In general use there is RMS (Root Mean Square) used to measure "average" (ie effective) voltage. For arbitrary k, you have the k-th root of average of x**k. The limit as k approaches infinity is the familiar mid-point. The limit as k approaches 0 should be the median (I'm too lazy to check it). The familiar (median, mean, RMS, midpoint) correspond to k=(0, 1, 2, infinity). The other values (not necessarily integral) are not in general use. The mode is not so tractable. Very slight change to a bi-modal distribution gives large difference in mode.
Amen from the chorus. Red Hat is taking the point, the cutting (bleeding?) edge. (but not as bloody as NT;-) Red Hat is marketing Linux to the suits. World Domination? Looks more and more plausible, but it is *NOT* a Red Hat domination. It is not a Debian domination. It is not a FreeBSD domination. It a domination of computers, of computers that work, of computers that do what they are supposed to do. It is a domination of computer by mankind, not the other way around. Maybe I'm dreaming, or hoping, but this keeps looking better and better.
OK, I'll try a shot at putting this into perspective. This is a bit exagerated for effect. This is coming from second and third generation mainframe experience. With mainframes, or other big iron, the situation is "little computer, big problem", so you throw everything you can afford at it. With PCs, the situation is "big computer, little problem", more like a toy. Current PCs have much more power than second generation mainframes, but they do not come close to doing what was done on those second generation mainframes. Mainframes attract what you *must* do. You have to solve all the problems. PCs attract what you *want* to do. You solve a few easy or interesting problems. This doesn't automatically make a mainframe OS better or more reliable, but there is a suble distinction in what is a bug and what is an unacceptable bug. Above all, a mainframe OS must perform under load, heavy load. With a PC, the idea of load doesn't really come up. There are real differences and apparent differences and hidden differences, with no good way to sort them out. These differences cannot be seen from the PC side. They exist in places where PCs don't have places. It is more subtle than features. NT has features,... well the appearance of features. The commercial Unices are pretty well proven under load. Linux is unproven, but that situation is changing. Beowulf proved a few things. Hope this helps a bit. Hang on to your hat, red or otherwise, these are interesting times.
Consider this possibility. IBM, Oracle, SAP et al never see a dime return on their investment in Red Hat. Any direct return on their investment would be insignificant on their bottom line. If they are able to depend on a stable, supported, and supportable OS, the impact on their bottom line can be enormous. This looks a bit like IBM, Oracle, SAP et al doing to Microsoft what Microsoft did to Netscape. Red Hat is not a non-profit organization, at least not yet, but it does act somewhat like one. At any rate it is *not* a get-rich-quick scheme. The only problem I can see with Open Source, Free Source, or whatever you call it, is how to fund it. These big boys may have figured out that it is to their advantage that it *be* funded, and decided to chip in. There is a lot of work still to be done, methinks, dull, boring and expensive. Think of Red Hat as carrying the Linux flag for the benefit of corporate management, PHBs, and suits. The "fragmentation" of un*x is more apparent than real, but will bother corporate types. (Compare libc versions to DLL versions -- now exactly *which* version of windows are you running? If you need incompatible versions running together, nothing like statically linked libraries -- do I have that right?) Red Hat becoming another M$? With Cheapbytes, "knock-off" distros, and free downloads, no way. Besides, the big boys would *not* like it. One mild caution. This cannot be a "change of heart" of the big boys. I cannot imagine the big boys willingly supporting a "hacker's OS". They are not interested in free (speech *or* beer). However, there is a large common ground in having a base level that is *not* wonky, *not* flakey, *is* well known and understood. If it is not messed up too badly, everybody (except Microsoft) gains. Yep, these are exciting times.
Sorry, I didn't make myself clear enough. The railroad, which has a monopoly on transportation, goes into the grain business where is does not have a monopoly. The railroad by various means can drive drive out any competition in the grain business. The grain producers play ball or have NO transportation to market.
>Microsoft has EVERY right to lean on PC manufacturers that want to ship other products. It's THEIR products! If Microsoft is selling to the public, their rights to discriminate are severly restricted.
The monopoly per se is not bad, but it does open certain possibilities for abuse. Lets say the railroad decides to go into the grain business. If its competitors have strange problems transporting grain, the railroad has an unfair advantage and can drive its competitors out. Legitimate businesses in a monopoly or oligarchy (i think that's the right word) situation tread *very* carefully to avoid even the appearance of wrongdoing. Microsoft does not (quite) have a monopoly on Operating Systems or on applications. Microsoft does have a monopoly on the average consumer "Experience". When Microsoft leans on a pc manufacturer who distributes Corel Office Suite, it is over the line of acceptable practice.
Agreed, but Microsoft will *never* go along. Microsoft doesn't even know their own APIs, although they do know rather more than is published. The rules for monopolies *should* be different.
>Does anyone else get the feeling that this whole DOJ thing is an exercise in futility? Maybe so, but letting it go unchallenged is even worse. I still remember the cartoon of someone reading a newspaper headline of the start of trial captioned "There *is* a god."
>Microsoft has played dirty for too long, and people resent it. Fires are popping up left and right now, and Bill can't put all of them out at the same time any more. Very true. The DOJ case by itself will probably *not* fix the situation. Nor the states (somehow this Auctioning off Windows Source looks like a hoax). The key is a *lot* of fires, some will fizzle, Bill will put some of them out. With enough light thrown on the subject, more and more people will see the microsoft bubble as a bubble. Just hope there are enough penquins around to help pick up the pieces.
This could be a real embarassment for Microsoft. I suspect the real value of the code base is well under $.02. Somehow this whole thing smells like a hoax. From Seattle Time Sunday edition. Dateline SEATTLE (CNNfn).
Microsoft does not support it anyway. Read the EULA.
While I am on a rant, WINE is a good idea, if for no other reason that "because it's there", to see how far the idea can be pushed. It will even run some (most?) legacy Windows aps. Having the source to Windows would help some, maybe. Emulator running simulator running legacy ap. It's been done before.
Best bet is good clean aps for Linux (and thence *BSD and other un*xes).
>... Microsoft's own engineers too scared to fix the base of the code itself, so every patch is an extension to the current code base. I have been expecting and looking for this, but this is the first reference I've come across. I've seen the phenomenon before, from a distance. It is the kiss of death, not quick and easy, but slow and agonizing. -- "The evil that men do lives on after them. The good is oft interred with their bones"
>A good unix-compatible but newer-implemented OS will come along rather soon (I hope) and kill all others. At the moment, that OS *is* Linux. AT&T has always diligently protected the trademark UNIX. There was a long and bitter battle between Berkeley and AT&T over the independently derived BSD Un*x. Linux is an independent implementation of the Un*x standard. Because of the trademark and copyrights, you have *BSD and Linux, not UNIX. When that better OS comes along, the good parts will find their way into Linux and into *BSD. From the introduction of FreeBSD. "For copyright reasons, FreeBSD may not be called UNIX. You be the judge of how much difference this makes." Linux *has* to insist that Linux is not UNIX. GNU, GNU's Not Unix. Same thing. Linux *is* UNIX. You just can't call it UNIX(tm).
>The world is advancing at an alarming rate, especially in the computer industry. Right. It's called Windows(tm). It is alarming.;-)
UNIX will still be obsolete thirty years from now. You will still have the requisite press announcements of UNIX's demise. AT&T ownned it and couldn't kill it. Microsoft can't kill it. IBM/TransMeta will not kill it.
OK, I'll try a stab at this. First, bugs are not created equal. Formal testing tries to use metrics, but this distorts reality. New user syndrome. A new user will stumble into bugs a more experienced user would never encounter. Open Source has at least two effects. First, the code is probably better thought out initially, otherwise it is flame-bait. Commercial quality code, at least IBM from long ago, is NOT that good, and from a few check points has actually deteriorated with time. Second, there should be a few people out who take a pathological delight in finding and eradicating certain types of bugs. With the source, they know where to look and what to look for. For the bugs that matter, somebody competent out there will run into it and fix it. In general, the user community will be as helpful as possible. Bugs are diverse. The user community is diverse and beats on it in a variety of ways. Open software will not catch all the bugs, but will tend to catch all of them that matter.
Closed software testing will start out with good intentions, but the same old same old tends more toward preservation of original bugs than their elimination. Further, there is an assumption that it's tested so it must be right. In addition the programmer wants his boss to think he did a good job. The whole mess creates a sort of arrogance so that if a user does find a bug, it's far easier to just try to bypass it. It's just not worth the hassle. Even if it does make it onto the list of things to be fixed, it will be months at least before the user will see the results.
In contrast, with Open Source, it is worth the hassle. The user gets his bug fixed, along with any fixes for other bugs. At this point, with a sensitized target, with communication lines open, you get the best regression testing imaginable. Everybody gains.
In short, with Closed Source, the forces are aimed at burying or trying to ignore bugs. With Open Source, the forces are aimed at 1) ameliorating effects of bugs and 2) exposure of bugs. Further, Closed Source is effectively paid by the line. There is a strong disincentive to take the effort to shorten and tighten the code.
Release early. Release often. Catch the really bad actor early. Cure the bug twice. First the bug itself. Second, fix it so the bug can't really hurt you. Another rule. Bugs are like snakes, they go around in pairs, so find the other one. It helps to be paranoid.
Be aware of the level. Cool on the desktop is ok. Cool for non-critical stuff is fine. Cool for critical stuff, particularly the kernel, is not ok. As I understand it, Linus Torvalds is a very nice guy. In the book excerpt from Open Sources, he uses the word "stupid" 4 times. The more I think about it, the more I am convinced that he is exactly right. At this level, it is much more important to not be stupid than it is to be brilliant. If you look at it the right way and it is stupid, it is stupid. Remember it. "The very basic rule is to avoid interfaces." This is dead on target. You cannot avoid them entirely, but less is better. Hmmm, maybe this is why Un*x cannot be killed off. "While the NT team knew the final result wouldn't approach a microkernel, they knew they had to pay lip service to the idea." And this starts the chain of deception. At the level of the kernel, and other critical stuff, you need to be [brutally] honest.
Crystal ball. (He who lives by the crystal ball shall learn to enjoy ground glass) Beware Windows2000 (W2K).
If you want to be able to recover contents of pre Service Pack 3 NT4, it is pretty much essential. Furthermore, you make an NT boot floppy. Format under NT. Copy to floppy files BOOT.INI NTDETECT.COM NTLDR NTBOOTDD.SYS Curiously, booting NT from floppy can work when NT cannot boot directly from Hard Drive.
>THAT is so wrong on SO many levels,... I agree, BUT am unwilling to risk losing access to the data. PS. I also hit the power switch when it gets its brains fried. Safer to not let it try to preserve the damage on hard drive.
cartoon. Feb 28,99. microwaving CDs. AOL makes an art form. NT crashed the microwave. I'm still laughing. (Microsoft does not appreciate unauthorized release of sensitive development information)
As I understand it, if the employee has the right to take the binary home with her/him, that employee has the right to take the source home with him/her. I think in general, the employee would not have either right. The right to the binary and the source belongs to the company. The company has no obligation to release anything to anybody. You can sell a modified GPL program, but the recipient has the rights to source and is free to modify, sell or give away the program, subject to the GPL. You cannot even give away a modified binary-only without making the source available. Realistically, you want to send back to the community any bug fixes or generally useful extensions. So much easier the next time around. Anything proprietary^H^H^H^H^H^H^H^H^H^H^H private goes nowhere further. The employees have no more rights than the general public. I think the key to understanding it is that the company itself, not the employees, owns the binaries. The above opinions are my own. GPL has no responsibility for them.
Explains a lot. Now, with all levels trying to put the best face on things, try to imagine how bad the W2K bug (or is it virus? nah, too big for a virus) really is.;-)
Ignore him and you will get more of him. He probably believes his own hype. What do you think his divisional managers tell him? What do their departmental managers tell them? Etc? (However the chain of command goes.) He is more like a carnival barker who has gotten the public (and PHBs who believe/hope the Gates magic will rub off) to believe. Wow! Look what I can do! (Just don't look too close. Don't stress it). Silence is affirmation. Satire and ridicule, a constant barrage, is the only solution. Seriously, would YOU want to be Bill Gates? Not me.
Contrast Bill with Linus. In a recent post, Linus is using the term "stupid" rather often. This is the same "stupid" as in KISS (Keep It Simple, STUPID). Translated it means "Do not outsmart yourself". The targets of KISS are not dumb or what would normally be thought of as stupid. The object is to try to keep unknown unknowns from killing people. Murphy's Law rules. Mother Nature sides with the hidden. Yeah, I know....flaw. That's because it's almost always the flaws that are hidden.
It is not as simple as it works or it doesn't. There is a vast spread between (among?) 1) it never works, by any definition of works 2) it sorta works some of the time 3) it usually works, for most things 4) works, except for exceptional cases 5) never fails, except isolated freaks 6) never fails any circumstances Bill's definition of works is about #3, aiming for #2. Linus' definition of works is #5 aiming for #6. You can count on the following: Windows is worse than you think it is. Linux is better than you think it is.
Microsoft has two things going for it. FUD and the BIG LIE. Combat FUD with information. Combat the BIG LIE with, preferably, the TRUTH, but in any event with something. Say it often enough, loud enough, and people will believe. Personally, I like to be able to, any day, any hour, go to slashdot and (live in living color) see what the people who understand computers are saying. BTW, I am a Windows (ab)user.
If Bill Gates tells lies about us, we will tell the truth about him.
Slashdot Mirror
In general use there is RMS (Root Mean Square) used to measure "average" (ie effective) voltage.
For arbitrary k, you have the k-th root of average of x**k. The limit as k approaches infinity is the familiar mid-point. The limit as k approaches 0 should be the median (I'm too lazy to check it). The familiar (median, mean, RMS, midpoint) correspond to k=(0, 1, 2, infinity). The other values (not necessarily integral) are not in general use.
The mode is not so tractable. Very slight change to a bi-modal distribution gives large difference in mode.
Amen from the chorus. ;-)
Red Hat is taking the point, the cutting (bleeding?) edge. (but not as bloody as NT
Red Hat is marketing Linux to the suits.
World Domination? Looks more and more plausible, but it is *NOT* a Red Hat domination. It is not a Debian domination. It is not a FreeBSD domination. It a domination of computers, of computers that work, of computers that do what they are supposed to do. It is a domination of computer by mankind, not the other way around.
Maybe I'm dreaming, or hoping, but this keeps looking better and better.
OK, I'll try a shot at putting this into perspective. This is a bit exagerated for effect. This is coming from second and third generation mainframe experience. ... well the appearance of features.
With mainframes, or other big iron, the situation is "little computer, big problem", so you throw everything you can afford at it.
With PCs, the situation is "big computer, little problem", more like a toy.
Current PCs have much more power than second generation mainframes, but they do not come close to doing what was done on those second generation mainframes.
Mainframes attract what you *must* do. You have to solve all the problems.
PCs attract what you *want* to do. You solve a few easy or interesting problems.
This doesn't automatically make a mainframe OS better or more reliable, but there is a suble distinction in what is a bug and what is an unacceptable bug. Above all, a mainframe OS must perform under load, heavy load. With a PC, the idea of load doesn't really come up. There are real differences and apparent differences and hidden differences, with no good way to sort them out. These differences cannot be seen from the PC side. They exist in places where PCs don't have places. It is more subtle than features. NT has features,
The commercial Unices are pretty well proven under load. Linux is unproven, but that situation is changing. Beowulf proved a few things.
Hope this helps a bit. Hang on to your hat, red or otherwise, these are interesting times.
Consider this possibility. IBM, Oracle, SAP et al never see a dime return on their investment in Red Hat. Any direct return on their investment would be insignificant on their bottom line. If they are able to depend on a stable, supported, and supportable OS, the impact on their bottom line can be enormous. This looks a bit like IBM, Oracle, SAP et al doing to Microsoft what Microsoft did to Netscape. Red Hat is not a non-profit organization, at least not yet, but it does act somewhat like one. At any rate it is *not* a get-rich-quick scheme.
The only problem I can see with Open Source, Free Source, or whatever you call it, is how to fund it. These big boys may have figured out that it is to their advantage that it *be* funded, and decided to chip in. There is a lot of work still to be done, methinks, dull, boring and expensive.
Think of Red Hat as carrying the Linux flag for the benefit of corporate management, PHBs, and suits. The "fragmentation" of un*x is more apparent than real, but will bother corporate types. (Compare libc versions to DLL versions -- now exactly *which* version of windows are you running? If you need incompatible versions running together, nothing like statically linked libraries -- do I have that right?)
Red Hat becoming another M$? With Cheapbytes, "knock-off" distros, and free downloads, no way. Besides, the big boys would *not* like it.
One mild caution. This cannot be a "change of heart" of the big boys. I cannot imagine the big boys willingly supporting a "hacker's OS". They are not interested in free (speech *or* beer). However, there is a large common ground in having a base level that is *not* wonky, *not* flakey, *is* well known and understood. If it is not messed up too badly, everybody (except Microsoft) gains.
Yep, these are exciting times.
http://www.root.org/melissa_virus.txt- Mar-99/melissa.macro.virus.txt
... don't show anyone else ;-)
http://www.genocide2600.com/~tattooman/exploits
Subject:
Important Message From
Body:
Here is that document you asked for
a bit obvious, don't you think.
Sorry, I didn't make myself clear enough. The railroad, which has a monopoly on transportation, goes into the grain business where is does not have a monopoly. The railroad by various means can drive drive out any competition in the grain business. The grain producers play ball or have NO transportation to market.
>Microsoft has EVERY right to lean on PC manufacturers that want to ship other products. It's THEIR products!
If Microsoft is selling to the public, their rights to discriminate are severly restricted.
The monopoly per se is not bad, but it does open certain possibilities for abuse.
Lets say the railroad decides to go into the grain business. If its competitors have strange problems transporting grain, the railroad has an unfair advantage and can drive its competitors out.
Legitimate businesses in a monopoly or oligarchy (i think that's the right word) situation tread *very* carefully to avoid even the appearance of wrongdoing.
Microsoft does not (quite) have a monopoly on Operating Systems or on applications. Microsoft does have a monopoly on the average consumer "Experience". When Microsoft leans on a pc manufacturer who distributes Corel Office Suite, it is over the line of acceptable practice.
Agreed, but Microsoft will *never* go along.
Microsoft doesn't even know their own APIs, although they do know rather more than is published. The rules for monopolies *should* be different.
>Does anyone else get the feeling that this whole DOJ thing is an exercise in futility?
Maybe so, but letting it go unchallenged is even worse. I still remember the cartoon of someone reading a newspaper headline of the start of trial captioned "There *is* a god."
>Microsoft has played dirty for too long, and people resent it. Fires are popping up left and right now, and Bill can't put all of them out at the same time any more.
Very true. The DOJ case by itself will probably *not* fix the situation. Nor the states (somehow this Auctioning off Windows Source looks like a hoax). The key is a *lot* of fires, some will fizzle, Bill will put some of them out. With enough light thrown on the subject, more and more people will see the microsoft bubble as a bubble. Just hope there are enough penquins around to help pick up the pieces.
This could be a real embarassment for Microsoft. I suspect the real value of the code base is well under $.02. Somehow this whole thing smells like a hoax. From Seattle Time Sunday edition. Dateline SEATTLE (CNNfn).
Microsoft does not support it anyway. Read the EULA.
While I am on a rant, WINE is a good idea, if for no other reason that "because it's there", to see how far the idea can be pushed. It will even run some (most?) legacy Windows aps. Having the source to Windows would help some, maybe. Emulator running simulator running legacy ap. It's been done before.
Best bet is good clean aps for Linux (and thence *BSD and other un*xes).
or heavy excise tax.
In any event, disallow the Microsoft EULA which makes the hardware reseller, not Microsoft, responsible for the software.
>... Microsoft's own engineers too scared to fix the base of the code itself, so every patch is an extension to the current code base.
I have been expecting and looking for this, but this is the first reference I've come across. I've seen the phenomenon before, from a distance. It is the kiss of death, not quick and easy, but slow and agonizing.
-- "The evil that men do lives on after them. The good is oft interred with their bones"
And if he didn't, he should have.
>A good unix-compatible but newer-implemented OS will come along rather soon (I hope) and kill all others.
;-)
At the moment, that OS *is* Linux. AT&T has always diligently protected the trademark UNIX. There was a long and bitter battle between Berkeley and AT&T over the independently derived BSD Un*x. Linux is an independent implementation of the Un*x standard. Because of the trademark and copyrights, you have *BSD and Linux, not UNIX. When that better OS comes along, the good parts will find their way into Linux and into *BSD.
From the introduction of FreeBSD. "For copyright reasons, FreeBSD may not be called UNIX. You be the judge of how much difference this makes."
Linux *has* to insist that Linux is not UNIX.
GNU, GNU's Not Unix. Same thing.
Linux *is* UNIX. You just can't call it UNIX(tm).
>The world is advancing at an alarming rate,
especially in the computer industry.
Right. It's called Windows(tm). It is alarming.
To get an idea of how slow, do you know *any* mathematics from the 20th century? Fundamental advances are *hard* and do not happen that often.
UNIX will still be obsolete thirty years from now. You will still have the requisite press announcements of UNIX's demise.
AT&T ownned it and couldn't kill it. Microsoft can't kill it. IBM/TransMeta will not kill it.
by daywalker: Al-Gore-rithyms -- priceless
OK, I'll try a stab at this.
First, bugs are not created equal. Formal testing tries to use metrics, but this distorts reality.
New user syndrome. A new user will stumble into bugs a more experienced user would never encounter.
Open Source has at least two effects. First, the code is probably better thought out initially, otherwise it is flame-bait. Commercial quality code, at least IBM from long ago, is NOT that good, and from a few check points has actually deteriorated with time. Second, there should be a few people out who take a pathological delight in finding and eradicating certain types of bugs. With the source, they know where to look and what to look for.
For the bugs that matter, somebody competent out there will run into it and fix it.
In general, the user community will be as helpful as possible.
Bugs are diverse. The user community is diverse and beats on it in a variety of ways.
Open software will not catch all the bugs, but will tend to catch all of them that matter.
Closed software testing will start out with good intentions, but the same old same old tends more toward preservation of original bugs than their elimination. Further, there is an assumption that it's tested so it must be right. In addition the programmer wants his boss to think he did a good job. The whole mess creates a sort of arrogance so that if a user does find a bug, it's far easier to just try to bypass it. It's just not worth the hassle. Even if it does make it onto the list of things to be fixed, it will be months at least before the user will see the results.
In contrast, with Open Source, it is worth the hassle. The user gets his bug fixed, along with any fixes for other bugs. At this point, with a sensitized target, with communication lines open, you get the best regression testing imaginable. Everybody gains.
In short, with Closed Source, the forces are aimed at burying or trying to ignore bugs. With Open Source, the forces are aimed at 1) ameliorating effects of bugs and 2) exposure of bugs. Further, Closed Source is effectively paid by the line. There is a strong disincentive to take the effort to shorten and tighten the code.
Release early. Release often. Catch the really bad actor early. Cure the bug twice. First the bug itself. Second, fix it so the bug can't really hurt you. Another rule. Bugs are like snakes, they go around in pairs, so find the other one. It helps to be paranoid.
Be aware of the level. Cool on the desktop is ok. Cool for non-critical stuff is fine. Cool for critical stuff, particularly the kernel, is not ok. As I understand it, Linus Torvalds is a very nice guy. In the book excerpt from Open Sources, he uses the word "stupid" 4 times. The more I think about it, the more I am convinced that he is exactly right. At this level, it is much more important to not be stupid than it is to be brilliant. If you look at it the right way and it is stupid, it is stupid. Remember it. "The very basic rule is to avoid interfaces." This is dead on target. You cannot avoid them entirely, but less is better. Hmmm, maybe this is why Un*x cannot be killed off. "While the NT team knew the final result wouldn't approach a microkernel, they knew they had to pay lip service to the idea." And this starts the chain of deception. At the level of the kernel, and other critical stuff, you need to be [brutally] honest.
Crystal ball. (He who lives by the crystal ball shall learn to enjoy ground glass) Beware Windows2000 (W2K).
If you want to be able to recover contents of pre Service Pack 3 NT4, it is pretty much essential. Furthermore, you make an NT boot floppy. Format under NT. Copy to floppy files BOOT.INI NTDETECT.COM NTLDR NTBOOTDD.SYS
Curiously, booting NT from floppy can work when NT cannot boot directly from Hard Drive.
>THAT is so wrong on SO many levels,...
I agree, BUT am unwilling to risk losing access to the data.
PS. I also hit the power switch when it gets its brains fried. Safer to not let it try to preserve the damage on hard drive.
(signed) Windows NT (ab)user.
cartoon. Feb 28,99. microwaving CDs. AOL makes an art form. NT crashed the microwave. I'm still laughing. (Microsoft does not appreciate unauthorized release of sensitive development information)
As I understand it, if the employee has the right to take the binary home with her/him, that employee has the right to take the source home with him/her. I think in general, the employee would not have either right. The right to the binary and the source belongs to the company. The company has no obligation to release anything to anybody. You can sell a modified GPL program, but the recipient has the rights to source and is free to modify, sell or give away the program, subject to the GPL. You cannot even give away a modified binary-only without making the source available.
Realistically, you want to send back to the community any bug fixes or generally useful extensions. So much easier the next time around. Anything proprietary^H^H^H^H^H^H^H^H^H^H^H private goes nowhere further. The employees have no more rights than the general public.
I think the key to understanding it is that the company itself, not the employees, owns the binaries.
The above opinions are my own. GPL has no responsibility for them.
Explains a lot. ;-)
Now, with all levels trying to put the best face on things, try to imagine how bad the W2K bug (or is it virus? nah, too big for a virus) really is.
Ignore him and you will get more of him. He probably believes his own hype. What do you think his divisional managers tell him? What do their departmental managers tell them? Etc? (However the chain of command goes.)
...flaw. That's because it's almost always the flaws that are hidden.
He is more like a carnival barker who has gotten the public (and PHBs who believe/hope the Gates magic will rub off) to believe. Wow! Look what I can do! (Just don't look too close. Don't stress it). Silence is affirmation. Satire and ridicule, a constant barrage, is the only solution. Seriously, would YOU want to be Bill Gates? Not me.
Contrast Bill with Linus. In a recent post, Linus is using the term "stupid" rather often. This is the same "stupid" as in KISS (Keep It Simple, STUPID). Translated it means "Do not outsmart yourself". The targets of KISS are not dumb or what would normally be thought of as stupid. The object is to try to keep unknown unknowns from killing people. Murphy's Law rules. Mother Nature sides with the hidden. Yeah, I know.
It is not as simple as it works or it doesn't. There is a vast spread between (among?)
1) it never works, by any definition of works
2) it sorta works some of the time
3) it usually works, for most things
4) works, except for exceptional cases
5) never fails, except isolated freaks
6) never fails any circumstances
Bill's definition of works is about #3, aiming for #2.
Linus' definition of works is #5 aiming for #6.
You can count on the following: Windows is worse than you think it is. Linux is better than you think it is.
Microsoft has two things going for it. FUD and the BIG LIE. Combat FUD with information. Combat the BIG LIE with, preferably, the TRUTH, but in any event with something. Say it often enough, loud enough, and people will believe. Personally, I like to be able to, any day, any hour, go to slashdot and (live in living color) see what the people who understand computers are saying. BTW, I am a Windows (ab)user.
If Bill Gates tells lies about us, we will tell the truth about him.
almost as good as google search of "evil empire"
>what would happen if those technical decisions were made by software engineers?
Windows might be useable.
PS. Don't tell M$