How many Sun machines are on people's desktops nowdays? Almost none, and declining rapidly. Let's face it, the desktop commercial marketplace is dead. Dead dead DEAD. Replaced by NT, or replaced by Linux. Which means that developers of Motif toolkits and interface builders are dead too -- unless something happens to make Motif popular on Linux.
Make no mistake about it: the "opening" of Motif for the Open Source operating systems was at the instigation of Motif toolkit and interface builder vendors, who saw their market evaporating. The only place where commercial Unix is used nowdays (with the exception of some legacy installations) is in the server room, and you can't sell too many Motif interface builders into the server room.
their FAQ is pretty specific about what their license is and isn't. It doesn't meet the terms of the Open Source Definition because it only applies when Motif is being run or distributed for Open Source operating systems (the OSD does not allow such a restriction). Other than that, it appears to be a descendent of the GPL -- e.g., you can make modifications and sell your modified version, but you must make the source available. As long as your modified version is for an Open Source operating system. (And yes, they have a very good definition for what they consider "Open Source" -- Solaris's "Community Source" need not apply!).
I suggest you read the FAQ. It *IS* free for commercial use -- on Open Source operating systems.
The FAQ says that the "on Open Source operating systems" part is why the license is not a "true" Open Source license. Still, given that every closed source Unix already *comes* with the Motif libraries, it's certainly not any big killer to anybody interested in writing Motif apps. (What? You want to write Motif apps for W2K? SICK!).
Yes, Microsoft employees are propogandized by their employer and generally believe their employer's bullshit about "Microsoft innovation" and "our job is to change the world". In that respect, Microsoft is like a Borg collective, whose members all share a common belief in the general rightness of what they are doing. But as for the working conditions part...from everything I've been told by former Microsoft employees, the working conditions are generally pretty reasonable. As you'd expect, since most of Microsoft's full-time employees (as vs. the PermaTemps) are now in their thirties -- yeah, they were hired right out of college, but that was 10 years ago!
Note that AMD would make their chips work in Intel motherboards in a split second, if Intel had not patented the socket and/or slots used on their motherboards for the CPU chip.
Remember the whole reason for Slot 1: Because AMD had come out with processors that fit in Intel's Socket 7 spec, and Intel needed a patented processor socket to keep AMD from doing that again.
at Enhanced Software Technologies in Phoenix, Arizona.
I jumped on my mountain bike this morning, rode out my front door, crossed the street, and was immediately in the midst of a mountain biker's wonderland of twisty single-track and everything from modest whoop-de-whoops to crazed insane climbs/downhills for the next three hours -- in the middle of the city! For 360 days a year! (It's raining the other five days, grin). Gosh, aren't you tired of spending all that time indoors in the rainy northwest?
We need a NT/W2K God, someone who knows NT/W2K internals on an intimate basis, and we need a GUI designer. And if Microsoft fired you for sending information to slashdot, that's a +5 on the algorithm used to score your resume:-).
> By being the company with the most power in its market, > it has no choice but to exercise that power
Oh puh-LEEZE. Having power does not mean you have to use it. Else we'd all be radioactive dust -- both the United States and the Soviet Union had the power to nuke us all until we glowed in the dark, and never used it.
Presumably the powers that be at Microsoft are human beings. (Unless Microsoft Research has in fact created an AI that is simulating the Bill'n'Balmer show for us!). If they are human beings, then they possess free will. If they possess free will, then they have a choice as to whether they use what power they possess. To say that they do not is to put human beings into the same class as sheep dogs, who are victims of their genetics (they have no choice as to whether they will herd sheep -- put a city-bred sheep dog into a pasture full of sheep, and he will herd them!).
Frankly, I have no respect for those who claim that they're not responsible for their own actions because "something else made me do it". I have no respect for the murderer who claims "my abusive mom and dad made me do it", and I have no respect for corporate executives who claim that acting ethically is not an option because "the corporate environment doesn't work like that." Are you a human being? Or are you a sheep dog? Sheesh.
The page size is a compile-time option in PostGreSQL 6.5 and below.
I don't know whether the limitation has been lifted in 7.0. It most certainly will be in 7.1, the guy in Russia who does the bottom end has been working on an entirely new storage manager.
One reason companies are reluctant to release the source code to their products is because some components may (or may not) be patented. With the current screwed-up state of the software patent biz, code that's been in your product for 10 years might violate some patent or another (no matter how many patent-savvy consultants look it over), and you'll never know it until the attorney for the patent-owner sues you for half a billion dollars.
A lot of closed-source vendors would feel a lot happier about releasing source for their products if they didn't fear getting their pants sued off for their efforts.
I've evaluated Interbase. It is roughly equivalent to PostGreSQL, a few % faster at writes, about the same speed for reads, about the same foot print, about the same feature set (each has a few features the other doesn't have, but no big deal). But the thing is, its source code is not currently available, so that rules it out for use as a production Open Source database -- at the moment.
Basically, the reason MySQL gets more press is because most current use of Open Source databases is in the realm of Internet services, and MySQL is more useful in a CGI environment. Like most "real" Unix databases, PostGreSQL takes a while to establish connections, whereas MySQL is super-quick at establishing connections. Once connections are established things even out a bit, but if you can only accept 8 connections per second because that's all PostGreSQL can open, you pretty much rule out PostGreSQL as your web site's back end.
MySQL is lightweight and very good at what it does. I feel somewhat ill, though, when I hear it called "SQL". It's not. It's a very limited subset of SQL, chosen in order to maximise speed, and it succeeds very well at that, thank you. But I wouldn't want to build an accounting package using it. Lack of transaction support alone, for example, immediately rules it out for "real" work.
You can write stored procedures in PL/SQL, TCL, Python, or "C". How complex do you want to get?!
Granted, this *DOES* need to be better documented. That's another lack of the PostGreSQL project -- very few of the advanced features are adequately documented. You should have seen how long it took me to get my first PL/SQL stored procedures working:-(.
I've been using PostGreSQL off and on since 1995, when it was a VERY buggy program called "Postgres95". With the 6.5 series, I believe it competes quite well with low-to-midrange databases from other vendors, such as, e.g., Raima. The speed is still about 10-20% slower, but for many applications that's not an issue, and the reliability and features *DEFINITELY* are there.
I am probably stepping out on a limb, but here it is: the EST development team evaluated a variety of embedded databases for use in BRU Professional, everything from plain old 'gdbm' to embedded Oracle. PostGreSQL beat them all on features and reliability, and lagged only slightly in performance. It has taken five years, but PostGreSQL is a winner.
Oh, my PostGreSQL wish-list:
Replication. Right now, replicating PostGreSQL databases is a pain in the @#$%@. I end up manually replicating them -- I have a 'modified' timestamp in every record, and my database routines timestamp that flag every time an 'insert' or 'update' is done to the database. Then I can do a simple query to pull out all the records changed between times and , and replicate that way. But that's a major pain in the @#$%@.
Hot Backups. Right now, we have to shut down PostGreSQL when it's time to back up the PostGreSQL databases themselves. Pain, pain pain.
Larger maximum record size. I believe 8192 bytes is the biggest record you can do with PostGreSQL at the moment. Doesn't bite me in my particular application, but someone who wants to keep big texts in PostGreSQL might be bitten. (This is being worked on, BTW).
Faster write speed! In fairness, this is being worked on. I haven't had a chance to benchmark PostGreSQL 7.0 yet, so I don't know whether the changes made it there.
*INTEGRATION WITH LDAP!*. I want PostGreSQL to be able to authenticate against a LDAP directory! This business of having both a Unix password, and a PostGreSQL password, is for the birds!
Someday. Someday. Oh well, back to doing encrypted network connections again (sigh... talk about something that's a pain in the #@$%@ to debug:-).
The Bill of Rights does not apply to businesses. It only applies to government.
There are various privacy acts and such prohibiting unauthorized wiretaps by private parties and such (just ask Linda Tripp:-), but it is unlikely that any such laws apply here. In any event, it's not a Constitutional issue, since the Bill of Rights restricts what government can do, not what businesses can do. The only restriction upon businesses are the laws passed by Congress and the individual states -- bodies which increasingly are owned by big businesses and which rarely today serve the people.
Granted, a non-profit corporation is engaged in corporation. Granted, free how-to guides on the Internet are directly in competition with IDG books. But: a gift is *NOT* commerce.
I suggest you quit trying to snow people with your shady two-stepping. If there is no exchange of goods, services, or money, there is no commerce. A gift (giving away something for free, with no good or service received in exchange) is *NOT* commerce.
This isn't rocket science. This is Business Law 101. That's why the relative renting my land in Louisiana pays me $1 per year for the privilige... otherwise, there would not be a valid contractual transaction, no commerce, and various nasty legal consequences that I won't go into here (issues of liability for his actions, basically).
Please note that commerce involves the exchange of goods and services. If it's free/freely available with no payment or exchange of goods or services involved, please tell me what definition of the word "commerce" you do not understand?
As for free stuff on the Internet being competition against printed "how to" guides published by IDG, awe, poor baby! Unfortunately, "competition" != "commercial". Without an exchange of goods, services, or money, there is no sale insofar as contract law is concerned, and if there is no sale, there is no commerce. A gift does *NOT* qualify as commerce, BTW, because it is one way (no exchange).
As I recall, Apple did have to pay a sum of money to McIntosh (the makers of the high end stereo equipment) due to the fact that McIntosh has a trademark on the name "McIntosh" as a name for home electronics equipment....
But if I want to put up a non-commercial site called "Macintosh Follies" or "Baiting IDG Lawyers for Dummies", that's protected use under the trademark law, because I am not using it in commerce -- i.e., there are no goods, services, or moneys changing hands.
Nice job of dodging the point, Mr. IDG Lawyer. The description of how to do screenprinting is non-commercial because there is no exchange of goods, services, or cash. You definitely stretch the point when you say that it somehow "competes" with the "for Dummies" series. You can't "compete" if no goods, services, or cash is being exchanged (look up the definition of "commercial", please).
Of course, we're talking about Australia here, which has totally different trademark laws, so it's rather irrelevant. Still, I'm tempted to put up an "IDG lawyer baiting for dummies" page on my web site with true-life examples of how to bait IDG lawyers into acting like idiots:-).
The only problem here is that LDAP is a pain in the #@$%@#$. Been there. Done that. Have the scars to show it (working on an LDAP-enabled application at the moment). Flexibility isn't always the end-all and be-all of everything... simply authenticating a user against an LDAP directory, for example, is a pain in the @Q#$%@ compared to making a simple 1-line SQL query.
Some of us need a stable kernel and Linux distribution because we need to get work done, we don't need to twiddle with the Linux kernel all day long. My router, for example, is still running the *2.0.37* kernel... there just hasn't been any reason to upgrade it, it just plain works, why mess with stuff that works?
Most of the changes in 2.2.15 were actually back-ported from the 2.3 series kernels. For example, I helped Kai with some changes to the 2.3 series tape driver, and those changes have been backported to the 2.2.x series kernels (but not in time for 2.2.15, sigh... maybe 2.2.16).
Both the successor agency to the KGB and the CIA both view "economic intelligence" (e.g. filching airplane plans from other countries) to be a major part of their mission now that the Cold War is mostly over (except for Elian!).
The big differences in the Buran design that I saw:
1) Buran has no engines. It uses that space for cargo instead.
2) Buran still uses the old tile method for heat shielding. The Space Shuttle no longer uses the original asbestos tile. Now they have a spray-on version that doesn't fall off like the tiles did.
DOesn't really matter, space shuttles are a loser's game anyhow right now. It costs too much to launch payload into orbit for most purposes -- big dumb rockets can be built for cheap to launch most payloads (except for people, who need safer handlikng). The only use for a space shuttle is for a manned space program, and even there, the Russians did just fine with "dumb" capsules for many years, building Mir without a shuttle even.
Buran, alas, is one of those great ideas that won't ever really fly.
The problem, though, is that neither instruction set has the classic symmetry and simplicity of the PDP-11 instruction set.
The Motorola 68000 was similar to the PDP-11 in its addressing modes and memory model, but it had that funky divide between data registers and address registers. For example, to do an indexed read off a pointer to an array, you'd load the pointer into an address register (like A1), then you'd load the offset into a data register (like D1), then
move (A1)[D1],D0
(the above is not the 68000 assembler's format, BTW, it's been too long:=-( ).
Anyhow, even with that address/data register split, it was still head and shoulders above the (blech) 8086. But the one I really wanted to get my hands on was the ?NCR?AMD??? 32032, there was a big write-up in Byte Magazine on the chipset and it made me slobber (even if I can't remember who made the stupid chip!). Had a MMU that implemented true paged virtual memory, had a symmetric instruction set that greatly resembled a VAX, etc... this was right after Motorola introduced the 68000, which had no MMU and thus really wasn't well suited for Unix. Unfortunately, the maker of the chip never managed to ship them in volume or with adequate performance. Kind of the same story as with Zilog and the Z8000, making microprocessors back then was a lot of hand-drawing masks and stuff, and many of the old-line companies just couldn't scale their design process to the "new" 16-bit microprocessors. Probably the only reason Motorola managed the 68000 was because they gave up and microcoded most of the instructions, and even then, the 68000 was late to market and thus missed the IBM design win (because IBM needed something available right then and there, and the 8088 was "good enough"). I still think we would be better off if Motorola had beat Intel to market... even the Pentium III and Xeon suffers from a serious lack of available processor registers (makes GCC's optimizer make aweful noises and die messily from time to time -- ask the kernel guys about all the work-arounds they've had to do when the optimizer craps on their code). One thing you could not accuse the 68000 of was a shortage of registers (it had 16 -- 8 address and 8 data, though 2 of the address registers were reserved for stack and program counter, and 1 of the address registers was usually used as an offset to the current stack frame).
Hate to tell you this, but a Vax/780 running BSD 4.2 crawled to a halt whenever it got above 20 users. Yeah, 30 people could log in, but they would be seriously uncomfortable if all 30 attempted to be active. I seriously doubt that a PDP-11 could support 30 timeshared users in any sort of comfort, even the big PDP-11 (the 11/65, was it?). Hold it, maybe it could, since everybody was using teletypes rather than terminals, and 'ed' rather than 'vi', but I'd hardly call that 'using':-).
Of course, even 20 timeshared users for tens of thousands of $$ would be a big deal when the mainframes STARTED at $1M.... and actually I think USL paid around $250K for their Vax 780, complete with two of the big 500Mb drives. That was more storage space than their $6M mainframe had at the time!
It wasn't Slashdot that was feeding the frenzy over the "back door". The mainline press did not interview the Taco. They interviewed NTBugTraq's big kahona.
I will state that most Open Source programmers had nothing to do with the feeding frenzy on Slashdot. A few "luminaries" did, but in general they acted upon what information was reported by Microsoft and NTBugTraq. Given that Microsoft itself was calling it a "back door", I can hardly fault ESR for putting out a long essay about the problem.
Finally: To accuse Open Source people of "corporatism" is silly. People who release code under the GPL do so that others *can't* take ownership and hide it from view, which is what corporatism is all about. Yes we get excited when we see our beliefs vindicated, but this has nothing to do with money. It is interesting that many former Microsoft employees, albeit working in other places hundreds of miles away from Redmond, will still defend their former employer, for the exact same reason: pride of ownership. It is "their" product, and they want to tell the world that it's good stuff and that those who criticize it are weenies. No Borg mind-washing required.
About the only lesson we can learn here is that there would have been no story if it were OSS. The Wall Street Journal would have contacted a local security guru, who would have looked at the source code of the module in question, and said "There's no back door there." No story. The only reason there was a story was because only one company had the source code to this module -- Microsoft -- and the Wall Street Journal had to rely on Microsoft's word. And Microsoft was saying it was a back door.
Slashdot Mirror
How many Sun machines are on people's desktops nowdays? Almost none, and declining rapidly. Let's face it, the desktop commercial marketplace is dead. Dead dead DEAD. Replaced by NT, or replaced by Linux. Which means that developers of Motif toolkits and interface builders are dead too -- unless something happens to make Motif popular on Linux.
Make no mistake about it: the "opening" of Motif for the Open Source operating systems was at the instigation of Motif toolkit and interface builder vendors, who saw their market evaporating. The only place where commercial Unix is used nowdays (with the exception of some legacy installations) is in the server room, and you can't sell too many Motif interface builders into the server room.
-E
-E
The FAQ says that the "on Open Source operating systems" part is why the license is not a "true" Open Source license. Still, given that every closed source Unix already *comes* with the Motif libraries, it's certainly not any big killer to anybody interested in writing Motif apps. (What? You want to write Motif apps for W2K? SICK!).
-E
-E
Remember the whole reason for Slot 1: Because AMD had come out with processors that fit in Intel's Socket 7 spec, and Intel needed a patented processor socket to keep AMD from doing that again.
-E
I jumped on my mountain bike this morning, rode out my front door, crossed the street, and was immediately in the midst of a mountain biker's wonderland of twisty single-track and everything from modest whoop-de-whoops to crazed insane climbs/downhills for the next three hours -- in the middle of the city! For 360 days a year! (It's raining the other five days, grin). Gosh, aren't you tired of spending all that time indoors in the rainy northwest?
We need a NT/W2K God, someone who knows NT/W2K internals on an intimate basis, and we need a GUI designer. And if Microsoft fired you for sending information to slashdot, that's a +5 on the algorithm used to score your resume :-).
Send me mail if you're interested!
-E
> it has no choice but to exercise that power
Oh puh-LEEZE. Having power does not mean you have to use it. Else we'd all be radioactive dust -- both the United States and the Soviet Union had the power to nuke us all until we glowed in the dark, and never used it.
Presumably the powers that be at Microsoft are human beings. (Unless Microsoft Research has in fact created an AI that is simulating the Bill'n'Balmer show for us!). If they are human beings, then they possess free will. If they possess free will, then they have a choice as to whether they use what power they possess. To say that they do not is to put human beings into the same class as sheep dogs, who are victims of their genetics (they have no choice as to whether they will herd sheep -- put a city-bred sheep dog into a pasture full of sheep, and he will herd them!).
Frankly, I have no respect for those who claim that they're not responsible for their own actions because "something else made me do it". I have no respect for the murderer who claims "my abusive mom and dad made me do it", and I have no respect for corporate executives who claim that acting ethically is not an option because "the corporate environment doesn't work like that." Are you a human being? Or are you a sheep dog? Sheesh.
-E
I don't know whether the limitation has been lifted in 7.0. It most certainly will be in 7.1, the guy in Russia who does the bottom end has been working on an entirely new storage manager.
-E
A lot of closed-source vendors would feel a lot happier about releasing source for their products if they didn't fear getting their pants sued off for their efforts.
-E
_E
MySQL is lightweight and very good at what it does. I feel somewhat ill, though, when I hear it called "SQL". It's not. It's a very limited subset of SQL, chosen in order to maximise speed, and it succeeds very well at that, thank you. But I wouldn't want to build an accounting package using it. Lack of transaction support alone, for example, immediately rules it out for "real" work.
-E
Granted, this *DOES* need to be better documented. That's another lack of the PostGreSQL project -- very few of the advanced features are adequately documented. You should have seen how long it took me to get my first PL/SQL stored procedures working :-(.
-E
I am probably stepping out on a limb, but here it is: the EST development team evaluated a variety of embedded databases for use in BRU Professional, everything from plain old 'gdbm' to embedded Oracle. PostGreSQL beat them all on features and reliability, and lagged only slightly in performance. It has taken five years, but PostGreSQL is a winner.
Oh, my PostGreSQL wish-list:
- Replication. Right now, replicating PostGreSQL databases is a pain in the @#$%@. I end up manually replicating them -- I have a 'modified' timestamp in every record, and my database routines timestamp that flag every time an 'insert' or 'update' is done to the database. Then I can do a simple query to pull out all the records changed between times and , and replicate that way. But that's a major pain in the @#$%@.
- Hot Backups. Right now, we have to shut down PostGreSQL when it's time to back up the PostGreSQL databases themselves. Pain, pain pain.
- Larger maximum record size. I believe 8192 bytes is the biggest record you can do with PostGreSQL at the moment. Doesn't bite me in my particular application, but someone who wants to keep big texts in PostGreSQL might be bitten. (This is being worked on, BTW).
- Faster write speed! In fairness, this is being worked on. I haven't had a chance to benchmark PostGreSQL 7.0 yet, so I don't know whether the changes made it there.
- *INTEGRATION WITH LDAP!*. I want PostGreSQL to be able to authenticate against a LDAP directory! This business of having both a Unix password, and a PostGreSQL password, is for the birds!
Someday. Someday. Oh well, back to doing encrypted network connections again (sigh... talk about something that's a pain in the #@$%@ to debug-E
There are various privacy acts and such prohibiting unauthorized wiretaps by private parties and such (just ask Linda Tripp :-), but it is unlikely that any such laws apply here. In any event, it's not a Constitutional issue, since the Bill of Rights restricts what government can do, not what businesses can do. The only restriction upon businesses are the laws passed by Congress and the individual states -- bodies which increasingly are owned by big businesses and which rarely today serve the people.
-E
I suggest you quit trying to snow people with your shady two-stepping. If there is no exchange of goods, services, or money, there is no commerce. A gift (giving away something for free, with no good or service received in exchange) is *NOT* commerce.
This isn't rocket science. This is Business Law 101. That's why the relative renting my land in Louisiana pays me $1 per year for the privilige... otherwise, there would not be a valid contractual transaction, no commerce, and various nasty legal consequences that I won't go into here (issues of liability for his actions, basically).
-E
As for free stuff on the Internet being competition against printed "how to" guides published by IDG, awe, poor baby! Unfortunately, "competition" != "commercial". Without an exchange of goods, services, or money, there is no sale insofar as contract law is concerned, and if there is no sale, there is no commerce. A gift does *NOT* qualify as commerce, BTW, because it is one way (no exchange).
-E
But if I want to put up a non-commercial site called "Macintosh Follies" or "Baiting IDG Lawyers for Dummies", that's protected use under the trademark law, because I am not using it in commerce -- i.e., there are no goods, services, or moneys changing hands.
-E
Of course, we're talking about Australia here, which has totally different trademark laws, so it's rather irrelevant. Still, I'm tempted to put up an "IDG lawyer baiting for dummies" page on my web site with true-life examples of how to bait IDG lawyers into acting like idiots :-).
-E
-E
Most of the changes in 2.2.15 were actually back-ported from the 2.3 series kernels. For example, I helped Kai with some changes to the 2.3 series tape driver, and those changes have been backported to the 2.2.x series kernels (but not in time for 2.2.15, sigh... maybe 2.2.16).
-E
Spooks don't disappear, they just change targets.
-E
1) Buran has no engines. It uses that space for cargo instead.
2) Buran still uses the old tile method for heat shielding. The Space Shuttle no longer uses the original asbestos tile. Now they have a spray-on version that doesn't fall off like the tiles did.
DOesn't really matter, space shuttles are a loser's game anyhow right now. It costs too much to launch payload into orbit for most purposes -- big dumb rockets can be built for cheap to launch most payloads (except for people, who need safer handlikng). The only use for a space shuttle is for a manned space program, and even there, the Russians did just fine with "dumb" capsules for many years, building Mir without a shuttle even.
Buran, alas, is one of those great ideas that won't ever really fly.
-E
The Motorola 68000 was similar to the PDP-11 in its addressing modes and memory model, but it had that funky divide between data registers and address registers. For example, to do an indexed read off a pointer to an array, you'd load the pointer into an address register (like A1), then you'd load the offset into a data register (like D1), then
move (A1)[D1],D0
(the above is not the 68000 assembler's format, BTW, it's been too long :=-( ).
Anyhow, even with that address/data register split, it was still head and shoulders above the (blech) 8086. But the one I really wanted to get my hands on was the ?NCR?AMD??? 32032, there was a big write-up in Byte Magazine on the chipset and it made me slobber (even if I can't remember who made the stupid chip!). Had a MMU that implemented true paged virtual memory, had a symmetric instruction set that greatly resembled a VAX, etc... this was right after Motorola introduced the 68000, which had no MMU and thus really wasn't well suited for Unix. Unfortunately, the maker of the chip never managed to ship them in volume or with adequate performance. Kind of the same story as with Zilog and the Z8000, making microprocessors back then was a lot of hand-drawing masks and stuff, and many of the old-line companies just couldn't scale their design process to the "new" 16-bit microprocessors. Probably the only reason Motorola managed the 68000 was because they gave up and microcoded most of the instructions, and even then, the 68000 was late to market and thus missed the IBM design win (because IBM needed something available right then and there, and the 8088 was "good enough"). I still think we would be better off if Motorola had beat Intel to market... even the Pentium III and Xeon suffers from a serious lack of available processor registers (makes GCC's optimizer make aweful noises and die messily from time to time -- ask the kernel guys about all the work-arounds they've had to do when the optimizer craps on their code). One thing you could not accuse the 68000 of was a shortage of registers (it had 16 -- 8 address and 8 data, though 2 of the address registers were reserved for stack and program counter, and 1 of the address registers was usually used as an offset to the current stack frame).
_E
Of course, even 20 timeshared users for tens of thousands of $$ would be a big deal when the mainframes STARTED at $1M.... and actually I think USL paid around $250K for their Vax 780, complete with two of the big 500Mb drives. That was more storage space than their $6M mainframe had at the time!
-E
I will state that most Open Source programmers had nothing to do with the feeding frenzy on Slashdot. A few "luminaries" did, but in general they acted upon what information was reported by Microsoft and NTBugTraq. Given that Microsoft itself was calling it a "back door", I can hardly fault ESR for putting out a long essay about the problem.
Finally: To accuse Open Source people of "corporatism" is silly. People who release code under the GPL do so that others *can't* take ownership and hide it from view, which is what corporatism is all about. Yes we get excited when we see our beliefs vindicated, but this has nothing to do with money. It is interesting that many former Microsoft employees, albeit working in other places hundreds of miles away from Redmond, will still defend their former employer, for the exact same reason: pride of ownership. It is "their" product, and they want to tell the world that it's good stuff and that those who criticize it are weenies. No Borg mind-washing required.
About the only lesson we can learn here is that there would have been no story if it were OSS. The Wall Street Journal would have contacted a local security guru, who would have looked at the source code of the module in question, and said "There's no back door there." No story. The only reason there was a story was because only one company had the source code to this module -- Microsoft -- and the Wall Street Journal had to rely on Microsoft's word. And Microsoft was saying it was a back door.
-E