Personally, I believe if a company rewards people who have strong social skills, a decent moral backing, and adequate technical skills they're probably going to have the best chance at having good management.
I would say that a company which thinks "management" is a promotion from "non-management" has no chance at having good management *or* good workers. Selecting your management via some kind of reward process is a fundamentally flawed concept - the skills required are entirely different. You should select your management as part of the hiring process. While people can conceivably change into or out of the management roles, this should be seen as a "sideways" move in the organisation, like any other change of deparments - neither promotion nor demotion.
Sadly, very few companies work this way. One notable company that does is IBM - and their management staff is appreciably better than average, as most workers there will attest. That's not to say that getting this right solves all problems, but it almost certainly does help significantly.
If the government has 25-year-old secrets that would cause a war if revealed, we're probably all better off with the war than with the secrets. Anything that could cause that after so long would be so bad that we could not possibly justify continuing to do nothing about it.
zfs has WAFL style snapshots and replication - these aren't available anywhere outside of netapp filers
Constant-time (few seconds) filesystem snapshots? Got those. Been using those on production Linux servers for nearly three years now. IBM implemented them for Linux, because AIX had them (back in the 1990s) and their AIX customers wanted to use them on Linux. Pretty nearly everybody but Solaris has had them for ages. Just because Solaris didn't have a given feature didn't mean that the feature didn't exist elsewhere. But unlike Solaris, which only supports it on one filesystem, Linux supports it on every local disk filesystem (more or less - I wouldn't try it with NTFS and other weird stuff like that).
Replication? Everybody including Solaris has had that for ages. Probably the most commonly used unix filesystem with replication is AFS. Most people don't use filesystem-level replication because it's fundamentally slow. Application-level replication is superior for most real-world purposes.
That's a popular myth, but Ubuntu has not existed for 5 years. A more correct statement would be "Ubuntu states an intention to provide 5 years of support". That's not the same thing as a real support contract from a company that you're sure will still be around in 5 years time. Anybody could state an intention to provide N years of support, but that doesn't mean that you'll actually get it. (Redhat - not Fedora - have been providing such long-term support contracts to certain corporate customers, although I'm not sure what their longest duration one is. They just don't provide them to everybody who happens to buy an RHEL license)
DTrace is not "miles" beyond what linux has today (systemtap), it's inches. IBM and RedHat started implementing that functionality for linux over a year ago, and they're nearly finished (it works today, there's just some features missing).
ZFS is behind what linux has today - bolting a filesystem onto an lvm does not magically generate features that a system with a separate lvm and filesystem did not already have, but it does mean that they haven't managed to get group quotas working yet. Pretty much all of the other features in ZFS were implemented in linux first - although most sysadmins do not bother to deploy them.
Linux development does not happen "in your moms[sic] basement". If you pulled your head out of your kool-aid for long enough you might have noticed that it's mostly happening at IBM and RedHat, who have got more manpower on the problem than Sun's entire employee base, now that IBM have moved their focus from AIX to linux.
which begs the question, would we be safe if exposed to bird flu?
Given that there isn't any evidence of it being ever transmitted by humans, yes, probably (unless you spend a lot of time with your hands in the guts of sick or dead birds).
Ridiculous bit of deliberate media hysteria, that one. A couple of people die from handling dead birds, the scientists say "well, it's not a threat right now, but just like every other damn virus in the world it could mutate into something that was a serious problem", a drug company realises "hey, we've got something that might be effective against things like that, although obviously we've never tried it because the virus we need DOESN'T EXIST YET", and the next thing you know all the media companies are reporting that this drug company needs to be given lots of money or the world will end.
There is no "bird flu" threat. There is only the possibility that it may one day mutate into a threat - which is a trait it shares with every single other virus in existence.
For most of their active life, as far as they were concerned, space flight is an everyday occurance.
It's more subtle than that. Space flight as they know it is an everyday waste of money. It accomplishes nothing of value - it's just NASA sinking money into their retirement funds. It's no surprise at all that people don't really care about it any more.
Yes. You may have to spend time carefully crafting ACLs that will work with the programs you need to run. Some programs just don't support the concept of a multi-user system at all.
Such programs are nominally broken for causing this. If you want to get this working, you have to incorporate it into your approval process before buying new software (ie, don't buy stuff that won't work with it).
There are numerous precedents to the contrary, although not in every jurisdiction. The question of whether non-humans have property rights is one that has been largely ignored by legislation, so the courts have been forced to make it up as they go along. While there are some dissenters, the courts are usually tolerant of the idea (although you may be required to appoint a guardian to manage their estate) - on the basis that if a person wants to provide for an animal's welfare by giving them something, they should be permitted to do so.
Inanely, a default Windows install permits everybody to write files anywhere outside \windows, \progra~1, and \docume~1, regardless of admin rights. You have to change the filesystem ACLs by hand to fix this stupidity. Most Windows "admins" don't know about it, so they don't fix it.
Swap is encrypted on its own, in a smarter way./tmp is held in memory and never written to the disk (if appropriately configured) - and anyway, nothing should be writing sensitive data there. What kind of fucked up unix has a/usr/tmp? That's just wrong. Stop using it./usr is a read-only filesystem.
Swap encryption is handled differently, because it's more secure this way: at boot, the OS generates a random encryption key. This key is held in memory and never written to disk or displayed to the user. It's used to create a fresh encrypted swap area. When you switch the computer off, it's gone. When you boot it next time, it creates a new key. Absolutely no way to recover any data from an old swap image.
In a typical apartment complex with 16 units per building, all fire risks are multiplied 16x, because a single tenant can burn down all 16 tenants' apartments.
I don't know where you live or what the building codes are like there, but my current day job is with a company that (among other things) builds apartment complexes. We have about two feet of solid concrete between each apartment (plus several layers of other stuff mixed in), on all interior sides. We are not allowed to use any less. The exterior walls are actually thinner than the interior ones. It completely blocks all cellphone and wireless network signals (which makes my job harder). Fire does not pass it, which is the whole point. The only way a single tenant could burn down all the apartments is by going into each one and starting a new fire there.
(b) the police *love* picking up people for breaking these limits and are usually staked out along the stretch (c) half of the signs have speed cameras on them that activate along with the camera anyway.
Furthermore, since your prediction of average speeds will be poorly informed and therefore wrong, the very best thing to do is to drive slower than that. Taking as a given that you're currently in a traffic jam, all the cars around you are almost certainly going too fast for the jam to clear (if there's no obstruction ahead of you, and all the cars in the jam are going slowly enough, it will clear up in the time it takes one car to drive from the back of the jam to the front at that speed - typically a minute or two).
But we can do better than this. Have you ever noticed those electronic signs on busy roads, where they can display a variable speed limit? If they're being controlled by an intelligent person/system (which is, sadly, not always the case) then they'll have computed a speed at which this will occur, based on actual measurements of where all the cars are, and the signs are showing a limit of that speed.
Ever notice how everybody ignores those signs and drives as fast as they can?
That's why traffic jams happen. Obey the damn signs, you arseholes.
Despite the fact that it's a modern invention, nobody can even come close to guaranteeing that the DVD medium (the disc) can withstand storage for long times.
Nobody even tries. Optical media is not for archiving. Wrong technology for the purpose. And even if we had good archive-grade optical technology, the DVD would be a stupid form factor to archive anything in - it's designed for portability and retail advantages, and is entirely wrong for long-term storage. It's just too thin to be tough enough. (As an aside, an archival form would probably look rather more like a laserdisk, only about 10mm thick - much tougher, and better data storage in the same physical space - but it would be too expensive to be worthwhile, given the alternatives)
Can any slashdotter convince me that if I had properly stored important video media on a disc in say 20 years ago, this disc would still be readable now? With proper storage, the video cartridge would still be readable now after that long.
Where did you get that idea? VHS tapes do not last 20 years without severe quality degradation, sometimes to the point of destruction. It doesn't matter a damn how you store them, magnetic tape media is fundamentally volatile and VHS tapes are particularly bad at it (due to layout and encoding issues). They're dodgy after 20 years and gone after 50. Magnetic media is not for archiving either. Even worse than optical technology for the purpose.
Archival media is chemical media. Principally microfiche, and archive-grade film stock (all those reports you hear of old films degrading and needing expensive restoration are because nobody bothered to copy it to archival-grade stock - the regular stuff only lasts a decade or so). You get expected lifetimes of over 100 years with that stuff. If you want to keep your data for extended periods of time, fork over the money to get it archived properly.
If you want all that without spending any money on it, then I'd like a pony.
Without patents, who is going to come up with the immense sums required to bring a drug from investigational status to clinical reality?
The government, exactly the same as they do today. Those things cost money, but the government pays for them (via grants and the welfare state). It is a myth that private industry is pouring money into these things. It isn't.
I'd have to disagree with the article when they say the voice interfaces, such as those used in Star Trek, would be inefficient. If the machine is able to understand natural language,...
And therein lies the issue. *If* your system is able to parse natural language correctly, all your interface problems are solved. The problem is that saying "understand natural language" actually means "think like a human", which means the problem is AI-complete. At that point the statement becomes less interesting - it says that if we only have to deal with human-like entities, then we won't have a problem interfacing with computers.
Furthermore, if the system can do that, it's probably going to be self-aware, so it will want to be paid. While there may be some advantages to employing an AI instead of a regular meatbag, you can already hire somebody to translate your spoken commands into terms a computer can understand, so nothing fundamental will have changed.
On the other hand, if you only deal with approximate natural language parsers, you'll probably find that while an expert user can get by very well with them, the ignorant 99% of the workforce will barely be able to use them at all, because they don't know how to structure their thoughts in a way that the parser can understand.
I would suggest that class scheduling and enrollment/registration software might be another area. Universities and schools pay millions for this software, and it's usually pretty primative stuff.
Purchasing of software for this purpose is often highly political. No attempt is made to choose the best software for the job - instead, the one which pays the largest kickbacks is chosen. Replacing it is not feasible at most sites.
Inventory management and cash register software might be another area.
Cash registers are primarily a hardware device, not a software one. They're proprietary hardware that won't run anything other than what the manufacturer wants to ship. If you're running a cash-register-based system, your inventory management will probably have to be tightly integrated with them, so the manufacturer will leverage that to force their product on you across the whole enterprise.
On the other hand, if you can just slap a regular computer down at each sales point, then compiere (erp/crm/pos) or oscommerce (web-driven sales) already provides a complete service, including full support and consultancy packages.
That's not "useful", it's another art course. Someone is willing to pay six-figure amounts for all kinds of artwork - but most people aren't going to be creating it.
Only in the US do people seriously think that you can "own technologies". The rest of the world has, so far, rejected the notion of software patents (despite much effort from the US to get them more widely accepted).
Have you ever been to this little place called Europe?
The US has tried, twice, to push software patents onto Europe. They have failed, twice (right now they're trying again). Software patents are not legal anywhere in Europe. Strangely, the European Patent Office rubber-stamps them anyway (perhaps because they expect them to be legal in the future), but they're worthless paper in court. All significant patent lawsuits in Europe to date have been targeting the sale of hardware devices.
Of course, RPM is a giant pile of shit that should never have been invented - who is the fuck-ass who thought up using cpio with a fucked up header on it so you have to use dd (or something) before you can even manually unpack the archive? He needs a serious ass-kicking. But that's a digression and something I can forgive them for:)
If you think that's the real problem with RPM, you've never looked at the source. I can forgive the screwey and kinda limited file format - but the unmaintainable pile of vomit and demon bile that they call a source tree is a mortal sin against humanity. It contains slightly-forked copies of (and this list is incomplete): libdb, file, popt and sqlite (and the changes to these are not really documented). Variable names are frequently either hungarian, or just random single characters. Comments are almost non-existent in most of the code, except for doxygen tags.
Most of the reasons why RPM is irritatingly limited stem from the fact that very few people can bear to do any work on the thing. It's not something you'd want to touch unless you were getting paid really well.
Astroturfing is when a corporation pretends to be a grassroots campaign. That doesn't appear to be the case here - this guy is just a regular marketdroid, no pretending to be anything. Still a slimeball, but a different breed of slimeball.
Linux has a lot of great things, and OSS and GPL has a lot of great things, but there are also flies in the ointment, and to ignore them is becoming part of a cult.
Show me an OS without flies in the ointment, and I'll show you which cult you belong to. That's not a meaningful objection to any of them.
You can find lots of stuff that is in Linux distributions that are on shaking ground from Xerox technology to even Fat32, and there are tons of people that do own these technologies and could potential ask they stop being used and shutting down Linux distributions all over the place.
Only in the US do people seriously think that you can "own technologies". The rest of the world has, so far, rejected the notion of software patents (despite much effort from the US to get them more widely accepted). There will not be any Linux distributions being shut down, there will only be development moving out of the US.
Software patents are troll material anyway. There is no such thing as a piece of software that does not violate numerous software patents without a license. That includes Windows; Microsoft get sued for patent violations two or three times a year, and they usually just settle (it's not usually newsworthy). The system is so broken that you can't win, unless you simply refuse to play.
The really sad part is that if he's found guilty, the system will still be just as broken, but people will think it's fair. Which is why the system remains broken.
Slashdot Mirror
I would say that a company which thinks "management" is a promotion from "non-management" has no chance at having good management *or* good workers. Selecting your management via some kind of reward process is a fundamentally flawed concept - the skills required are entirely different. You should select your management as part of the hiring process. While people can conceivably change into or out of the management roles, this should be seen as a "sideways" move in the organisation, like any other change of deparments - neither promotion nor demotion.
Sadly, very few companies work this way. One notable company that does is IBM - and their management staff is appreciably better than average, as most workers there will attest. That's not to say that getting this right solves all problems, but it almost certainly does help significantly.
If the government has 25-year-old secrets that would cause a war if revealed, we're probably all better off with the war than with the secrets. Anything that could cause that after so long would be so bad that we could not possibly justify continuing to do nothing about it.
Constant-time (few seconds) filesystem snapshots? Got those. Been using those on production Linux servers for nearly three years now. IBM implemented them for Linux, because AIX had them (back in the 1990s) and their AIX customers wanted to use them on Linux. Pretty nearly everybody but Solaris has had them for ages. Just because Solaris didn't have a given feature didn't mean that the feature didn't exist elsewhere. But unlike Solaris, which only supports it on one filesystem, Linux supports it on every local disk filesystem (more or less - I wouldn't try it with NTFS and other weird stuff like that).
Replication? Everybody including Solaris has had that for ages. Probably the most commonly used unix filesystem with replication is AFS. Most people don't use filesystem-level replication because it's fundamentally slow. Application-level replication is superior for most real-world purposes.
That's a popular myth, but Ubuntu has not existed for 5 years. A more correct statement would be "Ubuntu states an intention to provide 5 years of support". That's not the same thing as a real support contract from a company that you're sure will still be around in 5 years time. Anybody could state an intention to provide N years of support, but that doesn't mean that you'll actually get it. (Redhat - not Fedora - have been providing such long-term support contracts to certain corporate customers, although I'm not sure what their longest duration one is. They just don't provide them to everybody who happens to buy an RHEL license)
DTrace is not "miles" beyond what linux has today (systemtap), it's inches. IBM and RedHat started implementing that functionality for linux over a year ago, and they're nearly finished (it works today, there's just some features missing).
ZFS is behind what linux has today - bolting a filesystem onto an lvm does not magically generate features that a system with a separate lvm and filesystem did not already have, but it does mean that they haven't managed to get group quotas working yet. Pretty much all of the other features in ZFS were implemented in linux first - although most sysadmins do not bother to deploy them.
Linux development does not happen "in your moms[sic] basement". If you pulled your head out of your kool-aid for long enough you might have noticed that it's mostly happening at IBM and RedHat, who have got more manpower on the problem than Sun's entire employee base, now that IBM have moved their focus from AIX to linux.
Given that there isn't any evidence of it being ever transmitted by humans, yes, probably (unless you spend a lot of time with your hands in the guts of sick or dead birds).
Ridiculous bit of deliberate media hysteria, that one. A couple of people die from handling dead birds, the scientists say "well, it's not a threat right now, but just like every other damn virus in the world it could mutate into something that was a serious problem", a drug company realises "hey, we've got something that might be effective against things like that, although obviously we've never tried it because the virus we need DOESN'T EXIST YET", and the next thing you know all the media companies are reporting that this drug company needs to be given lots of money or the world will end.
There is no "bird flu" threat. There is only the possibility that it may one day mutate into a threat - which is a trait it shares with every single other virus in existence.
It's more subtle than that. Space flight as they know it is an everyday waste of money. It accomplishes nothing of value - it's just NASA sinking money into their retirement funds. It's no surprise at all that people don't really care about it any more.
Yes. You may have to spend time carefully crafting ACLs that will work with the programs you need to run. Some programs just don't support the concept of a multi-user system at all.
Such programs are nominally broken for causing this. If you want to get this working, you have to incorporate it into your approval process before buying new software (ie, don't buy stuff that won't work with it).
There are numerous precedents to the contrary, although not in every jurisdiction. The question of whether non-humans have property rights is one that has been largely ignored by legislation, so the courts have been forced to make it up as they go along. While there are some dissenters, the courts are usually tolerant of the idea (although you may be required to appoint a guardian to manage their estate) - on the basis that if a person wants to provide for an animal's welfare by giving them something, they should be permitted to do so.
Inanely, a default Windows install permits everybody to write files anywhere outside \windows, \progra~1, and \docume~1, regardless of admin rights. You have to change the filesystem ACLs by hand to fix this stupidity. Most Windows "admins" don't know about it, so they don't fix it.
Swap is encrypted on its own, in a smarter way.
Swap encryption is handled differently, because it's more secure this way: at boot, the OS generates a random encryption key. This key is held in memory and never written to disk or displayed to the user. It's used to create a fresh encrypted swap area. When you switch the computer off, it's gone. When you boot it next time, it creates a new key. Absolutely no way to recover any data from an old swap image.
I don't know where you live or what the building codes are like there, but my current day job is with a company that (among other things) builds apartment complexes. We have about two feet of solid concrete between each apartment (plus several layers of other stuff mixed in), on all interior sides. We are not allowed to use any less. The exterior walls are actually thinner than the interior ones. It completely blocks all cellphone and wireless network signals (which makes my job harder). Fire does not pass it, which is the whole point. The only way a single tenant could burn down all the apartments is by going into each one and starting a new fire there.
Not where I live, they don't.
Furthermore, since your prediction of average speeds will be poorly informed and therefore wrong, the very best thing to do is to drive slower than that. Taking as a given that you're currently in a traffic jam, all the cars around you are almost certainly going too fast for the jam to clear (if there's no obstruction ahead of you, and all the cars in the jam are going slowly enough, it will clear up in the time it takes one car to drive from the back of the jam to the front at that speed - typically a minute or two).
But we can do better than this. Have you ever noticed those electronic signs on busy roads, where they can display a variable speed limit? If they're being controlled by an intelligent person/system (which is, sadly, not always the case) then they'll have computed a speed at which this will occur, based on actual measurements of where all the cars are, and the signs are showing a limit of that speed.
Ever notice how everybody ignores those signs and drives as fast as they can?
That's why traffic jams happen. Obey the damn signs, you arseholes.
Nobody even tries. Optical media is not for archiving. Wrong technology for the purpose. And even if we had good archive-grade optical technology, the DVD would be a stupid form factor to archive anything in - it's designed for portability and retail advantages, and is entirely wrong for long-term storage. It's just too thin to be tough enough. (As an aside, an archival form would probably look rather more like a laserdisk, only about 10mm thick - much tougher, and better data storage in the same physical space - but it would be too expensive to be worthwhile, given the alternatives)
Where did you get that idea? VHS tapes do not last 20 years without severe quality degradation, sometimes to the point of destruction. It doesn't matter a damn how you store them, magnetic tape media is fundamentally volatile and VHS tapes are particularly bad at it (due to layout and encoding issues). They're dodgy after 20 years and gone after 50. Magnetic media is not for archiving either. Even worse than optical technology for the purpose.
Archival media is chemical media. Principally microfiche, and archive-grade film stock (all those reports you hear of old films degrading and needing expensive restoration are because nobody bothered to copy it to archival-grade stock - the regular stuff only lasts a decade or so). You get expected lifetimes of over 100 years with that stuff. If you want to keep your data for extended periods of time, fork over the money to get it archived properly.
If you want all that without spending any money on it, then I'd like a pony.
The government, exactly the same as they do today. Those things cost money, but the government pays for them (via grants and the welfare state). It is a myth that private industry is pouring money into these things. It isn't.
And therein lies the issue. *If* your system is able to parse natural language correctly, all your interface problems are solved. The problem is that saying "understand natural language" actually means "think like a human", which means the problem is AI-complete. At that point the statement becomes less interesting - it says that if we only have to deal with human-like entities, then we won't have a problem interfacing with computers.
Furthermore, if the system can do that, it's probably going to be self-aware, so it will want to be paid. While there may be some advantages to employing an AI instead of a regular meatbag, you can already hire somebody to translate your spoken commands into terms a computer can understand, so nothing fundamental will have changed.
On the other hand, if you only deal with approximate natural language parsers, you'll probably find that while an expert user can get by very well with them, the ignorant 99% of the workforce will barely be able to use them at all, because they don't know how to structure their thoughts in a way that the parser can understand.
Purchasing of software for this purpose is often highly political. No attempt is made to choose the best software for the job - instead, the one which pays the largest kickbacks is chosen. Replacing it is not feasible at most sites.
Cash registers are primarily a hardware device, not a software one. They're proprietary hardware that won't run anything other than what the manufacturer wants to ship. If you're running a cash-register-based system, your inventory management will probably have to be tightly integrated with them, so the manufacturer will leverage that to force their product on you across the whole enterprise.
On the other hand, if you can just slap a regular computer down at each sales point, then compiere (erp/crm/pos) or oscommerce (web-driven sales) already provides a complete service, including full support and consultancy packages.
That's not "useful", it's another art course. Someone is willing to pay six-figure amounts for all kinds of artwork - but most people aren't going to be creating it.
The US has tried, twice, to push software patents onto Europe. They have failed, twice (right now they're trying again). Software patents are not legal anywhere in Europe. Strangely, the European Patent Office rubber-stamps them anyway (perhaps because they expect them to be legal in the future), but they're worthless paper in court. All significant patent lawsuits in Europe to date have been targeting the sale of hardware devices.
If you think that's the real problem with RPM, you've never looked at the source. I can forgive the screwey and kinda limited file format - but the unmaintainable pile of vomit and demon bile that they call a source tree is a mortal sin against humanity. It contains slightly-forked copies of (and this list is incomplete): libdb, file, popt and sqlite (and the changes to these are not really documented). Variable names are frequently either hungarian, or just random single characters. Comments are almost non-existent in most of the code, except for doxygen tags.
Most of the reasons why RPM is irritatingly limited stem from the fact that very few people can bear to do any work on the thing. It's not something you'd want to touch unless you were getting paid really well.
Astroturfing is when a corporation pretends to be a grassroots campaign. That doesn't appear to be the case here - this guy is just a regular marketdroid, no pretending to be anything. Still a slimeball, but a different breed of slimeball.
Show me an OS without flies in the ointment, and I'll show you which cult you belong to. That's not a meaningful objection to any of them.
Only in the US do people seriously think that you can "own technologies". The rest of the world has, so far, rejected the notion of software patents (despite much effort from the US to get them more widely accepted). There will not be any Linux distributions being shut down, there will only be development moving out of the US.
Software patents are troll material anyway. There is no such thing as a piece of software that does not violate numerous software patents without a license. That includes Windows; Microsoft get sued for patent violations two or three times a year, and they usually just settle (it's not usually newsworthy). The system is so broken that you can't win, unless you simply refuse to play.
The really sad part is that if he's found guilty, the system will still be just as broken, but people will think it's fair. Which is why the system remains broken.