Virgil's paper talks about the HARDWARE verification (also important, but...) not the software verification.
The SCOMP OS was general purpose. It had a Unix emulation layer over its kernel similar to BSD over Mach. And the kernel WAS verified. The Unix System-call layer was modelled and verified against a security model. They have 10000 theorems, we had about 3500. They use Isabell, we used Boyer-Moore. Sounds like a fairly similar approach.
And you are right: It was a very cool piece of work for the time. I am not trying to denigrate the referenced paper. It sounds wonderful and I look forward to reading it. I just object to terms like "unique" and "first".
You are assuming that airport security is competent and doing something related to real security rather than performing meaningless security theater to calm the crowds.
I was going to post this as a followup to the earlier IBM thread, but this really needs an expansion.
Yes, IBM was the big bad monoply way back when. But we need to remember that the BIG anti trust finding with IBM that relates to the OS wars of today is that it was found to be illegal for IBM to bundle OS-360 with its IBM-360 hardware. The release of the OS from the 360's hardware was what allowed Gene Amdahl and others to split off and form IBM-360-clone companys. It was an anti-trust decision that required the unbundling of the OS.
The big difference here is that rather than one company (Microsoft) bundling its OS with its own hardware, Microsoft has contracts with all the PC vendors that require them to bundle. So it is one step removed from the IBM situation.
The question (that has been asked before by the likes of Judge Jackson) is: what can be done about these very private contracts?
Ok, I share your worry that it is difficult to bring up troubling questions without being labeled a Troll. I value discussion of all points of view.
However, PJ DID address the original suspicion in her first response.
She first stated simply and clearly "NO" "Wrong" "I Didn't"
Then she went into a lot of detail justifying the simple direct "addressing of the original suspicion" with what seems a very complete discussion of all the allegations. Such as IBM funding because of donating to Ibiblio and how a document might get to her hands before the official court website and so on.
So I think that the original suspicions have been very well addressed in a quick and complete way.
And then she did say "Nyeh-Nyeh". It is a little childish, but I wold probably say the same thing given the circumstances.
Micro$oft has both purchased a very large license for undisclosed Unix IP (can't they get anything that they need from BSD?) and has been implicated at least a little bit in under-the-table assurances to places like Baystar that their "investment" in SCO will be covered when it is lost.
It has probably been worth a great deal to spread IP FUD about Linux until Vista gets up and going. Do you think that they have gotten their money's worth in funding TSCOG's long slow suicide?
Those who forget history...
on
The Birth of vi
·
· Score: 5, Insightful
This is good history to remember. Those who weren't there find it hard to appreciate the tremendous leap forward of Unix Version 6 and ed on a PDP-11. We had been using teco on our PDP-10 and the cousin of ed that was on Multics, but we had been getting into PDP-11s for more and more things. Comparing ed on Unix with the line editors available on PDP-11 DOS/BATCH and that new-fangeled RT-11 thing was amazing. Along with all the other tools available on Unix, the PDP-11 went from a toy to a state-of-the-art (for then) development environment. We were mostly on DECwriters and TI-Silent 700s runing hardwired 1200 baud at work and 300 baud from home over the modems. We started to get VT-100s about the time vi was being released and it was again a great leap forward.
Thanks Bill Joy! I have used your work in the BSDs and Suns and all the followons over the years, but vi was a most important gift at an important time.
Linus has the one entry that is really a compliment.
Dennis Ritchie gave a nice talk on the 21st(??) birthday of Unix about how it is like a child growing up, leaving home, being all grown up and an adult... He felt a little like a proud parent.
What better compliment for Linus than to have created something that has grown and matured to the point that it is beyond the creator? I can imagine few more satisfying accomplishments in life.
The only people who need your SS# is your employer because they have to make the contributions. Your bank doesn't need it - they, as well as your mortgage company , broker, etc., can use a Taxpayer ID # to create 1099s and such for the IRS.
I will probably get modded down for inserting a few facts...
The SSN IS your taxpayer ID, unless you are a corporation in which case it is your EIN.
As you noted, your employer witholds taxes identified with your SSN. Properly allowed by law.
Banks pay interest which is reported to the IRS under your SSN. Properly allowed by law.
Your broker may be dealing with 401K, 403B, etc types of accounts whose activities need to be reported to the IRS under your SSN. Properly allowed by law.
Mortgage Companies are payed interest which may be deductable and is reported to the IRS under your SSN. properly allowed by Law.
The problem starts when this convieniently avaliable identifier is then used by the institution for a whole lot of other primary keys just because it is already there and the law does not restrict its use unless the institution is a governmental one. See the 1974 Privacy Act section 7.
We need to get the restrictions of the 1974 Privacy Act extended to non-governmental use as well as governmental use along with penalties for an organization that improperly uses the SSN as an identifier when not authorized by a federal law.
Much of the debate on the 1974 Privacy Act revolved around the fact that the SSN was NOT to be used as a universal identifier. Paragraph 7 (if my memory serves) restricted the use of SSNs to those things either grandfathered (allowed by federal, state, or local law) before 1974 or explicitly named and allowed in a federal law; and in either case including a requirement that the requestor tell you the basis for the request. (Note that folks blanketly refusing to give the SSN are usually not on strong legal ground. Much better is to refuse until the requestor provides the legal basis for the request as provided for in the Privacy Act. IANAL etc...).
The loophole was that this act only restricted government not the private sector. Thus banks, insurance companies, universities, employers, local pizza joints, all ask for the SSN and can refuse service unless you provide it.
It would be a good start to debate if we could base a new law on the existing historical basis for the limitations in the 1974 privacy act, and then extend those restrictions to ALL use of the SSN by anyone.
>That rule is a good start. All of my personal info is covered by copyright,
Now I am a strong believer in the GPL, so I would use copyleft. That way if there is (for example) an error in my SSN, anyone can go in and just fix it and the world will be a better place...
I Started using GNUcash on Linux for exactly the reasons that were stated. I wanted basic checkbook balance, a view of how I manage and mismanage my paycheck, and a little projection of how the bills would eat into the bank account by next paycheck so as to budget toys.
The learning curve was fairly high, not being an accountant and not knowing anything about this Double Entry Bookkeeping thing.
I had used Quicken before, and found it OK, but I don't normally run windows...
Once I made it through the Included tutorials and documentation, I have used GNUcash regularly for the last year or so and am a very happy user.
It is unfortunate that GNUcash is not on enough radar screens to show up in a weblog like this one. A REAL comparison would be nice.
The Multics Relational Data Store (MRDS, The French loved the name) was the first commercial database system, marketed by Honeywell on the Multics in 1977. It had an early SQL as the standards bodies churned the standard into shape.
I know, Oracle was early, but as in so many other things, Multics was first.
A collection of previously un-copyrightable associations between names and some data is now a copyrightable item.
So If SCO can claim the original ownership due to first "collection" of say
#define EPERM 1
#define ENOENT 2
and so on, then SCO suddenly can find that errno.h is a copyright violation of System-V code in Linux, where there was none before. It is not the file "errno.h"; Linus can state under oath that his is an independant creation. But now the Database itself contained in the file is copyrighted... Just what Darl is trying to claim.
I don't think that this needs the tinfoil hat. This is way too subtle for Darl.
Averment 14: After successful in-house use of the UNIX software, AT&T began to license UNIX as a commercial product for use in enterprise applications by other large companies.
The Unix developers at AT&T Bell Labs had little idea what to do with Unix, and they consequently licensed it to universities for study, instruction and, as a natural byproduct, additional development. Sales of an operating system were not a priority for the telephone monopoly.
Sales of the OS were FORBIDDEN by the 1956 consent decree that allowed the monopoly telephone company. It was not until the Bell breakup that they it was even legally possible to sell Unix licenses.
Lots of comments so far with bits of the puzzle. Let's put them all together.
Microsoft Office with DRM
DRM includes application, client and SERVER (e.g.OS) pieces
DMCA prevents reverse engineering of protocols
(The one no one has mentioned)
Microsoft owns a patent on "A DRM Operating System"
So it's protected, you can't break it without violating the DMCA, and if you reverse engineer it is some way that skirts the DMCA, you are violating a Microsoft patent.
You do the Math, but I think that any ability to interoperate with Microsoft Office on a non-Microsoft-approved platform just evaporated.
Microsoft is not as dumb as some here on/. would have you believe.
The basic premise of the article, that overall survival depends on more than technical merit is accurate. However, the author omits one important factor that is also seen in the internet of today.
P0rn!
Sony was hesitant to license, or make available, the format to major porn makers. VHS was chosen. The main initial market for those $1500 players and $100 tapes was that normal horney people could finally see adult content in the privacy of their own home. Go check out some of those 1979-1980 Penthouse magazines on eBay and look in the back at the first tape advertisements. All VHS!
Those recording the history of the internet are hesitant to document the importance of adult content e.g. to developing secure credit card mechanisms. This was critical to the rise of the internet we know today.
If one is to learn from history, the history must be available in a complete form.
The poster asked for an Architecture:
The Common Lisp Interface Manager is far from perfect but it does use a very nice multiple inheritance scheme in CLOS to implement specialized methods based on both the object and a point-of-view of the object.
CLIM has been used to implement complex interfaces as the poster described and requested.
One example architecture... Google will point farther down the path.
The famous PDP-6 was asynch logic. It made a very fast machine out of very few transistors, but was a nightmare to maintain. The follow-on PDP-10 was syncronous logic.
There must be some history out there somewhere of the problems DEC had with the asynchronous logic. Any old MIT research notes?
Slashdot Mirror
Virgil's paper talks about the HARDWARE verification (also important, but...) not the software verification.
The SCOMP OS was general purpose. It had a Unix emulation layer over its kernel similar to BSD over Mach. And the kernel WAS verified. The Unix System-call layer was modelled and verified against a security model. They have 10000 theorems, we had about 3500. They use Isabell, we used Boyer-Moore. Sounds like a fairly similar approach.
And you are right: It was a very cool piece of work for the time. I am not trying to denigrate the referenced paper. It sounds wonderful and I look forward to reading it. I just object to terms like "unique" and "first".
Honywell SCOMP Early 1980s Was intended to me a secure front-end to Multics.
Verified by NSA et all as part of the first Orange-book A1 level certification.
For the time it was a magnificent bit of work.
One word: Multics. Way too far ahead of its time. Those who forget history will have to try to re-invent it. Badly.
You are assuming that airport security is competent and doing something related to real security rather than performing meaningless security theater to calm the crowds.
Yes, IBM was the big bad monoply way back when. But we need to remember that the BIG anti trust finding with IBM that relates to the OS wars of today is that it was found to be illegal for IBM to bundle OS-360 with its IBM-360 hardware. The release of the OS from the 360's hardware was what allowed Gene Amdahl and others to split off and form IBM-360-clone companys. It was an anti-trust decision that required the unbundling of the OS.
The big difference here is that rather than one company (Microsoft) bundling its OS with its own hardware, Microsoft has contracts with all the PC vendors that require them to bundle. So it is one step removed from the IBM situation.
The question (that has been asked before by the likes of Judge Jackson) is: what can be done about these very private contracts?
However, PJ DID address the original suspicion in her first response.
She first stated simply and clearly "NO" "Wrong" "I Didn't"
Then she went into a lot of detail justifying the simple direct "addressing of the original suspicion" with what seems a very complete discussion of all the allegations. Such as IBM funding because of donating to Ibiblio and how a document might get to her hands before the official court website and so on.
So I think that the original suspicions have been very well addressed in a quick and complete way.
And then she did say "Nyeh-Nyeh". It is a little childish, but I wold probably say the same thing given the circumstances.
Micro$oft has both purchased a very large license for undisclosed Unix IP (can't they get anything that they need from BSD?) and has been implicated at least a little bit in under-the-table assurances to places like Baystar that their "investment" in SCO will be covered when it is lost.
It has probably been worth a great deal to spread IP FUD about Linux until Vista gets up and going. Do you think that they have gotten their money's worth in funding TSCOG's long slow suicide?
This is good history to remember. Those who weren't there find it hard to appreciate the tremendous leap forward of Unix Version 6 and ed on a PDP-11. We had been using teco on our PDP-10 and the cousin of ed that was on Multics, but we had been getting into PDP-11s for more and more things. Comparing ed on Unix with the line editors available on PDP-11 DOS/BATCH and that new-fangeled RT-11 thing was amazing. Along with all the other tools available on Unix, the PDP-11 went from a toy to a state-of-the-art (for then) development environment. We were mostly on DECwriters and TI-Silent 700s runing hardwired 1200 baud at work and 300 baud from home over the modems. We started to get VT-100s about the time vi was being released and it was again a great leap forward.
Thanks Bill Joy! I have used your work in the BSDs and Suns and all the followons over the years, but vi was a most important gift at an important time.
Linus has the one entry that is really a compliment.
Dennis Ritchie gave a nice talk on the 21st(??) birthday of Unix about how it is like a child growing up, leaving home, being all grown up and an adult... He felt a little like a proud parent.
What better compliment for Linus than to have created something that has grown and matured to the point that it is beyond the creator? I can imagine few more satisfying accomplishments in life.
I will probably get modded down for inserting a few facts...
The SSN IS your taxpayer ID, unless you are a corporation in which case it is your EIN.
As you noted, your employer witholds taxes identified with your SSN. Properly allowed by law.
Banks pay interest which is reported to the IRS under your SSN. Properly allowed by law.
Your broker may be dealing with 401K, 403B, etc types of accounts whose activities need to be reported to the IRS under your SSN. Properly allowed by law.
Mortgage Companies are payed interest which may be deductable and is reported to the IRS under your SSN. properly allowed by Law.
The problem starts when this convieniently avaliable identifier is then used by the institution for a whole lot of other primary keys just because it is already there and the law does not restrict its use unless the institution is a governmental one. See the 1974 Privacy Act section 7.
We need to get the restrictions of the 1974 Privacy Act extended to non-governmental use as well as governmental use along with penalties for an organization that improperly uses the SSN as an identifier when not authorized by a federal law.
Much of the debate on the 1974 Privacy Act revolved around the fact that the SSN was NOT to be used as a universal identifier. Paragraph 7 (if my memory serves) restricted the use of SSNs to those things either grandfathered (allowed by federal, state, or local law) before 1974 or explicitly named and allowed in a federal law; and in either case including a requirement that the requestor tell you the basis for the request. (Note that folks blanketly refusing to give the SSN are usually not on strong legal ground. Much better is to refuse until the requestor provides the legal basis for the request as provided for in the Privacy Act. IANAL etc...).
The loophole was that this act only restricted government not the private sector. Thus banks, insurance companies, universities, employers, local pizza joints, all ask for the SSN and can refuse service unless you provide it.
It would be a good start to debate if we could base a new law on the existing historical basis for the limitations in the 1974 privacy act, and then extend those restrictions to ALL use of the SSN by anyone.
Our stock just dipped below $1.00. What to do?? What to do?
1. Call this guy I know named Darl.
2. Find a bunch of our 3D graphics code that has somehow gotten into Linux.
3. Find some deep pockets; maybe IBM... and sue them for giving our secrets to the great unwashed masses
4. PROFIT!!! It Can't fail!!!
Man! These Phishing schemes get wilder every day.
DELETE
I wanted to comment that N-Gage is just too small and talk about the G-Gage in my garden.
And then there is the question: are there any moderators also into model railroading who can give these posts the +5 humorous that they deserve?
Now I am a strong believer in the GPL, so I would use copyleft. That way if there is (for example) an error in my SSN, anyone can go in and just fix it and the world will be a better place...
The learning curve was fairly high, not being an accountant and not knowing anything about this Double Entry Bookkeeping thing.
I had used Quicken before, and found it OK, but I don't normally run windows...
Once I made it through the Included tutorials and documentation, I have used GNUcash regularly for the last year or so and am a very happy user.
It is unfortunate that GNUcash is not on enough radar screens to show up in a weblog like this one. A REAL comparison would be nice.
The Multics Relational Data Store (MRDS, The French loved the name) was the first commercial database system, marketed by Honeywell on the Multics in 1977. It had an early SQL as the standards bodies churned the standard into shape.
I know, Oracle was early, but as in so many other things, Multics was first.
A collection of previously un-copyrightable associations between names and some data is now a copyrightable item.
So If SCO can claim the original ownership due to first "collection" of say
#define EPERM 1
#define ENOENT 2
and so on, then SCO suddenly can find that errno.h is a copyright violation of System-V code in Linux, where there was none before. It is not the file "errno.h"; Linus can state under oath that his is an independant creation. But now the Database itself contained in the file is copyrighted... Just what Darl is trying to claim.
I don't think that this needs the tinfoil hat. This is way too subtle for Darl.
Never wrestle with a Pig.
You both get dirty...
and the Pig likes it!
Well, for a little while longer, maybe...
Sales of the OS were FORBIDDEN by the 1956 consent decree that allowed the monopoly telephone company. It was not until the Bell breakup that they it was even legally possible to sell Unix licenses.
Lots of comments so far with bits of the puzzle. Let's put them all together.
Microsoft Office with DRM
DRM includes application, client and SERVER (e.g.OS) pieces
DMCA prevents reverse engineering of protocols
(The one no one has mentioned) Microsoft owns a patent on "A DRM Operating System"
So it's protected, you can't break it without violating the DMCA, and if you reverse engineer it is some way that skirts the DMCA, you are violating a Microsoft patent.
You do the Math, but I think that any ability to interoperate with Microsoft Office on a non-Microsoft-approved platform just evaporated.
Microsoft is not as dumb as some here on /. would have you believe.
P0rn!
Sony was hesitant to license, or make available, the format to major porn makers. VHS was chosen. The main initial market for those $1500 players and $100 tapes was that normal horney people could finally see adult content in the privacy of their own home. Go check out some of those 1979-1980 Penthouse magazines on eBay and look in the back at the first tape advertisements. All VHS!
Those recording the history of the internet are hesitant to document the importance of adult content e.g. to developing secure credit card mechanisms. This was critical to the rise of the internet we know today.
If one is to learn from history, the history must be available in a complete form.
The Common Lisp Interface Manager is far from perfect but it does use a very nice multiple inheritance scheme in CLOS to implement specialized methods based on both the object and a point-of-view of the object.
CLIM has been used to implement complex interfaces as the poster described and requested.
One example architecture... Google will point farther down the path.
The famous PDP-6 was asynch logic. It made a very fast machine out of very few transistors, but was a nightmare to maintain. The follow-on PDP-10 was syncronous logic.
There must be some history out there somewhere of the problems DEC had with the asynchronous logic. Any old MIT research notes?