The 256K limit was in DOS 1.0 and the quoted person was Bill Gates. And I am that old. The first serious machine I worked with had 16K of usable memory.
It isn't just the raw machine cycles that count. How much of the machine is used to just run the OS? On the machines I worked with (200+ nodes), over 1/2 of the computing power was being used to run the OS.
And we had a "master node". The master node was a single point of failure for the whole parallel thing. It did the workload management. By my rough calculations at the time, a 400 node cluster would have pegged that machine in the day.
I'm sure that the individual cluster machines are faster, but these are valid things to consider in reviewing one machine vs another.
Still runs, and as short a time as 3 years ago it was still the only way to communicate with the mainframe where I was working at the time (Try Disney!). I've moved on, but I had to keep the machine for a while longer to win the class action lawsuit (got a $100 merchandise certificate from Toshiba for just proving I was the original owner of it).
I still keep it around because it can read 360K floppies, 720K floppies, and it can do direct hardware reads on files on floppies with the old Norten Utilities still on it.
As for software, I still run and use Wordstar 7.0C. I was a beta tester for it, and for certain types of programming, nothing has ever beaten it.
Yup, it has a 1200baud modem, 20 meg drive and is an 8088, but it keeps on trucking. Back in 1994, I had the power supply replaced, otherwise, no problems ever.
To me, being there to bail them out is part of the tech skills, not part of social skills. It is part of the package of pride in being a professional. Pride in doing the job correctly the first time.
But some of us have enough to do without helping those that won't help themselves. As for being nice to everyone all the time, do the bosses want that, or do they want the guy who can bail them out when it hits the fan?
I can remember back a few years, having a boss type on my terminal while I was on the phone and walking him thru a database restore. He wouldn't even know what manual to look in, let alone be able to run the job! And the restore was part of the payroll system for a multi-billion dollar company.
Heck, I'm still at work now, because I am restoring a test database from production so a programmer, who screwed up, can have the copy of production data to test her correction program against.
USB drives require USB 1.1 and some 2.0 Many people out there don't even support USB 1.0, let alone higher.
Out of 4 machines at my house, 2 do not run USB, one runs 1.0 and the other 1.1
Dongles are worse than product activation! If you had lived thru the Parallel port versions you would know that.
Last year, when my Dad wanted to do his taxes that way, his only 2 machines didn't support USB at all.
I'll stick to pencil and paper thanks. I mean these people can't even figure out that my state is one of 2 that has NO state income tax (and release a product specifically geared for the state income tax as well)!
E-mail has the attachment on your machine, until you read it and delete it. Many e-mail programs work this way. The attachment is simply a file in your inbasket, but a dumb enough scan would still catch it and blame the receiptiant.
Your list of abuses is pretty good, but you missed a major hole in this setup. If the Icarus software is scanning the PCs, then by default it is scanning all the files in the PCs.
Let's say that John down the hall decides to "get even" with you for something? He sends you an MP3 file to your e-mail. Even a short one. After your machine is taken down and comes back online, he does it again. The third time, you are diconnected permenantly. Yet, you never did anything.
By storing the MP3 on a CD and only putting it online long enough to mail it, John is pretty immune from getting scanned himself. Send the file after you go to sleep, and he has 8 hours for it to be scanned before you wake up and find it.
Back around 1990 I was stuck with the problem of sending a binary file thru an e-mail system that could only handle text.
First, I sent a basica program to put the file back together, then sent the "packaged" binary to do the same thing, but faster, then sent the file.
I just wrote each byte as a comma delimited file with values from 0 to 255. I wrote each byte value as a text representaqtion of the number. I also wrote the file name and a checksum at the end. It worked rather well, though the files were a bit long (which didn't matter to the e-mail system).
The file in transit wasn't a program, as it could not be run, but upon re-assembly, it was fine.
I hope someone took that sucker down. They were granted the patent in 1997 of 1998.
I personally had written prior art that I came up with on my own in 1985. I'm not claiming to be the first one to create that particular idea either, just that I created it on my own in 1985.
Due to "work for hire" clauses, that implementation would belong to General Electric. Talk about having deep enough pockets to sue!
The intertec Superbrains and Compustars had a pair of RED keys - one at each end of the keyboard. You had to depress BOTH of them to get a reboot. I was working on these machines when the IBM PC hit the market, as one of my then bosses went to the show where they were announced.
I think those machines had been around for 3 or 4 years by then. I know they pre date 1981 when I was working on them, as the Compustar was the "new and improved" version of the Superbrain.
And these machines were probably copying someone else as well, but we will never know who, because Intertec went the way of the dinosaur....
6000 line and later on the DPS-8 line. The people I got the story about Multics from were from TIPO, when I worked for the General (83-85). Very probably they were talking about the Minis.
However, many commands, even in the DPS-8 subsystem were programs stored in the parent directory above the user directories.
Anyway, the original comment was supposed to be humorous, at SCO's expense, and it worked.
Yup, I remember 18 and 36 bits.... And who can forget the batch programs that ran with FUTIL UTIL as the name.....
In the 81-83 timeframe I worked a Government contractor. I even managed to find a real bug in the Honeywell Cobol-74 compiler. If you did the right thing in your program (which was easiest to do if you were doing a syntax check on an incomplete program) you could put the compiler into an infinite loop!
I worked GCOS from 1978 thru 1985. The look and feel was like both UNIX and DOS. I was recruited to work on a cut down version of GCOS by Honeywell, but turned them down. From 1978 thru 1981, I was at Honeywell's beta site.
I did work with some people at one point who were in a position to know. Just because you are aware of one parent of something, doesn't mean you are aware of ALL of them.
My PDP experience was all running the Digital OSes. (OS/8 and OS/11) I was never a fan on UNIX, even when I was at AT&T.
I've got manuals too, but what does that prove?
Multics had several parents, of which GCOS was one
on
SCO's Plan Examined
·
· Score: 1
You may be correct as well, as Multics was supposed to be a group project (which failed).
In the long run, GCOS had more the look and feel that evolved into UNIX anyway.
>>Renaissance also bought the story -- hook, line and sinker -- that SCO owned the UNIX tree trunk, so to speak, and that all other versions of Unix were branches, or derivatives, off of their tree, including, so they imagined, Linux.
If their logic were correct, then by the same logic, Unix is a derivative of Multics, which in turn is a derivative of GCOS. Thus, by SCO's logic, General Electric owned the whole trunk, but sold it to Honeywell, who sold it to Bull of France.....
Maybe it is time to trade in my GE stock for Bull stock?
The only tele-marketer calls I am getting these days are from the local phone company trying to get me to switch to them for Long distance. Been that way for over a year now.
I also went to the trouble of removing myself from 6 of the biggest online phone directories. That might have contributed as well.
Think what you might, but if I only get 4 or 5 calls a YEAR, I say it is working.
I have an unlisted number, which works very well thank you. If I had signed up, I would now be fair game for all those tele-marketers based in Oklahoma.
The Government has not considered anything about the amounts of storage that may be required. The only quibble I have is the 10 year calculation. I see where your numbers are coming from, but the performance curve has some other obsticles that it is running into that your calculations haven't shown.
Sorry, you miss the point. If a company like AT&T can not store 3 months of long distance information (NO such machine exists to hold it all), then the government CAN NOT analyze data that does not exist. It does exist, but only on many cartridges of tape. Reading said tape would take many hours, even if you knew what you were looking for. And before that, you have to retrieve some of the tape from "the bunker", which takes a truck to haul it back to the data center.
When you scale this up to the whole population of the country, and then add in all transactions, it can not be done with current technology. Poindexter is smoking something really good to think that it can be done. Thus, this is one of the biggest wastes of tax dollars going today. Can anyone say "Pork Barrel!".
Oh, and long term storage is ABSOLUTELY NECCESSARY to see patterns in transactions, as described by the authors of the idea.
The amount of data thatis available is much to large to even store on a single or any group of machines available to the kind of money that has been allocated to the project.
In fact, I was involved with a project to capture the billing records of AT&T for a 3 month rolling store some years ago. The largest theoretical system at the time was too small to handle the amount of data, by a factor of 1/3.
Scale this up to all transactions by all people. No computer in existance can handle the volume of data, even before the government tries to determine which data is important.
I'm not really worried about this program for that reason. It will never fly.
INTRODUCTION OF THE ``SOCIAL SECURITY NUMBER PRIVACY AND IDENTITY THEFT PREVENTION ACT OF 2003'' -- (Extensions of Remarks - July 25, 2003)
[Page: E1637] GPO's PDF ---SPEECH OF HON. E. CLAY SHAW, JR. OF FLORIDA IN THE HOUSE OF REPRESENTATIVES FRIDAY, JULY 25, 2003 Mr. SHAW. Mr. Speaker, use of Social Security numbers is rampant. When Social Security numbers were created in 1936, their only purpose was to track a worker's earnings so that Social Security benefits could be calculated. But today, we literally have a culture of dependence on Social Security numbers. Businesses and governments use the number as the primary way of identifying individuals. All of us know how difficult it is to conduct even the most mundane transactions without having to provide our Social Security number first. It's no wonder identity theft has become the fastest growing white collar crime. Worse yet, terrorists, including those responsible for the September 11th attacks, misuse SSNs in order to assimilate into our society. Barely a day goes by without hearing more examples of the truly devastating effects of identity theft. Just this month, at a Ways and Means Subcommittee on Social Security hearing, we learned about a widow whose husband died in the September 11th attacks on the World Trade Center--an illegal immigrant used her deceased husband's Social Security number to get a driver's license and to work. We also heard about individuals whose credit was ruined, who were arrested for crimes they did not commit, and who spent years and hundreds or even thousands of dollars out of their own pockets trying to clear their names because of identity theft often facilitated by obtaining the individual's Social Security number. Concerns about identity theft are increasing dramatically. According to the Federal Trade Commission, identity theft is the number one consumer complaint--amounting to 43 percent of complaints received in 2002. In fact, my state, Florida, is sixth in the nation in the number of identity theft victims per 100,000 people. Clearly, there is need for a comprehensive law to better protect the privacy of Social Security numbers and protect the American public from being victimized. Today, I re-introduce the ``Social Security Number Privacy and Identity Theft Prevention Act of 2003,'' which is similar to bipartisan legislation introduced during the last Congress. In the public and private sector, the bill would restrict the sale and public display of Social Security numbers, limit dissemination of Social Security numbers by credit reporting agencies, make it more difficult for businesses to deny services if a customer refuses to provide his or her Social Security number and establish civil and criminal penalties for violations. Based on the thoughtful comments we have received, this new legislation reflects a small number of fair and appropriate modifications, including the following: In response to concerns about potentially preventing necessary disclosures of the SSN and the impact on businesses, customers, and the economy, the U.S. Attorney General will be able to authorize the sale, purchase and display of SSNs only when necessary and with restrictions to assure the Social Security number would not be used to commit fraud or crime and to prevent risk of individual harm. Based on feedback from employee benefit plan administrators, the legislation makes clear that sale and purchase of Social Security numbers does not include its submission for administering employee benefits. In response to concerns regarding vulnerabilities in the Social Security Administration's process of issuing Social Security numbers, the bill tightens controls by requiring a photo ID; raising the standards for issuing Social Security numbers to babies; and restricting reissuance of Social Security number cards. In response to concerns about the need for stronger, clearer penalties for SSN misuse, the legislation provides enhanced criminal penalties for repeat offenders and fo
Slashdot Mirror
You can modify them to read as you see fit, as long as you don't distribute them afterwards.
The 256K limit was in DOS 1.0 and the quoted person was Bill Gates. And I am that old. The first serious machine I worked with had 16K of usable memory.
of disk space will be all anyone would ever need. But really, what requires 25G flops?
Maybe if we decide to model "Life, the Universe and Everything?"
Besides, it isn't real until it is used in some computer somewhere
It isn't just the raw machine cycles that count. How much of the machine is used to just run the OS? On the machines I worked with (200+ nodes), over 1/2 of the computing power was being used to run the OS.
And we had a "master node". The master node was a single point of failure for the whole parallel thing. It did the workload management. By my rough calculations at the time, a 400 node cluster would have pegged that machine in the day.
I'm sure that the individual cluster machines are faster, but these are valid things to consider in reviewing one machine vs another.
Still runs, and as short a time as 3 years ago it was still the only way to communicate with the mainframe where I was working at the time (Try Disney!). I've moved on, but I had to keep the machine for a while longer to win the class action lawsuit (got a $100 merchandise certificate from Toshiba for just proving I was the original owner of it).
I still keep it around because it can read 360K floppies, 720K floppies, and it can do direct hardware reads on files on floppies with the old Norten Utilities still on it.
As for software, I still run and use Wordstar 7.0C. I was a beta tester for it, and for certain types of programming, nothing has ever beaten it.
Yup, it has a 1200baud modem, 20 meg drive and is an 8088, but it keeps on trucking. Back in 1994, I had the power supply replaced, otherwise, no problems ever.
To me, being there to bail them out is part of the tech skills, not part of social skills. It is part of the package of pride in being a professional. Pride in doing the job correctly the first time.
But some of us have enough to do without helping those that won't help themselves. As for being nice to everyone all the time, do the bosses want that, or do they want the guy who can bail them out when it hits the fan?
I can remember back a few years, having a boss type on my terminal while I was on the phone and walking him thru a database restore. He wouldn't even know what manual to look in, let alone be able to run the job! And the restore was part of the payroll system for a multi-billion dollar company.
Heck, I'm still at work now, because I am restoring a test database from production so a programmer, who screwed up, can have the copy of production data to test her correction program against.
USB drives require USB 1.1 and some 2.0 Many people out there don't even support USB 1.0, let alone higher.
Out of 4 machines at my house, 2 do not run USB, one runs 1.0 and the other 1.1
Dongles are worse than product activation! If you had lived thru the Parallel port versions you would know that.
Last year, when my Dad wanted to do his taxes that way, his only 2 machines didn't support USB at all.
I'll stick to pencil and paper thanks. I mean these people can't even figure out that my state is one of 2 that has NO state income tax (and release a product specifically geared for the state income tax as well)!
E-mail has the attachment on your machine, until you read it and delete it. Many e-mail programs work this way. The attachment is simply a file in your inbasket, but a dumb enough scan would still catch it and blame the receiptiant.
Your list of abuses is pretty good, but you missed a major hole in this setup. If the Icarus software is scanning the PCs, then by default it is scanning all the files in the PCs.
Let's say that John down the hall decides to "get even" with you for something? He sends you an MP3 file to your e-mail. Even a short one. After your machine is taken down and comes back online, he does it again. The third time, you are diconnected permenantly. Yet, you never did anything.
By storing the MP3 on a CD and only putting it online long enough to mail it, John is pretty immune from getting scanned himself. Send the file after you go to sleep, and he has 8 hours for it to be scanned before you wake up and find it.
Back around 1990 I was stuck with the problem of sending a binary file thru an e-mail system that could only handle text.
First, I sent a basica program to put the file back together, then sent the "packaged" binary to do the same thing, but faster, then sent the file.
I just wrote each byte as a comma delimited file with values from 0 to 255. I wrote each byte value as a text representaqtion of the number. I also wrote the file name and a checksum at the end. It worked rather well, though the files were a bit long (which didn't matter to the e-mail system).
The file in transit wasn't a program, as it could not be run, but upon re-assembly, it was fine.
I hope someone took that sucker down. They were granted the patent in 1997 of 1998.
I personally had written prior art that I came up with on my own in 1985. I'm not claiming to be the first one to create that particular idea either, just that I created it on my own in 1985.
Due to "work for hire" clauses, that implementation would belong to General Electric. Talk about having deep enough pockets to sue!
The intertec Superbrains and Compustars had a pair of RED keys - one at each end of the keyboard. You had to depress BOTH of them to get a reboot. I was working on these machines when the IBM PC hit the market, as one of my then bosses went to the show where they were announced.
I think those machines had been around for 3 or 4 years by then. I know they pre date 1981 when I was working on them, as the Compustar was the "new and improved" version of the Superbrain.
And these machines were probably copying someone else as well, but we will never know who, because Intertec went the way of the dinosaur....
6000 line and later on the DPS-8 line. The people I got the story about Multics from were from TIPO, when I worked for the General (83-85).
Very probably they were talking about the Minis.
However, many commands, even in the DPS-8 subsystem were programs stored in the parent directory above the user directories.
Anyway, the original comment was supposed to be humorous, at SCO's expense, and it worked.
Yup, I remember 18 and 36 bits.... And who can forget the batch programs that ran with FUTIL UTIL as the name.....
In the 81-83 timeframe I worked a Government contractor. I even managed to find a real bug in the Honeywell Cobol-74 compiler. If you did the right thing in your program (which was easiest to do if you were doing a syntax check on an incomplete program) you could put the compiler into an infinite loop!
I worked GCOS from 1978 thru 1985. The look and feel was like both UNIX and DOS. I was recruited to work on a cut down version of GCOS by Honeywell, but turned them down. From 1978 thru 1981, I was at Honeywell's beta site.
I did work with some people at one point who were in a position to know. Just because you are aware of one parent of something, doesn't mean you are aware of ALL of them.
My PDP experience was all running the Digital OSes. (OS/8 and OS/11) I was never a fan on UNIX, even when I was at AT&T.
I've got manuals too, but what does that prove?
You may be correct as well, as Multics was supposed to be a group project (which failed).
In the long run, GCOS had more the look and feel that evolved into UNIX anyway.
>>Renaissance also bought the story -- hook, line and sinker -- that SCO owned the UNIX tree trunk, so to speak, and that all other versions of Unix were branches, or derivatives, off of their tree, including, so they imagined, Linux.
If their logic were correct, then by the same logic, Unix is a derivative of Multics, which in turn is a derivative of GCOS. Thus, by SCO's logic, General Electric owned the whole trunk, but sold it to Honeywell, who sold it to Bull of France.....
Maybe it is time to trade in my GE stock for Bull stock?
The only tele-marketer calls I am getting these days are from the local phone company trying to get me to switch to them for Long distance. Been that way for over a year now.
I also went to the trouble of removing myself from 6 of the biggest online phone directories. That might have contributed as well.
Think what you might, but if I only get 4 or 5 calls a YEAR, I say it is working.
A credit card is used by me to buy commodities that can not be purchased any other way, so the number of them is small.
Further, I think the CC may still have the old phone number listed, so that trail is cold for tele-marketers.
I have an unlisted number, which works very well thank you. If I had signed up, I would now be fair game for all those tele-marketers based in Oklahoma.
The Government has not considered anything about the amounts of storage that may be required. The only quibble I have is the 10 year calculation. I see where your numbers are coming from, but the performance curve has some other obsticles that it is running into that your calculations haven't shown.
Sorry, you miss the point. If a company like AT&T can not store 3 months of long distance information (NO such machine exists to hold it all), then the government CAN NOT analyze data that does not exist. It does exist, but only on many cartridges of tape. Reading said tape would take many hours, even if you knew what you were looking for. And before that, you have to retrieve some of the tape from "the bunker", which takes a truck to haul it back to the data center.
When you scale this up to the whole population of the country, and then add in all transactions, it can not be done with current technology. Poindexter is smoking something really good to think that it can be done. Thus, this is one of the biggest wastes of tax dollars going today. Can anyone say "Pork Barrel!".
Oh, and long term storage is ABSOLUTELY NECCESSARY to see patterns in transactions, as described by the authors of the idea.
The amount of data thatis available is much to large to even store on a single or any group of machines available to the kind of money that has been allocated to the project.
In fact, I was involved with a project to capture the billing records of AT&T for a 3 month rolling store some years ago. The largest theoretical system at the time was too small to handle the amount of data, by a factor of 1/3.
Scale this up to all transactions by all people. No computer in existance can handle the volume of data, even before the government tries to determine which data is important.
I'm not really worried about this program for that reason. It will never fly.
Here is the text:
INTRODUCTION OF THE ``SOCIAL SECURITY NUMBER PRIVACY AND IDENTITY THEFT PREVENTION ACT OF 2003'' -- (Extensions of Remarks - July 25, 2003)
[Page: E1637] GPO's PDF
---SPEECH OF
HON. E. CLAY SHAW, JR.
OF FLORIDA
IN THE HOUSE OF REPRESENTATIVES
FRIDAY, JULY 25, 2003
Mr. SHAW. Mr. Speaker, use of Social Security numbers is rampant. When Social Security numbers were created in 1936, their only purpose was to track a worker's earnings so that Social Security benefits could be calculated. But today, we literally have a culture of dependence on Social Security numbers.
Businesses and governments use the number as the primary way of identifying individuals. All of us know how difficult it is to conduct even the most mundane transactions without having to provide our Social Security number first. It's no wonder identity theft has become the fastest growing white collar crime.
Worse yet, terrorists, including those responsible for the September 11th attacks, misuse SSNs in order to assimilate into our society.
Barely a day goes by without hearing more examples of the truly devastating effects of identity theft. Just this month, at a Ways and Means Subcommittee on Social Security hearing, we learned about a widow whose husband died in the September 11th attacks on the World Trade Center--an illegal immigrant used her deceased husband's Social Security number to get a driver's license and to work. We also heard about individuals whose credit was ruined, who were arrested for crimes they did not commit, and who spent years and hundreds or even thousands of dollars out of their own pockets trying to clear their names because of identity theft often facilitated by obtaining the individual's Social Security number.
Concerns about identity theft are increasing dramatically. According to the Federal Trade Commission, identity theft is the number one consumer complaint--amounting to 43 percent of complaints received in 2002. In fact, my state, Florida, is sixth in the nation in the number of identity theft victims per 100,000 people.
Clearly, there is need for a comprehensive law to better protect the privacy of Social Security numbers and protect the American public from being victimized. Today, I re-introduce the ``Social Security Number Privacy and Identity Theft Prevention Act of 2003,'' which is similar to bipartisan legislation introduced during the last Congress. In the public and private sector, the bill would restrict the sale and public display of Social Security numbers, limit dissemination of Social Security numbers by credit reporting agencies, make it more difficult for businesses to deny services if a customer refuses to provide his or her Social Security number and establish civil and criminal penalties for violations.
Based on the thoughtful comments we have received, this new legislation reflects a small number of fair and appropriate modifications, including the following:
In response to concerns about potentially preventing necessary disclosures of the SSN and the impact on businesses, customers, and the economy, the U.S. Attorney General will be able to authorize the sale, purchase and display of SSNs only when necessary and with restrictions to assure the Social Security number would not be used to commit fraud or crime and to prevent risk of individual harm.
Based on feedback from employee benefit plan administrators, the legislation makes clear that sale and purchase of Social Security numbers does not include its submission for administering employee benefits.
In response to concerns regarding vulnerabilities in the Social Security Administration's process of issuing Social Security numbers, the bill tightens controls by requiring a photo ID; raising the standards for issuing Social Security numbers to babies; and restricting reissuance of Social Security number cards.
In response to concerns about the need for stronger, clearer penalties for SSN misuse, the legislation provides enhanced criminal penalties for repeat offenders and fo