which optical scanners cannot easily read, postal officials said.
The big question is: will the post office stop delivering mail that doesn't have a valid return address?
In the time of the Unabomer, the PO stopped delivering mail that weighs over one pound and came from a collection box. Mail that weighs over one pound has to be brought in person to a post office.
The post office OCR's the mail and keeps the scans. They also apply a unique bar code to each piece of first class mail. Then they record all this information and save it for a while.
Tracking of Anthrax Letter Yields Clues
I also remember reading that they save the sender's information as well. It was in an anthrax story that said they went to all the curbside mailboxes where all the pieces that were close to an anthrax-related piece had been sent.
Monday: Red Hat sues SCO.
Thursday: IBM sues SCO.
Friday: Oracle announces commitment to Linux.
This is a good way to do PR: a rolling wave from different sources.
Objectively, it's the same total amount of commitment to Linux whether everybody does it in one day or they do it three weeks apart. But this timing feeds the news cycle better.
I'm hoping that Google will issue a press release soon. And then a behemoth retailer, like Home Depot. And then a brokerage firm, like Merrill Lynch.
Don't drag Sun into this.
Sun dragged themselves into it.
Sun expands Unix deal with SCO
Sun and Microsoft paid SCO $8.3 million last quarter, with contracts to pay them $5 million more in the next three quarters.
Sun also received warrants to purchase 210,000 shares of SCOX at $1.83 per share.
Although they have flung some FUD in the SCO soap opera, they probably are not financing SCO in any way.
Sun purchased an SCO Source license in 2003.
Sun expands Unix deal with SCO
I wonder why Sun's spokesman didn't mention that?
Sun's spokesman also didn't mention that Sun's deal with SCO includes a warrant to purchase 210,000 shares of SCOX at $1.83 per share. Which is quite a bargain given the current price of SCOX.
Naturally, Sun does whatever is good for Sun. Sometimes that hurts Linux (they finance SCO and get ownership of 2% of the comany) and sometimes that helps Linux (Open Office).
SCO are pissing off their Unixware and Open Server customers, and their resellers, and their ISV's. Who wants to be a business partner of SCO? Heard much about customer wins for SCO/X Web Services lately? Me neither.
Next question: why would SCO do such a thing? Answer: because 40% of their revenue, and all of their net profit, comes from "SCO Source" contracts with Microsoft and Sun.
SCO gets paid precisely *for* this FUD activity. They don't even need to sell any of these $699 licenses, they just need to get every IT director in the world to link "Linux" and "IP lawsuit" together.
The SCO Group is a weird aberration of a company. It's actually a sock puppet of The Canopy Group, where all the money is flowing (that's where all the Caldera International lawsuit money from Microsoft went, too).
IBM is a more normal company. They sell products and services. Unlike SCO, they make a profit doing so. If they started screwing with Linux the way that SCO is, IBM would lose profitable business.
You bet that IBM is in it for the money. It so happens that supporting Linux, in a big way, makes IBM lots of money.
What ownership of the company does 210K shares represent?
SCO has 12,262,003 shares outstanding as of March 28, 2003. So 210,000 shares is 1.7%.
It was Sun, rather than Microsoft, who got the warrant.
BTW, the guy who was talking about stock that receives dividends but doesn't vote -- that would be preferred stock. Preferred stock doesn't vote, but they are guaranteed to get their dividends paid before common stock gets any dividends paid. They are "preferred" in that sense.
This is a warrant to buy common stock. Common stock does vote, and it is last in line to get paid when the company pays out money. If I recall correctly, the order is: taxes, employees, creditors, senior bonds, junior bonds, preferred stock, common stock. That generally only matters when the company is insolvent and has to choose which people to pay because they can't pay everyone.
Did you read post #2?
...
Kean Johnston said that SCO's current working version of gcc is gcc 2. That is more than two years old now. SCO have not upgraded their customers to gcc 3 yet and Kean is aiming for gcc 3.4 to do that.
Which means their is a window of opportunity where disengaging from SCO starts hurting them right now.
It will be relatively easy to maintain
I disagree with your assessment about how easy "relatively easy" will be.