The word length was 64-bits but the address space could be either 32 or 64-bits. Earliest OpenVMS/AXP implementaions came from the VAX so it was logical to start with a 32-bit address space. However Digital's Unix went 64-bit pure quickly as did Linux.
If you have source code, adapting for sizeof * = long long rather than just long wasn't a major problem but it could be tedious.
McBride is a stock manipulator, shyster, confidence trickester and a copyright pirate. He is probably an ill person. He is also deeply ignorant about computers, programs and operating systems.
First, there is no concept under US law called "Intellectual Property". There is patent law and copyright law and that is it. As far as trade secrets are concerned, that is a matter between contracting partners. McBride has no claim against partners with whom his company or their predecessors weren't contracted.
The problem is that the Opensource model means that companies must make money out of support. They can even compete to provide support as all have source access. SCO as a company (like Microsoft) do not provide good support. Actually, SCO don't provide anything these days except entertainment.
I actually started programming around the time that Unix was written. I was never at a place with source code access, as this was before that time. On the other hand I worked with other system software where I had source access and many have their own solutions to the problems of multiprocessing, reliable file systems and clustering. These solutions were from the early eighties before AT&T had anything in S4.
There is *nothing* special about Unix. Some things were so basic that two programmers working on different ends of the planet would come up with identical answers (they were both probably reading Knuth).
BSD was derived from Unix but was in many ways a parallel development. Linux is not Unix at all and even by McBride's standards until IBM/SGI started working on it, it was clean.
I wish the project luck. I really feel that airships have has a bad deal after Hindenburg. The Cargolifter project ended up annoying both the government who provided development aid and the public who provided much of the investment.
This summer, the Zeppelin NT was flying around. This is a smaller, primarily passenger carrying craft taking a total of 14 people and requiring no infrastructure other than a mobile mast.
The question is documentation. Brown worked for DSC for ten years. If he had documented his idea before starting and had dated it (the old post-marked sealed envelope thing is useful here, but a notary is better). Without evidence, it is just down to his word against the company's that it wasn't produced in the company's time.
I really don't understand Heinlein in your last point. Analog computers were in wide use during WW2 and after. Heinlein was in the navy and surely would have come across fire directors and torpedo computers. These provided coupling mechanisms so computation could be carried out directly from the sighting equipment. They were horrendously complex bits with funny gear wheels and motors but they worked and were also capable of limited motion prediction in the case of AA fire directors.
First forget people apart from the crew. The ill-fated Cargolifter project was looking at 100-tonne loads with minimal infrastructure.
They screwed up due to a number of issues but the technology was sound and there is still a market which can not be addressed by conventional transport. In particular the ability to heavy lift almost anywhere is useful to oil and mineral exploration companies, aid agencies as well as others.
I agree, to apply updates is relatively easy, however you would still be suprised how many don't. The problem is that if you have 4-5 hrs of downloads if you have high-speed access or much longer if you have slower access.
My point is that you are hit before you have updated your box. Locking down a brand-new XP system is possible but it isn't trivial.
A computer isn't the only complex device that I own which requires "Patches". I have a car. A car doesn't leave the showroom until it has been updated with the latest fixes. I don't quite want that but if Billg wants to sell PCs like televisions, then he should make sure they work like televisons. I want to see XP with the firewall pre-configured and unnecessary services locked down (we have been promised). I want to see systems shipped pre-patched to the date of manufacture and I want the updating of OS masters to be put on the OEM agreements.
The use of the weather-codes to provide 'cribs' for Enigma and Fish was well documented elsewhere as well as who was responsible for the work. Suffice to say that given the speed of the computers in 1942, exhaustive searches were really out of the question. However, to go from 'cribs' to the full code books for the day still represented an incredible amount of work and Turing was repsonsible for a lot of the mathematical basis.
On a broadband or DSL address pool it isn't unusual to see one attack attempt per minute.
The point is that XP is insecure out of the box. You shouldn't have to be an expert to update it. If you click 'windows update', why can the rest of the world break into your box? At least with current RH installs, it doesn't turn services on unless you request them and the built in firewall is enabled by default.
Given that MS aren't totally sure about which services are needed or not, locking down a Windows PC isn't easy. OK, you can turn of the web server, ftp, but what about all the other stuff. The firewall on XP isn't bad, but why doesn't it default to 'on'?
On the subject of free 'C', well we did have Tiny-C. But this was as the name suggests a very 'small' implementation. As mentioned a lot of substandard compilers appeared but as they weren't open source, few really had the chance to polish them. Commercial implementations of Unix were cursed with very expensive development kits. One key issue with GCC was that it was easy to rehost and retarget across 32-bit architectures. I even used it on a VAX under VMS.
EGCS was a fork of GCC, it was *not* a new compiler. A major fork, but a fork nevertheless which came out of a number of development streams. The mainline gcc continued because this was pushing stability. When egcs became stable it was then merged into the mainline gcc. This is how open source projects evolve.
If you rewrite libc from scratch then you can license it how you want. However, glibc wasn't a rewrite. It was much improved but that is all. I read the linked page, and to be honest I think Drepper is perhaps worrying too much.
On the subject of 'snatching' it really comes down to where people want to get their code from. If you want a fork which you have control over - there is nothing to stop you doing so.
The FSF is a kind of umbrella organisation and GNU is a movement. If you don't want to put the "G" in front, you can choose not to as the original author. However if the project already exists, then usually the original author is often credited with ownership of new modules for copyright purposes, even if the person didn't write them.
The why is simple. If some 'idiot' rips off my GPLed code, how can I prosecute this if I have to get every contributor to say that they agree that the idiot should go to court.
The thing is with glibc is that it is a direct offshoot of the original work done on gcc. RMS doesn't throw his weight around about stuff that isn't in his sphere, but there he does. If someone screws-up with glibc licensing, that may mean a big hack on gcc to get it to work with a different library.
However active Stallman is with gcc now is largely academic. However, it is the basis of all the following work on the GPL.
No the problem is that if I have a WiFi router and an ADSL connection. For reasonable P2P performance, I need a port which means action on my part. We can't just say that someone has hooked up to my access point. The poster before you had a point it is possible to do *some* sharing without an exposed port - but performance tends to suck.
An open AP isn't really a major problem if you have it hanging off a DMZ. WiFi security is kind of sucky anyway, so it would be reasonable to separate those systems so they can only SSL to your server but they can talk to the ADSL line without problems.
In the end, the question comes down what the court thinks is reasonable. The problem is that the RIAA/MPAA have been using civil suits.
First of all without GNU, there wouldn't have been a GCC and I haven't exactly seen a lot of other choices floating around. There were a *lot* of really crap compilers (every other CS student's undergraduate project) that then seemed to be sold.
GCC worked because of the GPL. Cygnus did a lot of work on it, but they didn't write it. Drepper, if anything, is only talking about the C library which has changed in major ways over the years. RH are doing a lot of work on GCC, but so are very many other people. Without Stallman's development model and his emphasis on portability in the original design, it wouldn't have happened.
I did some hacks on GCC many years ago (early 2.x) to fix some issues with a port and whilst a lot of people had contributed - it was clear that structure came from Stallman.
I can be attacked and have my machine taken over, even before I have patched it up to date. XP is horribly broken out of the box. In the time it takes to get it 'up to date' and AV software installed, it may be compromised.
In order to expose a particular port for the Internet (needed by many P2P protocols for proper function) you must configure a route through your router/NAT. That is if I must expose, say port 9999 for P2P, I must tell my router that all requests to 9999 go to 192.168.2.15:9999. This means that not only must I leave my WiFi AP open, I have to have a route configured as well, so the AP must be reprogrammable. Again, if I wanted to appear a total idiot, I *could* have left the AP open for programming.
It isn't easy to be anonymous now. The Islamic (Hawala) remittance system uses personal trust between intermediaries. As little cash as possible is actually shipped, rather an informal netting mechanism is used at each end to minimise the actual cash transferred. The old cash in a briefcase isn't really guaranteed now over a few thousand.
If you outsource your software, you are placeing your company's balls on a plate. Generally speaking, you protect your interest by speccing your system as component subsystems and then giving different subsystems to different outsourcing companies and you audit the product.
Actually, having work for programmers in these countries keeps them out of trouble. Very few people would *want* to work with the mafya and with that kind of money, they would demand to be involved, whether the programmer likes it or not.
The local law enforcement people aren't that sophisticated. If you have that kind of knowledge, chances are you are working with a reasonable pay check.
This has changed. There are now fields in electronic money transfers for reporting the true beneficiary. It is easier to trace but you really need a cooperative law engorcement officer to open the records.
Another trick (to borrow one from the feds) is to suggest that extortion of this magnitude may be from terrorists. Many such groups do offer 'protection' as a sideline.
Until the popular operating systems are locked down sufficiently, we will continue to see zombies. While there are zombies, it is difficult but not impossible to trace who is initiating the attack.
Rate limiting SYN packets is one answer, but you can DDOS someone just with HTTP GETs if you have enough machine. Just ask a recent/. effect victim.
The other thing is to just follow the money. This is where the FBI come in. It is *very* difficult now to make a transfer of more than a few thousand dollars through the banking system anonymously. Ironically, the only way that works are the informal methods used by overseas workers (and Al Quaida) to send cash home.
You would pay a whole lot less than $40K for a local hitman. About $10K tops for Russia and about $2K for the Ukraine. However you would have to find the right target to hit.
Unless you live in Alaska, you may notice that leaving desktops running contributes a fair bit of heating to the room. In summer that means pumping that 100 watts outside, which itself is taking energy. Take that from the home to the office where there are hundreds of PCs then the heat output is extremely important.
The thing is that data is stored in a mechanism whereby charge is topped up because of leakages but never dumped. Effectively the power usage (and thus the heat) should then be minimal, being the amount needed to maintain the charge and the switching losses.
Actually I have seen asynch logic used for non-trivial stuff - how about engineering design? Of course, it was in 1976 and that computer was to be retired in favour of a synchronous machines. Many early computers used asynchronous design because of their physical size. This particular one was an Atlas.
Slashdot Mirror
If you have source code, adapting for sizeof * = long long rather than just long wasn't a major problem but it could be tedious.
Actually, when Digital looked like they were going belly up, a bunch of the engineers went to AMD. I therefore have high hopes for Opteron/Athalons.
First, there is no concept under US law called "Intellectual Property". There is patent law and copyright law and that is it. As far as trade secrets are concerned, that is a matter between contracting partners. McBride has no claim against partners with whom his company or their predecessors weren't contracted.
The problem is that the Opensource model means that companies must make money out of support. They can even compete to provide support as all have source access. SCO as a company (like Microsoft) do not provide good support. Actually, SCO don't provide anything these days except entertainment.
I actually started programming around the time that Unix was written. I was never at a place with source code access, as this was before that time. On the other hand I worked with other system software where I had source access and many have their own solutions to the problems of multiprocessing, reliable file systems and clustering. These solutions were from the early eighties before AT&T had anything in S4. There is *nothing* special about Unix. Some things were so basic that two programmers working on different ends of the planet would come up with identical answers (they were both probably reading Knuth).
BSD was derived from Unix but was in many ways a parallel development. Linux is not Unix at all and even by McBride's standards until IBM/SGI started working on it, it was clean.
Yes, but this does not mean a 'pump and dump' for his personal short term enrichment. Generally speaking the SEC do not look kindly upon such actions.
This summer, the Zeppelin NT was flying around. This is a smaller, primarily passenger carrying craft taking a total of 14 people and requiring no infrastructure other than a mobile mast.
The thing that got me was how quiet they are.
The question is documentation. Brown worked for DSC for ten years. If he had documented his idea before starting and had dated it (the old post-marked sealed envelope thing is useful here, but a notary is better). Without evidence, it is just down to his word against the company's that it wasn't produced in the company's time.
Perhaps he just didn't see A/D position decoders?
They screwed up due to a number of issues but the technology was sound and there is still a market which can not be addressed by conventional transport. In particular the ability to heavy lift almost anywhere is useful to oil and mineral exploration companies, aid agencies as well as others.
My point is that you are hit before you have updated your box. Locking down a brand-new XP system is possible but it isn't trivial.
A computer isn't the only complex device that I own which requires "Patches". I have a car. A car doesn't leave the showroom until it has been updated with the latest fixes. I don't quite want that but if Billg wants to sell PCs like televisions, then he should make sure they work like televisons. I want to see XP with the firewall pre-configured and unnecessary services locked down (we have been promised). I want to see systems shipped pre-patched to the date of manufacture and I want the updating of OS masters to be put on the OEM agreements.
The use of the weather-codes to provide 'cribs' for Enigma and Fish was well documented elsewhere as well as who was responsible for the work. Suffice to say that given the speed of the computers in 1942, exhaustive searches were really out of the question. However, to go from 'cribs' to the full code books for the day still represented an incredible amount of work and Turing was repsonsible for a lot of the mathematical basis.
The point is that XP is insecure out of the box. You shouldn't have to be an expert to update it. If you click 'windows update', why can the rest of the world break into your box? At least with current RH installs, it doesn't turn services on unless you request them and the built in firewall is enabled by default.
Given that MS aren't totally sure about which services are needed or not, locking down a Windows PC isn't easy. OK, you can turn of the web server, ftp, but what about all the other stuff. The firewall on XP isn't bad, but why doesn't it default to 'on'?
EGCS was a fork of GCC, it was *not* a new compiler. A major fork, but a fork nevertheless which came out of a number of development streams. The mainline gcc continued because this was pushing stability. When egcs became stable it was then merged into the mainline gcc. This is how open source projects evolve.
If you rewrite libc from scratch then you can license it how you want. However, glibc wasn't a rewrite. It was much improved but that is all. I read the linked page, and to be honest I think Drepper is perhaps worrying too much.
On the subject of 'snatching' it really comes down to where people want to get their code from. If you want a fork which you have control over - there is nothing to stop you doing so.
The why is simple. If some 'idiot' rips off my GPLed code, how can I prosecute this if I have to get every contributor to say that they agree that the idiot should go to court.
The thing is with glibc is that it is a direct offshoot of the original work done on gcc. RMS doesn't throw his weight around about stuff that isn't in his sphere, but there he does. If someone screws-up with glibc licensing, that may mean a big hack on gcc to get it to work with a different library.
However active Stallman is with gcc now is largely academic. However, it is the basis of all the following work on the GPL.
An open AP isn't really a major problem if you have it hanging off a DMZ. WiFi security is kind of sucky anyway, so it would be reasonable to separate those systems so they can only SSL to your server but they can talk to the ADSL line without problems.
In the end, the question comes down what the court thinks is reasonable. The problem is that the RIAA/MPAA have been using civil suits.
First of all without GNU, there wouldn't have been a GCC and I haven't exactly seen a lot of other choices floating around. There were a *lot* of really crap compilers (every other CS student's undergraduate project) that then seemed to be sold.
GCC worked because of the GPL. Cygnus did a lot of work on it, but they didn't write it. Drepper, if anything, is only talking about the C library which has changed in major ways over the years. RH are doing a lot of work on GCC, but so are very many other people. Without Stallman's development model and his emphasis on portability in the original design, it wouldn't have happened.
I did some hacks on GCC many years ago (early 2.x) to fix some issues with a port and whilst a lot of people had contributed - it was clear that structure came from Stallman.
I can be attacked and have my machine taken over, even before I have patched it up to date. XP is horribly broken out of the box. In the time it takes to get it 'up to date' and AV software installed, it may be compromised.
In order to expose a particular port for the Internet (needed by many P2P protocols for proper function) you must configure a route through your router/NAT. That is if I must expose, say port 9999 for P2P, I must tell my router that all requests to 9999 go to 192.168.2.15:9999. This means that not only must I leave my WiFi AP open, I have to have a route configured as well, so the AP must be reprogrammable. Again, if I wanted to appear a total idiot, I *could* have left the AP open for programming.
It isn't easy to be anonymous now. The Islamic (Hawala) remittance system uses personal trust between intermediaries. As little cash as possible is actually shipped, rather an informal netting mechanism is used at each end to minimise the actual cash transferred. The old cash in a briefcase isn't really guaranteed now over a few thousand.
Actually, having work for programmers in these countries keeps them out of trouble. Very few people would *want* to work with the mafya and with that kind of money, they would demand to be involved, whether the programmer likes it or not.
The local law enforcement people aren't that sophisticated. If you have that kind of knowledge, chances are you are working with a reasonable pay check.
Another trick (to borrow one from the feds) is to suggest that extortion of this magnitude may be from terrorists. Many such groups do offer 'protection' as a sideline.
Rate limiting SYN packets is one answer, but you can DDOS someone just with HTTP GETs if you have enough machine. Just ask a recent /. effect victim.
The other thing is to just follow the money. This is where the FBI come in. It is *very* difficult now to make a transfer of more than a few thousand dollars through the banking system anonymously. Ironically, the only way that works are the informal methods used by overseas workers (and Al Quaida) to send cash home.
You would pay a whole lot less than $40K for a local hitman. About $10K tops for Russia and about $2K for the Ukraine. However you would have to find the right target to hit.
The thing is that data is stored in a mechanism whereby charge is topped up because of leakages but never dumped. Effectively the power usage (and thus the heat) should then be minimal, being the amount needed to maintain the charge and the switching losses.
Actually I have seen asynch logic used for non-trivial stuff - how about engineering design? Of course, it was in 1976 and that computer was to be retired in favour of a synchronous machines. Many early computers used asynchronous design because of their physical size. This particular one was an Atlas.