Very sweet idea, but no form of insurance is socialist.
The very first insurance, back in the 1700s when a number of ship owners met in a coffee house in London to discuss paying into a pot so if their ship was lost at sea they could pay for it out of the pot - yes, that may have been vaguely socialist.
But as soon as the amount of money in the pot and the amount of people putting money in became so great that there was some real sense in a man piping up "Tell you what. Why don't I look after the pot, keep track of who's paid in and who hasn't, pay out for people who've lost their ships and I get to keep whatever's left at the end of the year for doing it?" insurance has been all about making sure that as much as possible is left in the pot at the end of the year.
1) Extreme sports such as race car driving, motocross, ski jumping.
That's fair enough. 2) Cancer due to smoking
Plenty of non-smokers come down with a cancer in their respiratory system. The only way you'll prove that it's not smoking related is to pay a doctor to examine you, write you a letter saying "doesn't look smoking related" and take that letter to the insurance company. Who, if you're lucky, will accept it at face value. If you're unlucky, you'll have to take them to court.
While all of this is going on, you'll have to fund your own treatment. Who do you pay first, the doctor keeping you alive or the lawyer whose work will (hopefully) ultimately pay the doctor? Assuming you choose "pay doctor first", how are you then going to pay the lawyer?
3) Liver failure due to alcoholic consumption
Same problem there. Yes, cirrhosis is usually caused by too much booze - but again, if the insurance company gets the slightest whiff that this may be the case for you, I bet you anything you like they won't pay for a thing unless and until you've proven otherwise.
Don't bother - that was such an obvious troll it's unreal.
Doesn't much matter what happens with the Linux kernel anyhow - much of the userland (including nice things like glibc and gcc) will be GPLv3 if they're not already. And they are major work to replace.
Granted, GPLv3's been in the offing for some time. But I reckon so was the MS/Novell deal - these things don't happen overnight. Version 3 of the GPL actually has the potential to bring the OSS community closer together by making clear the issues surrounding things like software patents and preventing (or at least severely curatailing) similar deals.
We should be grateful that the only major player to take the Microsoft pill was Novell - it would be far worse if Redhat and Canonical had as well.
'Twasn't me, it was a former manager. I sincerely doubt it was bulletproof - Excel on its own has an "Open with VB disabled for debugging" option to see to that, and OpenOffice completes the deal nicely.
However, if Sab-Ox (or indeed any legislation) decreed that everything be 100% tamper proof (or at least that any tampering must be guaranteed, 100% evident) such that no man, however determined, could ever cook the books, all you US folk would be submitting accounts chiselled into blocks of stone and getting them countersigned by Aretha Franklin. It's called "reasonable care" and "due diligence" - you do all you can to make it hard to mess around with the system. The more you do, the more people have to be involved in any attempt to break it - and the more likely it is to be spotted.
I have my doubts it will help - the directors of Enron, Global Crossing et al were quite happy to break the law as it stands, why should they be bothered about breaking one more? - but it puts your regulators in a much stronger position to spot these things before they become a big deal.
3 years ago I was looking for work. I found what looked like an interesting little role as IT manager (read: organ grinder and monkey, but mostly monkey) in a small company not far from where I live. The MD was clearly a bright man, business-wise. He'd taken a business which was hardly doing anything substantial and turned it into one which was making something like £6-10 million profit per annum with under 100 staff.
He was quite happy to explain that he'd done that by not spending money unnecessarily and by not employing people when he could get by without them. By way of example, he explained how they used a bunch of Excel spreadsheets with some clever scripting to handle a large chunk of the business - and about a year previously, he'd found that the workbook was getting rather unwieldy and difficult to work with. At the time, he was rather concerned that he may have to throw out the Excel spreadsheet and go out and buy a commercial package to do the work instead.
Then he discovered that it's possible to link formulas and cells between separate Excel files. Hallelujah! Suddenly he could make the spreadsheet much bigger without the management headache of so many separate sheets in one file!
I'm prepared to put £50 on the table now which says I could find half a dozen companies with a similar unholy mashup of spreadsheets in use somewhere in less than a day.
And you're telling me that these places could easily convert to OpenOffice?
IME, 9 times out of 10 the most sensible solution is "stick with what you've got, tweaking where necessary" unless there is major breakage with it that is fundamentally unfixable - you can't polish a turd, you can't make an inkjet printer cheap to run and you can't make 100 Windows 95/98 machines stable and reliable. Better the devil you know and all that.
On the other hand, I think it's rather amusing that I was trying to bring some upbeat, sensible advice to a question which had so far mostly seen answers along the lines of "you're wasting your time, don't bother" - and the first reply I got was "you're wasting your time, don't bother". I'd love to know where all these managers who get taken out for fancy lunches are exactly because I've never seen any real evidence for it in 5 years - and 2 of those as a manager.
Software price/performance/ease of use quite often isn't a straight line on a graph.
In fact, if you just model price (on the X axis) against out of the box functionality, it's often a bellcurve.
At the very cheap/free end, you've got the likes of Slakware, Debian and Fedora linux. Can do most things, but out of the box quite often needs further work.
Then you start spending a bit more money, and you get Windows and RHEL. Does quite a bit out of the box.
Then you start spending lots of money, and you get things like VMS and AIX (complete with the proprietary hardware they entail). Unlike with Windows, these products are built on the assumption that the purchaser doesn't need to be told how to run their systems, and so out of the box provide remarkably little. What they do provide is a framework for building your own infrastructure - and when such a system is configured to do X, it will generally sit there doing X for years on end quite happily. There's still more than one VMS system out there which has an uptime longer than Windows 2003 and its NT based predecessors have existed for.
With the advent of Linux and the popularity of the GPL, operating systems have had this curve messed around with substantially - and the "free" end of the market provides a lot more functionality than it did 10 or even 5 years ago. This is also spreading to other parts of IT such as CRM systems - witness vTiger, for instance. However, in systems which are still dominated by large software companies (things like financial software immediately spring to mind, though I'm sure there are plenty of others), it still holds true.
This is why Wine has been such an amazing success, why OpenOffice is practically indistinguishable from Microsoft Office and why Samba is able to replicate the functionality of a Windows 2000 Domain only 7 years after Windows 2000 was released so flawlessly.
Let me give you a suggestion what I hope will be more constructive.
Firstly, I'm going to discount the "fancy lunches" thing. I don't think it's as prevalent as some on here would have you believe, and even if it is in this case it's not the kind of thing you can easily fight against. This leaves us with "managing the business properly".
Any (sane) business owner/high level manager doesn't spend any serious quantity of money unless there is a clear business benefit. Remember those words: "Business benefit".
Now, a business benefit boils down to one of two things:
1. Helps the company make money. 2. Helps the company save money.
Every other reason, once you've drilled down far enough, ultimately boils down to this. For example, "Reduce risk to the business in the event of trouble" is just another way of saying "There's a strong chance that if something goes wrong, it will cost us a small fortune. This purchase either reduces the likelihood of something going wrong, or it reduces the size of that "small fortune". In other words, it saves money."
This, by the way, is precisely why management often have trouble understanding why software would be given away for free ("where's the business benefit?") and also why most of Microsoft's FUD has been along the lines of "Windows costs less than Linux".
Understanding this means that you can now ask yourself/your manager what the perceived business benefit of such a move would be. There is a possibility (unlikely but not entirely unknown) that there is a genuine business reason you haven't considered which, with the best will in the world, does provide a solid business reason. If this is the case: live with it or leave. You were employed to do a technical job, not preach a religion.
If not: get organised. List the pros and cons of each solution (including your current one), emphasising the things which are likely to be of concern to those higher up than you.
Getting upset and having a moan on/. won't solve anything.
The type of spreadsheet I'm talking about, Excel making the odd incorrect calculation is the least of your problems.
Besides, my understanding of Sab-Ox is that it makes spreadsheets an absolute minefield - because spreadsheets make it trivially easy to change things, save it under alternative names and otherwise mess about with the numbers with no audit trail. My former manager has apparently succeeded in making a specific spreadsheet compliant - that was with a team of a few people basically reskinning Excel with VB macros and the like so the user interface looked similar but kept audit trails, enforced per-user access control on parts of the spreadsheet and removed functionality which was completely at odds with the regulations.
"Compatability with office" is a difficult thing. Particularly as most computing people don't actually make full use of Office's capabilities.
OTOH, go to the accounts department in any reasonably-sized company and you'll see positively scary things going on with Excel. No, nobody needs 100% compatability. But there are enough different users using and abusing Office that no two people have the same list of "things which must work".
TBH, as soon as anyone says "$PRODUCT is not as Windows compatible as Windows", you can probably stop listening.
Windows is a proprietary software product. Much of what goes on under the hood is completely unknown - enough information has been reverse engineered for some interoperability (cf. Samba, ndiswrapper), but expecting any product to ever be as "Windows compatible as Windows" is asking for the moon.
I'm not married, but as soon as the remotest risk of me winning an argument comes up, my g/f always starts saying something like "Who cares, I'm not listening" and walks out. I'd wager that something similar is true for 90% of men in relationships.
OT, I know, but does that mean that someone who's living in a city could start to suffer respiratory problems (like asthma) despite no previous symptoms as a direct result of all the crap they're breathing in?
I don't see any earthly reason why they couldn't RAID them under the hood, as it were. One SATA port, a controller which presents to the host PC as a single drive but actually spreads the data across many flash chips.
Of course, then the controller on the drive would be a lot more complicated...
There have always been bad parents. Since the beginning of time. Even in the animal kingdom, you get the occasional mother which simply shows no interest whatsoever in her offspring.
Only these days, it's more obvious in people because it's the idiots who are breeding. Smarter teens have less sex.
How do you guarantee that nobody, no matter how determined, can possibly have interfered with the data that's on the underlying disk (and thus what the version control system returns when queried about a file's history)?
When the Channel Tunnel was being built, the company building it ran into financial difficulties. The bank was loath to push for bankruptcy, as they'd have been unlikely to get more than a fraction of their money back.
Just goes to show - if you owe the bank £100, you're in trouble. If you owe the bank £10,000,000, the bank's in trouble.
None of those problems you describe sound particularly fundamental to the package format. More, they are features(!) of the tool which installs them and the folks who create the packages.
Solaris does a lot of things very nicely. ZFS and DTrace spring to mind - and unlike Linux, Solaris AIUI can boot off ZFS.
A former lecturer when I was at university back in 2002 reckoned commercial Unix had 5-10 years left to live. Looks like he may not have been far off the mark - where they're not dying, commercial unixes are being made to look more like Linux and less commercial (witness OpenSolaris) by the month.
Slashdot Mirror
And pink unicorns might fly out of my butt.
If they do, can I have one?
Very sweet idea, but no form of insurance is socialist.
The very first insurance, back in the 1700s when a number of ship owners met in a coffee house in London to discuss paying into a pot so if their ship was lost at sea they could pay for it out of the pot - yes, that may have been vaguely socialist.
But as soon as the amount of money in the pot and the amount of people putting money in became so great that there was some real sense in a man piping up "Tell you what. Why don't I look after the pot, keep track of who's paid in and who hasn't, pay out for people who've lost their ships and I get to keep whatever's left at the end of the year for doing it?" insurance has been all about making sure that as much as possible is left in the pot at the end of the year.
I don't want to be insured for:
1) Extreme sports such as race car driving, motocross, ski jumping.
That's fair enough.
2) Cancer due to smoking
Plenty of non-smokers come down with a cancer in their respiratory system. The only way you'll prove that it's not smoking related is to pay a doctor to examine you, write you a letter saying "doesn't look smoking related" and take that letter to the insurance company. Who, if you're lucky, will accept it at face value. If you're unlucky, you'll have to take them to court.
While all of this is going on, you'll have to fund your own treatment. Who do you pay first, the doctor keeping you alive or the lawyer whose work will (hopefully) ultimately pay the doctor? Assuming you choose "pay doctor first", how are you then going to pay the lawyer?
3) Liver failure due to alcoholic consumption
Same problem there. Yes, cirrhosis is usually caused by too much booze - but again, if the insurance company gets the slightest whiff that this may be the case for you, I bet you anything you like they won't pay for a thing unless and until you've proven otherwise.
Don't bother - that was such an obvious troll it's unreal.
Doesn't much matter what happens with the Linux kernel anyhow - much of the userland (including nice things like glibc and gcc) will be GPLv3 if they're not already. And they are major work to replace.
I hope you won't mind if I say: BULLSHIT.
Granted, GPLv3's been in the offing for some time. But I reckon so was the MS/Novell deal - these things don't happen overnight. Version 3 of the GPL actually has the potential to bring the OSS community closer together by making clear the issues surrounding things like software patents and preventing (or at least severely curatailing) similar deals.
We should be grateful that the only major player to take the Microsoft pill was Novell - it would be far worse if Redhat and Canonical had as well.
Dog and horse semen are both very fragile; just whacking the container, or introducing a couple drops of water, will kill the sperm.
I hardly dare to ask this but... how do you know?
'Twasn't me, it was a former manager. I sincerely doubt it was bulletproof - Excel on its own has an "Open with VB disabled for debugging" option to see to that, and OpenOffice completes the deal nicely.
However, if Sab-Ox (or indeed any legislation) decreed that everything be 100% tamper proof (or at least that any tampering must be guaranteed, 100% evident) such that no man, however determined, could ever cook the books, all you US folk would be submitting accounts chiselled into blocks of stone and getting them countersigned by Aretha Franklin. It's called "reasonable care" and "due diligence" - you do all you can to make it hard to mess around with the system. The more you do, the more people have to be involved in any attempt to break it - and the more likely it is to be spotted.
I have my doubts it will help - the directors of Enron, Global Crossing et al were quite happy to break the law as it stands, why should they be bothered about breaking one more? - but it puts your regulators in a much stronger position to spot these things before they become a big deal.
Let me tell you a story.
3 years ago I was looking for work. I found what looked like an interesting little role as IT manager (read: organ grinder and monkey, but mostly monkey) in a small company not far from where I live. The MD was clearly a bright man, business-wise. He'd taken a business which was hardly doing anything substantial and turned it into one which was making something like £6-10 million profit per annum with under 100 staff.
He was quite happy to explain that he'd done that by not spending money unnecessarily and by not employing people when he could get by without them. By way of example, he explained how they used a bunch of Excel spreadsheets with some clever scripting to handle a large chunk of the business - and about a year previously, he'd found that the workbook was getting rather unwieldy and difficult to work with. At the time, he was rather concerned that he may have to throw out the Excel spreadsheet and go out and buy a commercial package to do the work instead.
Then he discovered that it's possible to link formulas and cells between separate Excel files. Hallelujah! Suddenly he could make the spreadsheet much bigger without the management headache of so many separate sheets in one file!
I'm prepared to put £50 on the table now which says I could find half a dozen companies with a similar unholy mashup of spreadsheets in use somewhere in less than a day.
And you're telling me that these places could easily convert to OpenOffice?
Thank you.
IME, 9 times out of 10 the most sensible solution is "stick with what you've got, tweaking where necessary" unless there is major breakage with it that is fundamentally unfixable - you can't polish a turd, you can't make an inkjet printer cheap to run and you can't make 100 Windows 95/98 machines stable and reliable. Better the devil you know and all that.
On the other hand, I think it's rather amusing that I was trying to bring some upbeat, sensible advice to a question which had so far mostly seen answers along the lines of "you're wasting your time, don't bother" - and the first reply I got was "you're wasting your time, don't bother". I'd love to know where all these managers who get taken out for fancy lunches are exactly because I've never seen any real evidence for it in 5 years - and 2 of those as a manager.
Software price/performance/ease of use quite often isn't a straight line on a graph.
In fact, if you just model price (on the X axis) against out of the box functionality, it's often a bellcurve.
At the very cheap/free end, you've got the likes of Slakware, Debian and Fedora linux. Can do most things, but out of the box quite often needs further work.
Then you start spending a bit more money, and you get Windows and RHEL. Does quite a bit out of the box.
Then you start spending lots of money, and you get things like VMS and AIX (complete with the proprietary hardware they entail). Unlike with Windows, these products are built on the assumption that the purchaser doesn't need to be told how to run their systems, and so out of the box provide remarkably little. What they do provide is a framework for building your own infrastructure - and when such a system is configured to do X, it will generally sit there doing X for years on end quite happily. There's still more than one VMS system out there which has an uptime longer than Windows 2003 and its NT based predecessors have existed for.
With the advent of Linux and the popularity of the GPL, operating systems have had this curve messed around with substantially - and the "free" end of the market provides a lot more functionality than it did 10 or even 5 years ago. This is also spreading to other parts of IT such as CRM systems - witness vTiger, for instance. However, in systems which are still dominated by large software companies (things like financial software immediately spring to mind, though I'm sure there are plenty of others), it still holds true.
Ah, no, you're quite correct.
This is why Wine has been such an amazing success, why OpenOffice is practically indistinguishable from Microsoft Office and why Samba is able to replicate the functionality of a Windows 2000 Domain only 7 years after Windows 2000 was released so flawlessly.
Let me give you a suggestion what I hope will be more constructive.
/. won't solve anything.
Firstly, I'm going to discount the "fancy lunches" thing. I don't think it's as prevalent as some on here would have you believe, and even if it is in this case it's not the kind of thing you can easily fight against. This leaves us with "managing the business properly".
Any (sane) business owner/high level manager doesn't spend any serious quantity of money unless there is a clear business benefit. Remember those words: "Business benefit".
Now, a business benefit boils down to one of two things:
1. Helps the company make money.
2. Helps the company save money.
Every other reason, once you've drilled down far enough, ultimately boils down to this. For example, "Reduce risk to the business in the event of trouble" is just another way of saying "There's a strong chance that if something goes wrong, it will cost us a small fortune. This purchase either reduces the likelihood of something going wrong, or it reduces the size of that "small fortune". In other words, it saves money."
This, by the way, is precisely why management often have trouble understanding why software would be given away for free ("where's the business benefit?") and also why most of Microsoft's FUD has been along the lines of "Windows costs less than Linux".
Understanding this means that you can now ask yourself/your manager what the perceived business benefit of such a move would be. There is a possibility (unlikely but not entirely unknown) that there is a genuine business reason you haven't considered which, with the best will in the world, does provide a solid business reason. If this is the case: live with it or leave. You were employed to do a technical job, not preach a religion.
If not: get organised. List the pros and cons of each solution (including your current one), emphasising the things which are likely to be of concern to those higher up than you.
Getting upset and having a moan on
The keyword here is "knowingly".
The type of spreadsheet I'm talking about, Excel making the odd incorrect calculation is the least of your problems.
Besides, my understanding of Sab-Ox is that it makes spreadsheets an absolute minefield - because spreadsheets make it trivially easy to change things, save it under alternative names and otherwise mess about with the numbers with no audit trail. My former manager has apparently succeeded in making a specific spreadsheet compliant - that was with a team of a few people basically reskinning Excel with VB macros and the like so the user interface looked similar but kept audit trails, enforced per-user access control on parts of the spreadsheet and removed functionality which was completely at odds with the regulations.
"Compatability with office" is a difficult thing. Particularly as most computing people don't actually make full use of Office's capabilities.
OTOH, go to the accounts department in any reasonably-sized company and you'll see positively scary things going on with Excel. No, nobody needs 100% compatability. But there are enough different users using and abusing Office that no two people have the same list of "things which must work".
TBH, as soon as anyone says "$PRODUCT is not as Windows compatible as Windows", you can probably stop listening.
Windows is a proprietary software product. Much of what goes on under the hood is completely unknown - enough information has been reverse engineered for some interoperability (cf. Samba, ndiswrapper), but expecting any product to ever be as "Windows compatible as Windows" is asking for the moon.
Not a million miles from the truth.
I'm not married, but as soon as the remotest risk of me winning an argument comes up, my g/f always starts saying something like "Who cares, I'm not listening" and walks out. I'd wager that something similar is true for 90% of men in relationships.
In the 20 years of doing this, not one of their employees has had any lung problems.
Or if they have, they've died so quickly that nobody had a chance to think "hang on, they've spent years working in a plant full of printer toner..."
OT, I know, but does that mean that someone who's living in a city could start to suffer respiratory problems (like asthma) despite no previous symptoms as a direct result of all the crap they're breathing in?
I don't see any earthly reason why they couldn't RAID them under the hood, as it were. One SATA port, a controller which presents to the host PC as a single drive but actually spreads the data across many flash chips.
Of course, then the controller on the drive would be a lot more complicated...
Let me tell you a little secret.
There have always been bad parents. Since the beginning of time. Even in the animal kingdom, you get the occasional mother which simply shows no interest whatsoever in her offspring.
Only these days, it's more obvious in people because it's the idiots who are breeding. Smarter teens have less sex.
But if there is, I'm sure the ICBM will be easy to use and the resulting explosion very pretty.
How do you guarantee that nobody, no matter how determined, can possibly have interfered with the data that's on the underlying disk (and thus what the version control system returns when queried about a file's history)?
Not always true.
When the Channel Tunnel was being built, the company building it ran into financial difficulties. The bank was loath to push for bankruptcy, as they'd have been unlikely to get more than a fraction of their money back.
Just goes to show - if you owe the bank £100, you're in trouble. If you owe the bank £10,000,000, the bank's in trouble.
None of those problems you describe sound particularly fundamental to the package format. More, they are features(!) of the tool which installs them and the folks who create the packages.
Solaris does a lot of things very nicely. ZFS and DTrace spring to mind - and unlike Linux, Solaris AIUI can boot off ZFS.
A former lecturer when I was at university back in 2002 reckoned commercial Unix had 5-10 years left to live. Looks like he may not have been far off the mark - where they're not dying, commercial unixes are being made to look more like Linux and less commercial (witness OpenSolaris) by the month.