Whenever someone says, "I understand it, I just can't articulate it," what they really mean is, "I don't understand it."
"There are various patterns that crop up that I have a knowlege of, but I have no words that describe the system, because I dont think in that manner."
"And knowing how to move the stuff around without messing up a peice is something that I have in physical memory."
What you are articulating, and using, is a concrete application of abstract algebra, at a level deep and fundamental enough that the algebraists don't really have the vocabulary. There is a thin thread between the concepts and the language. It stretches, well out of sight, but it does not go forever.
You can take a cube apart and flip one piece on an edge so the cube cannot be put back together. I'd guess you would know it couldn't be put back fairly quickly without moving anything. That would be using language even if it is just to yourself.
Probably the best evidence of the influence of language is the nearly simultaneous discoveries of major inventions, like calculus. It seems also that the practical is often far in advance of the theoretical.
Assume I have two computers, one monitor, one keyboard, one mouse. I will show up as 100% Windows anywhere over a large range of doing most everything on Windows to doing most everything on Linux. Further, competent browsers can lie as to their identity. Collecting statistics is even less informative than anecdotal evidence.
He who lives by the crystal ball shall learn to enjoy the taste of ground glass, but I would predict that the balance of desktop usage will shift to Linux rather abruptly, like a teeter-totter shifting balance. Further I would predict that the shift will be Linux on new hardware with something that actually takes useable advantage of the new hardware, and is impractical with Windows. Other than that, the PC is really becoming more and more just a dumb terminal. The idea of thin clients is right, but it requires "just a browser", where the technology is sufficiently andvanced that nobody knows or cares which browser they are using.
What they ask for is a site that stands out from their competitors, usually they have a very large idea ... and that is who the site is developed for. Not for the customers.
The site much stand out from the competition on ONE computer. One browser. On broadband if not LAN speeds. Only after several iterations will the question of exactly who is supposed to be using this thing even come up.
I'm not clear on how the difference can be detected
If it were clear then mechanincal intelligence would suffice.
I send you this file to have your advice. Could be legitimate, but only if the file only makes sense coming specifically from me to specifically you and if I would be asking specifically for advice. Legitimacy of messages depends critically on context. Why is this messager here? Now?
What would be harder is to defend against: Subject: Hey stupid. Body: You forgot something. Attachment: payload.
Not very good security, but would be rather effective. Note: The first requirement for Security by Obscurity is OBSCURITY. Anything ubiquitous is NOT OBSCURE.
The advantage of standards should be that, other than a few select cases you choose to give special treatment, you shouldn't have to care which User Agent it is or claims to be. This is completely the wrong place to put self-advertisements.
That's the real reason Open Source tends to be more secure than closed source.
I would expect anything serious with Red Hat to show up on Red Hat's main page before Slashdot's main page. I can fully justify reading Slashdot at work just to keep on top of the latest in Microsoft wormage.
I suspect that the Open Source coders are better, but that is an "extra". Because of the response patterns, Open Source coders would have to be much worse to break even with Closed Source coders.
Put a 20' 2x12 on the ground and walk from one end to the other. Easy. Put same board 100' in the air and walk from one end to the other. I wouldn't. The difference is the cost of an error. And once you fall into the trap of trying to convince anyone else that there are no problems, I have no idea how you get out of it.
Until anything that can be tenuously blamed on microsoft happens.
You mean like "Windows is starting up"?
This lack of objectivity makes everyone here look as mature as an 8-year-old kid. Well, you might be an 8-year-old kid, but that was a long, long time ago for me. And I'm not the only old fart around this place. Some of us even come here specifically for the Microsoft bashing. When the objective is to find new and interesting ways in which to bash Microsoft, I'd say that Slashdot is certainly not lacking in objectivity.
Nerds get a bad enough rap as it is. To whatever extent that is true, I'd say that nerds tend to bring it on themselves. Probably phrased better by some ancient Greek philosopher, but if you can stand to live with yourself, nothing else matters much.
the background image shows what appears to be a way out of a window, whereas the whole point of the site is to show you a way into Windows (TM).
I'd guess that they tried a deal with Microsoft and eventually figured it wasn't getting them where they wanted to go.
Unisys joining the bandwagon is a measure of the strength of that bandwagon. Real life experience with Linux on a mainframe, Case Study: zLinux Cuts Costs at Controller's Office, gives a taste of the future. What works is slowly, carefully testing the waters, making progress, and consolidating the position. Not fast. Not easy. But quite doable, even with no prior experience.
"Like many Big-Iron Linux adopters, Idaho's initial deployment plans were relatively limited in scope and ambition. For example, says George Judge, a deputy controller in the Idaho State Controller's Office, his organization was looking for a way to streamline its highly distributed Intel-based infrastructure. "I was very concerned about the proliferation and the cost of the little... [Intel] boxes and the utilization of same," he confirms."
"Now finished and in production, Idaho's new zLinux-based Web publishing system taps a.PDF reporting format, PHP coding, and a MySQL repository. Best of all, says Marchant, it's both user-friendly and user-empowering: "We've been able to build user interfaces where report-owners actually perform all of the maintenance activities that will trigger their reports to be pushed [from] the mainframe to the application running on zLinux," she observes"
"Idaho's MySQL repository current stores almost 2.9 million pages, all of which are accessible from the Web, according to Judge. The new application currently supports as many as 600 concurrent users, a number that's expected to continue to grow as the Idaho State Controller's Office rolls the application out to a broader audience, such as vendors who provide goods and services to the state. "Our authentication database has 26,000 active users [state employees and others] over the course of a year, and any of those users are eligible to view online reports," notes Marchant."
From a cast-off P2 to a new z800 mainframe there are similarities and there are differences. The similarities are easy to spot. The differences are not easy to spot and cannot be seen or comprehended from the low-end. Those experienced on the high-end maybe have a clue. That's why big customers will pay good money to IBM and Red Hat and Novell and now apparently Unisys.
Then flip the first bit, hash the file, flip the first bit back, flip the second bit, hash again, and so on. Do this and eventually you'll run into a collision. Also, you'll only need to do N+1 calculations. N isn't that big either.
The sequence 100000 010000 001000 000100 000010 000001 does NOT exhaust the space.
Uhhh, orders of magnitude is whatever base you're working in, not necessarily stated or even known. With a range of 30-130 orders of magnitude, the 0103 of 0.30103 looks kinda silly. Base 2 to base 10 is a small difference. This is talking about big differences.
This of course assumes that OSS = high quality. That is definitely NOT always the case.
Even more distorted. There exists some OSS software which is not of high quality.
Wide availability of high quality, low cost software. There exists high quality, low cost software. It is widely available. Doesn't seem to carry any implication that all or even much of OSS = high quality. The statement holds if there is a little bit of high quality OSS software and that little bit is widely available.
Now there is more than a little truth to OSS implies high quality. It doesn't happen immediately, but time is on OSS's side. Closed Source, once made, tends to pretty much just sit there. There is little mechanism and little incentive to improve it. Whatever help facilities exist always have the assumption that the Closed Source is both correct and immutable and that it is the user of that Closed Source who must distort him or herself to fit. Open Source, once made, tends to be a bit more active. There are mechanisms and incentives to improve it. You can always fork it if nothing else. The unhelpful "help facilities" pretty much do not exist and the users who refuse to distort themselves can and will improve the situation. Of course, bad ideas tend to die out, probably faster with Open Source than with Closed.
Germaine to the topic of helping the world's economy, helping has to involve interactions between interested parties as opposed to isolating the various factions. Helping occurs when both parties gain from a transaction. Overall, the various parties can gain more when they are able to understand one another. If each individual has his own private code and can speak only to himself and be understook only by himself, the world opportunities are much less than if information can be conveyed and shared. If you can read what I write, maybe we can and should do business.
What makes you think F/OSS is any less secure? Because you can see how the lock is made? If it's made right, it shouldn't matter. If it's implemented right, it shouldn't matter.
You can secure stuff going from a secure location through unsecure connection to another secure location. You have immense difficulty with secure stuff going from a secure location through unsecure connection to an unsecure location.
If you can use X-Rays to see exactly how the lock is made, regardless of how it is implemented, you can easily open it.
Humanity is screeching to a halt. New technology has little to no effect on people's lives any more.
Hardly. The effects of technology take a while to become pervasive, now just as in the past. Some changes are radical and dramatic, most are evolutional.
Steam engine versus internal combustion engine. Try mowing your small lawn with a steam engine powered lawnmower.
Telegraph versus pony express versus telephone. All are vast improvements over hand-carrying a message in person. The ability to send a few words quickly is the most important. A 300 baud modem is much better than zero bits per second.
Computers have changed hardly at all in the last thirty years, even. Spreadsheets. Harness an awful lot of computer power to do a few simple calculations. What has changed is the value/price so that the world can now afford millions of them as opposed to the originally projected 5.
People don't get a host, and then decide they want something to be hosted, that makes no sense.
True, ONLY if there is ONE thing hosted.
The decision of where to host will be made, if technical factors are even considered at all, based on the initial guess at requirements. As requirements are ADDED to the existing set, periodically the question of whether to stay or to move will be examined, with inertia and fear of the unknown tending toward staying. The problem with moving is that everything you do not know is capable of wrecking havoc.
The problem with "the right tool for the job" is that there are many jobs and only a very limited amount of space in the toolbox.
Microsoft is probably right about the TCO of Linux being higher than Windows.
"Well my windows experience hasn't been that good. Just about every windows network I've seen has been slapped together backwards. Shit is totally installed wrong. And quite often people just barely got it working. As an aside I've also noticed that there are a lot of Windows programming shops that slap together some really crappy code with no error checking."
"But of course Linux could just as easily be in the same position, but for some reason I have yet to see it. Maybe it's because I deal with smaller businesses that can't afford an uber admin, or may not no any better on who to hire?"
With Microsoft Windows it just works. (barely) With Linux, if it just barely works, it tends to stick out and somebody is expected to do something about it and fix something. This costs more time and money.
With both Windows and Linux there are a lot of subliminal clues as to what is accecptable behavior and how much effort can be expected to produce acceptable results. There are also clues as to whether one should accept responsibility for ones actions or attempt to shift the blame elsewhere. "And quite often people just barely go it working" and felt quite pleased with themselves too, I'd bet. Unix philosophy seems to be that if you succeed in doing what you were supposed to do, shut up about it. However, if you failed you are required to die noisily. Microsoft seems to take the opposite approach which is deadly if you attempt to do much of anything that requires several things to work together.
I hope the major news outlets cover the huge difference in paradigm here- good cop instead of bad cop.
There's a huge difference in paradigm, but if the media does anything about it, it will be to bury it.
With the possible exception of some stuff by Knuth, everything has bugs, where possible inputs produce undesirable outputs.
Given that there are bugs, what's the better way to stumble into them? Something nasty and hidden? Something spectacular and harmless?
No, the media will be worse than useless. Since their livelihood depends on advertising, they are incapable of comprehending the mind of someone who keeps looking for hidden flaws.
There's a distinction between DATA and PROGRAMMING. Supplying the source for a login program does not imply supplying all your usernames and passwords. Confidential information about the company shouldn't be hardcoded into the programming anyway, otherwise every salesman who files an expense report exposes your confidential information much more so than the programming itself.
Unfortunatly it's management who will think to themselves "oh no, what if our compeditors start using it too, then we just helped them".
A bit shortsighted, methinks. What happens when my competitors are forced into doing things my way because they are using software developed with my priorities in mind? There are lots of "minor" decisions that are made when software is initially developed, subliminal maybe, but various assumptions are made as to the existing context and the desired direction. The sum total gives a rather large force that tends to dictate the way business will be done. Just watch the clash when two corporate cultures join because of a merger. Even if all the ground rules and objectives are the same, the partitioning will be different. (Partitioning is how mathematicians get bigger infinities;)
Grandparent's observation "Release it as open source - the payback is that you get to use the program, well maintained and all, even after the developer has moved" covers some of the ground. There are also all the implicit built-in assumptions that went into it that will continue to serve the originator's objectives. There is also the fact that it helps immensely to be part of a thriving industry, which has to mean that your competitors are doing well.
Bumper sticker on an old junker on a mountain road. I may be slow but I'm ahead of you.
IBM is just following standard change management/control procedures. Of course, they want to ensure that their own internal application still work after the apply of SP2. Whats wrong with that?
Nothing wrong with that. That's the way it has to be. It does however impact the window during which Microsoft Windows systems will remain unpatched and vulnerable. It doesn't matter how fast Microsoft reacts. The relevant timing is when the patches can be safely applied.
This is where Open Source has essentially an unfair advantage. It is almost trivial to know exactly what the patches do, what they might affect, and to be extremely selective in which parts of which patches are applied, even to the point of changing patches to work around problems where you cannot or don't want to patch.
Am I someone who knows whether something is actionable or not? Well, I've worked in publishing for some considerable time, and I know basic libel law.
As one who has worked in publishing for some considerable time, are you saying that Slashdot is publishing? My own impression is that Slashdot throws up some Headlines, links, and some provocative commentary to start the discussions, hardly what I'd consider to be publishing. In less that three hours of inital posting, which clarifies the situation as to someone within ZD-whatever overreacting to seeing their content elsewhere. Regardless of your view of what Slashdot should be doing, they seem to be doing very well just as they are, with possible quibbles about color schemes.
There's also the situation where you have a product which is both proprietary and GPL. The equilibrium state is where the GPL stuff is playing with the cutting edge and the proprietary stuff is holding back awaiting developments. The situation looks a bit screwey, but consider that the people paying for the proprietary version not only want stability but want someone around who can and will fix any problems that arise. If the advanced GPL version kills most of the bugs before the code is also the proprietary version, so much the better for the paying customers. Hackers playing with the advanced GPL version get to play with and use the state of the art, are actually in a good position to get support because the stuff is fresh and new.
Now in the above, would you have any intention of accepting something which is strictly GPL and then supporting simultaneously both forks caused by that minor improvement? You could take existing MySQL code under the GPL and improve it. However you'd have to keep maintaining it and try to keep up with MySQL AB. If I did manage to come up with a significant improvement which I was dependent on (Why else would I have gone to the trouble?) after a bit of consideration, I think I would pay MySQL AB to take it. Something's topsy-turvy here.
Anyone who knows anything whatsoever about graphic design knows...
that you use high-contrast for what you want to be easily readable and low contrast (like beige on light beige) for stuff that needs to be there but shouldn't draw that much attention to itself.
Looks to me like the scheme is doing what it should be doing. More emphasis on the comments and less on who posted them (and the links that are always there in the same places).
Slashdot Mirror
Whenever someone says, "I understand it, I just can't articulate it," what they really mean is, "I don't understand it."
"There are various patterns that crop up that I have a knowlege of, but I have no words that describe the system, because I dont think in that manner."
"And knowing how to move the stuff around without messing up a peice is something that I have in physical memory."
What you are articulating, and using, is a concrete application of abstract algebra, at a level deep and fundamental enough that the algebraists don't really have the vocabulary. There is a thin thread between the concepts and the language. It stretches, well out of sight, but it does not go forever.
You can take a cube apart and flip one piece on an edge so the cube cannot be put back together. I'd guess you would know it couldn't be put back fairly quickly without moving anything. That would be using language even if it is just to yourself.
Probably the best evidence of the influence of language is the nearly simultaneous discoveries of major inventions, like calculus. It seems also that the practical is often far in advance of the theoretical.
Assume I have two computers, one monitor, one keyboard, one mouse. I will show up as 100% Windows anywhere over a large range of doing most everything on Windows to doing most everything on Linux. Further, competent browsers can lie as to their identity. Collecting statistics is even less informative than anecdotal evidence.
He who lives by the crystal ball shall learn to enjoy the taste of ground glass, but I would predict that the balance of desktop usage will shift to Linux rather abruptly, like a teeter-totter shifting balance. Further I would predict that the shift will be Linux on new hardware with something that actually takes useable advantage of the new hardware, and is impractical with Windows. Other than that, the PC is really becoming more and more just a dumb terminal. The idea of thin clients is right, but it requires "just a browser", where the technology is sufficiently andvanced that nobody knows or cares which browser they are using.
What they ask for is a site that stands out from their competitors, usually they have a very large idea ... and that is who the site is developed for. Not for the customers.
The site much stand out from the competition on ONE computer. One browser. On broadband if not LAN speeds. Only after several iterations will the question of exactly who is supposed to be using this thing even come up.
I'm not clear on how the difference can be detected
If it were clear then mechanincal intelligence would suffice.
I send you this file to have your advice.
Could be legitimate, but only if the file only makes sense coming specifically from me to specifically you and if I would be asking specifically for advice. Legitimacy of messages depends critically on context. Why is this messager here? Now?
What would be harder is to defend against:
Subject: Hey stupid.
Body: You forgot something.
Attachment: payload.
Not very good security, but would be rather effective.
Note: The first requirement for Security by Obscurity is OBSCURITY.
Anything ubiquitous is NOT OBSCURE.
The advantage of standards should be that, other than a few select cases you choose to give special treatment, you shouldn't have to care which User Agent it is or claims to be. This is completely the wrong place to put self-advertisements.
IBM has put their valuable Intellectual Property into the public domain?
That doesn't sound like any IBM I've ever heard of.
That's the real reason Open Source tends to be more secure than closed source.
I would expect anything serious with Red Hat to show up on Red Hat's main page before Slashdot's main page.
I can fully justify reading Slashdot at work just to keep on top of the latest in Microsoft wormage.
I suspect that the Open Source coders are better, but that is an "extra". Because of the response patterns, Open Source coders would have to be much worse to break even with Closed Source coders.
Put a 20' 2x12 on the ground and walk from one end to the other. Easy.
Put same board 100' in the air and walk from one end to the other. I wouldn't.
The difference is the cost of an error. And once you fall into the trap of trying to convince anyone else that there are no problems, I have no idea how you get out of it.
Windows Update is the online extension of Windows that helps you get the most out of your computer.
You must be running a Microsoft Windows operating system in order to use Windows Update.
Small point maybe, but when you see that, you know that Microsoft is still not serious about security.
Until anything that can be tenuously blamed on microsoft happens.
You mean like "Windows is starting up"?
This lack of objectivity makes everyone here look as mature as an 8-year-old kid.
Well, you might be an 8-year-old kid, but that was a long, long time ago for me. And I'm not the only old fart around this place. Some of us even come here specifically for the Microsoft bashing. When the objective is to find new and interesting ways in which to bash Microsoft, I'd say that Slashdot is certainly not lacking in objectivity.
Nerds get a bad enough rap as it is.
To whatever extent that is true, I'd say that nerds tend to bring it on themselves. Probably phrased better by some ancient Greek philosopher, but if you can stand to live with yourself, nothing else matters much.
the background image shows what appears to be a way out of a window, whereas the whole point of the site is to show you a way into Windows (TM).
... [Intel] boxes and the utilization of same," he confirms."
.PDF reporting format, PHP coding, and a MySQL repository. Best of all, says Marchant, it's both user-friendly and user-empowering: "We've been able to build user interfaces where report-owners actually perform all of the maintenance activities that will trigger their reports to be pushed [from] the mainframe to the application running on zLinux," she observes"
I'd guess that they tried a deal with Microsoft and eventually figured it wasn't getting them where they wanted to go.
Unisys joining the bandwagon is a measure of the strength of that bandwagon. Real life experience with Linux on a mainframe, Case Study: zLinux Cuts Costs at Controller's Office, gives a taste of the future. What works is slowly, carefully testing the waters, making progress, and consolidating the position. Not fast. Not easy. But quite doable, even with no prior experience.
"Like many Big-Iron Linux adopters, Idaho's initial deployment plans were relatively limited in scope and ambition. For example, says George Judge, a deputy controller in the Idaho State Controller's Office, his organization was looking for a way to streamline its highly distributed Intel-based infrastructure. "I was very concerned about the proliferation and the cost of the little
"Now finished and in production, Idaho's new zLinux-based Web publishing system taps a
"Idaho's MySQL repository current stores almost 2.9 million pages, all of which are accessible from the Web, according to Judge. The new application currently supports as many as 600 concurrent users, a number that's expected to continue to grow as the Idaho State Controller's Office rolls the application out to a broader audience, such as vendors who provide goods and services to the state. "Our authentication database has 26,000 active users [state employees and others] over the course of a year, and any of those users are eligible to view online reports," notes Marchant."
From a cast-off P2 to a new z800 mainframe there are similarities and there are differences. The similarities are easy to spot. The differences are not easy to spot and cannot be seen or comprehended from the low-end. Those experienced on the high-end maybe have a clue. That's why big customers will pay good money to IBM and Red Hat and Novell and now apparently Unisys.
Then flip the first bit, hash the file, flip the first bit back, flip the second bit, hash again, and so on. Do this and eventually you'll run into a collision. Also, you'll only need to do N+1 calculations. N isn't that big either.
The sequence 100000 010000 001000 000100 000010 000001
does NOT exhaust the space.
You need 2^N+1 calculations
Uhhh, orders of magnitude is whatever base you're working in, not necessarily stated or even known.
With a range of 30-130 orders of magnitude, the 0103 of 0.30103 looks kinda silly. Base 2 to base 10 is a small difference. This is talking about big differences.
This of course assumes that OSS = high quality. That is definitely NOT always the case.
Even more distorted.
There exists some OSS software which is not of high quality.
Wide availability of high quality, low cost software.
There exists high quality, low cost software. It is widely available.
Doesn't seem to carry any implication that all or even much of OSS = high quality. The statement holds if there is a little bit of high quality OSS software and that little bit is widely available.
Now there is more than a little truth to OSS implies high quality. It doesn't happen immediately, but time is on OSS's side.
Closed Source, once made, tends to pretty much just sit there. There is little mechanism and little incentive to improve it. Whatever help facilities exist always have the assumption that the Closed Source is both correct and immutable and that it is the user of that Closed Source who must distort him or herself to fit.
Open Source, once made, tends to be a bit more active. There are mechanisms and incentives to improve it. You can always fork it if nothing else. The unhelpful "help facilities" pretty much do not exist and the users who refuse to distort themselves can and will improve the situation.
Of course, bad ideas tend to die out, probably faster with Open Source than with Closed.
Germaine to the topic of helping the world's economy, helping has to involve interactions between interested parties as opposed to isolating the various factions. Helping occurs when both parties gain from a transaction. Overall, the various parties can gain more when they are able to understand one another. If each individual has his own private code and can speak only to himself and be understook only by himself, the world opportunities are much less than if information can be conveyed and shared. If you can read what I write, maybe we can and should do business.
What makes you think F/OSS is any less secure? Because you can see how the lock is made? If it's made right, it shouldn't matter. If it's implemented right, it shouldn't matter.
You can secure stuff going from a secure location through unsecure connection to another secure location.
You have immense difficulty with secure stuff going from a secure location through unsecure connection to an unsecure location.
If you can use X-Rays to see exactly how the lock is made, regardless of how it is implemented, you can easily open it.
F/OSS is good, but nothing is that good.
Humanity is screeching to a halt. New technology has little to no effect on people's lives any more.
Hardly. The effects of technology take a while to become pervasive, now just as in the past. Some changes are radical and dramatic, most are evolutional.
Steam engine versus internal combustion engine. Try mowing your small lawn with a steam engine powered lawnmower.
Telegraph versus pony express versus telephone. All are vast improvements over hand-carrying a message in person. The ability to send a few words quickly is the most important. A 300 baud modem is much better than zero bits per second.
Computers have changed hardly at all in the last thirty years, even.
Spreadsheets. Harness an awful lot of computer power to do a few simple calculations. What has changed is the value/price so that the world can now afford millions of them as opposed to the originally projected 5.
People don't get a host, and then decide they want something to be hosted, that makes no sense.
True, ONLY if there is ONE thing hosted.
The decision of where to host will be made, if technical factors are even considered at all, based on the initial guess at requirements.
As requirements are ADDED to the existing set, periodically the question of whether to stay or to move will be examined, with inertia and fear of the unknown tending toward staying. The problem with moving is that everything you do not know is capable of wrecking havoc.
The problem with "the right tool for the job" is that there are many jobs and only a very limited amount of space in the toolbox.
Heroes or villians?
Heroes is better.
It's not for the dead, it's for the living.
If they did not die in vain, they are heroes.
Microsoft is probably right about the TCO of Linux being higher than Windows.
"Well my windows experience hasn't been that good. Just about every windows network I've seen has been slapped together backwards. Shit is totally installed wrong. And quite often people just barely got it working. As an aside I've also noticed that there are a lot of Windows programming shops that slap together some really crappy code with no error checking."
"But of course Linux could just as easily be in the same position, but for some reason I have yet to see it. Maybe it's because I deal with smaller businesses that can't afford an uber admin, or may not no any better on who to hire?"
With Microsoft Windows it just works. (barely)
With Linux, if it just barely works, it tends to stick out and somebody is expected to do something about it and fix something. This costs more time and money.
With both Windows and Linux there are a lot of subliminal clues as to what is accecptable behavior and how much effort can be expected to produce acceptable results. There are also clues as to whether one should accept responsibility for ones actions or attempt to shift the blame elsewhere. "And quite often people just barely go it working" and felt quite pleased with themselves too, I'd bet.
Unix philosophy seems to be that if you succeed in doing what you were supposed to do, shut up about it. However, if you failed you are required to die noisily. Microsoft seems to take the opposite approach which is deadly if you attempt to do much of anything that requires several things to work together.
I hope the major news outlets cover the huge difference in paradigm here- good cop instead of bad cop.
There's a huge difference in paradigm, but if the media does anything about it, it will be to bury it.
With the possible exception of some stuff by Knuth, everything has bugs, where possible inputs produce undesirable outputs.
Given that there are bugs, what's the better way to stumble into them?
Something nasty and hidden?
Something spectacular and harmless?
No, the media will be worse than useless. Since their livelihood depends on advertising, they are incapable of comprehending the mind of someone who keeps looking for hidden flaws.
There's a distinction between DATA and PROGRAMMING.
Supplying the source for a login program does not imply supplying all your usernames and passwords.
Confidential information about the company shouldn't be hardcoded into the programming anyway, otherwise every salesman who files an expense report exposes your confidential information much more so than the programming itself.
Unfortunatly it's management who will think to themselves "oh no, what if our compeditors start using it too, then we just helped them".
A bit shortsighted, methinks.
What happens when my competitors are forced into doing things my way because they are using software developed with my priorities in mind? There are lots of "minor" decisions that are made when software is initially developed, subliminal maybe, but various assumptions are made as to the existing context and the desired direction. The sum total gives a rather large force that tends to dictate the way business will be done. Just watch the clash when two corporate cultures join because of a merger. Even if all the ground rules and objectives are the same, the partitioning will be different. (Partitioning is how mathematicians get bigger infinities;)
Grandparent's observation "Release it as open source - the payback is that you get to use the program, well maintained and all, even after the developer has moved" covers some of the ground. There are also all the implicit built-in assumptions that went into it that will continue to serve the originator's objectives. There is also the fact that it helps immensely to be part of a thriving industry, which has to mean that your competitors are doing well.
Bumper sticker on an old junker on a mountain road.
I may be slow but I'm ahead of you.
IBM is just following standard change management/control procedures. Of course, they want to ensure that their own internal application still work after the apply of SP2. Whats wrong with that?
Nothing wrong with that. That's the way it has to be.
It does however impact the window during which Microsoft Windows systems will remain unpatched and vulnerable. It doesn't matter how fast Microsoft reacts. The relevant timing is when the patches can be safely applied.
This is where Open Source has essentially an unfair advantage. It is almost trivial to know exactly what the patches do, what they might affect, and to be extremely selective in which parts of which patches are applied, even to the point of changing patches to work around problems where you cannot or don't want to patch.
Am I someone who knows whether something is actionable or not? Well, I've worked in publishing for some considerable time, and I know basic libel law.
As one who has worked in publishing for some considerable time, are you saying that Slashdot is publishing?
My own impression is that Slashdot throws up some Headlines, links, and some provocative commentary to start the discussions, hardly what I'd consider to be publishing. In less that three hours of inital posting, which clarifies the situation as to someone within ZD-whatever overreacting to seeing their content elsewhere. Regardless of your view of what Slashdot should be doing, they seem to be doing very well just as they are, with possible quibbles about color schemes.
There's also the situation where you have a product which is both proprietary and GPL. The equilibrium state is where the GPL stuff is playing with the cutting edge and the proprietary stuff is holding back awaiting developments. The situation looks a bit screwey, but consider that the people paying for the proprietary version not only want stability but want someone around who can and will fix any problems that arise. If the advanced GPL version kills most of the bugs before the code is also the proprietary version, so much the better for the paying customers. Hackers playing with the advanced GPL version get to play with and use the state of the art, are actually in a good position to get support because the stuff is fresh and new.
Now in the above, would you have any intention of accepting something which is strictly GPL and then supporting simultaneously both forks caused by that minor improvement? You could take existing MySQL code under the GPL and improve it. However you'd have to keep maintaining it and try to keep up with MySQL AB. If I did manage to come up with a significant improvement which I was dependent on (Why else would I have gone to the trouble?) after a bit of consideration, I think I would pay MySQL AB to take it. Something's topsy-turvy here.
Anyone who knows anything whatsoever about graphic design knows ...
that you use high-contrast for what you want to be easily readable and low contrast (like beige on light beige) for stuff that needs to be there but shouldn't draw that much attention to itself.
Looks to me like the scheme is doing what it should be doing. More emphasis on the comments and less on who posted them (and the links that are always there in the same places).