Exactly, they are talking to companies who already use linux, and the conclusion is that they have had a positive enough experience of their existing linux use that they are considering increasing their use of linux. They're not hiding who they are and where their data comes from, unlike many of the supposedly "independent" surveys sponsored by microsoft which seem to come out every now and again.
It's also been many years since i've seen a large company which wasn't making at least some use of linux.
A lot of businesses are afraid of linux because it represents "the unknown"... It's not surprising that having tried it in some areas and found it to be up to the task businesses are losing that fear and are now considering expanding their use of linux.
If your replacing one server with another running the same os you aren't really adding any. It's also extremely common for companies to behave as you do, completely trashing and rebuilding windows servers rather than upgrading them - because windows installs tend to deteriorate with age. It's also more common to virtualise windows servers to keep applications separate from each other, and again on unix this is traditionally much less of a problem so its more common to have a more powerful server running multiple apps (not sure what your servers run, but you've already stated they're more powerful).
That said, the survey quoted percentages other than 100%, your company could easily have been among the percentage that isn't planning on accelerating linux use.
Vendor driven certs are worthless, wether they come from microsoft, cisco or redhat... Those vendors goals is not to educate people or even to ensure a high standard, they simply want more people out there promoting their products and having a large number of "qualified" cert holders helps more than a small number of "qualified and competent" cert holders.
Half is 50%, therefore 41% *is* less than half and 43% is nearly half. But you could argue the other way round, 43% is also less than half and 41% is also nearly half as both values are over 4/5 of the way to 50%.
One of the contributing factors towards Oracle/Sun having written most of the code was the terms under which you could submit code to the project... You had to assign copyright to Sun/Oracle, and many third parties were unwilling to do that.
Having a proprietary version for sale while being fully compatible with the free version actually helps improve the credibility of the suite among business users... There are a lot of suits who would never even consider downloading openoffice for free, but would be quite happy to buy staroffice from oracle.
RTF as originally specced was a very simple format.. What modern versions of word create when you tell them to save as rtf is nothing like the original spec for rtf, you would be better off saving in the doc format - just as proprietary, but more widely used so more effort goes into reverse engineering it.
When you talk about free software to the general public, most take it to mean "software that is available for no monetary cost"... Many users perceive free software to be pirated software, malware, or extremely poor quality binary only applications for windows downloadable from questionable sites.
I believe they are against stickers which advocate or promote proprietary software, eg "designed for windows"... Merely stating that a device is compatible with windows shouldn't be a problem.
Having open source drivers is generally enough, if those drivers are not included in the current version of redhat or slackware they will be included in the next, especially if the drivers get merged into the linux kernel.
There is a difference between saying hardware is designed for or made for a particular proprietary os, and saying the hardware is compatible with it. I have no objection with the latter assuming its true, but the former is active promotion of a third party product.
You are spot on about prisons being poor at rehabilitation... The current system actually works very poorly in this:
You go to prison for a relatively minor crime, and are thrown in with a bunch of people who are usually far more experienced criminals than you... You get mistreated (beaten up, raped etc) by the other prisoners and some of them teach you additional criminal skills. Once you come out, you've typically lost whatever you had before you went in, you are now more bitter, you are stigmatised by having a criminal record which makes it difficult for you to get legitimate work and to top it off you've learnt how to commit new or more effective crimes and have new criminal contacts. So for many, the only course of action open to them is to commit more crime, and this time they may get away with it for longer due to increased experience.
This is utterly stupid, why should the government (ie TAXPAYERS) pay content distributors? If anything, the government should put pressure on those distributors to lower prices, not pay them off.
Just because a company has been around for 30 years, doesn't mean it will still be around 15 years from now. Or even if it is, there is no guarantee they will still support their product.
Also support and lock-in are two completely different things, for example you can get Linux support from IBM, RedHat, Oracle, Novell, Canonical and many smaller players without needing to be locked in.
Support is often desirable, but lock-in never is. Having multiple choices for where you get that support from should always be a requirement.
By documentation, i assume you mean documentation regarding how the data is stored, so that worst case you can migrate that data into something else. There are literally thousands of companies out there who are tied into various legacy proprietary systems, and cannot get their data out of them in any useful form. Many of these systems are no longer supported, their vendors are long gone.
With OSS, the data is usually stored in as standard a format as possible because there is no incentive to keep it locked in.
In business you have to consider the worst case, and with OSS, the worst case is never as bad as with proprietary code. A company may seem financially stable and have a good reputation, but any investor knows things can go belly up very quickly.
Worst case with either OSS or proprietary, all the developers disappear or are killed etc... With OSS you have the sourcecode, and while it may be expensive and time consuming to hire developers and get them up to speed at least its possible.
Being FOSS is an advantage in itself... Everything else being equal, something which is FOSS is inherently superior to an equivalent product which is closed source.
Second rate software is everywhere, you just get to see it more easily when you can freely download the sourcecode from the internet.
I have worked with lots of proprietary software which is decidedly second rate, but still has flashy marketing behind it so you don't see just how bad it is until you get locked in.
The differences, are...
Open source is far less likely to try and lock you in, meaning you can easily get away from the second rate code when something better turns up. You don't pay a lot of money for it, people will often accept a slightly inferior product if it saves them money. You have the opportunity to fix it yourself, you aren't beholden to the vendor.
You however make a good point, a good balanced evaluation of multiple options is essential, and the inherent advantages of OSS are always a plus point in its favor, proprietary software has to be substantially better to win out over OSS in a balanced evaluation.
The problem is that very few places actually conduct proper evaluation, i have the misfortune to work with someone who's idea of evaluating products is to make a spreadsheet and compare the advertised feature set side by side while reading bought into evaluations from the likes of gartner. No OSS ever gets into the gartner lists directly (only when used as part of a commercial package) because it costs money to get on. He never actually tests anything properly, and is naive enough to believe that what vendors say in their marketing literature is 100% true and that the likes of gartner are actually impartial.
I have virtually never encountered any company which did a total cost of ownership study...
Windows requires a lot of initial setup and ongoing maintenance costs.
On linux the costs are different, but generally lower... The only area where linux costs are higher is the artificial costs of breaking out of windows lock-in, assuming you are migrating from windows. That said, breaking free of vendor lock-in is a very worthwhile action and will save huge amounts in the long run, whatever you migrate to.
If you're setting up a new network which is not already locked in, the setup time/costs for windows and linux are fairly comparable, but linux typically has far lower ongoing maintenance costs.
Windows can take a HUGE amount of work to setup correctly on a large deployment, you have to setup a domain, setup av, setup policies, setup deployment and configuration of apps, setup processes for updating windows and the applications, thoroughly test everything, setup a monitoring system...
Both systems allow you to skip important steps while still having an apparently working setup, and end up with shoddy results (99% of windows networks i've ever seen), so costs are often not added up properly. If you skimp out on the initial setup then your ongoing costs will be much higher and your security poor, and windows is generally much worse in this regard.
You also need competent staff, and sufficiently competent windows staff are just as expensive as good unix staff... Sure you can hire incompetent staff, and incompetent windows staff are far more widely available, but the end result will be extremely poor and they will end up costing more than hiring competent staff in the first place.
Most of the comparisons i've seen, were comparing a (fairly typical) poorly configured windows network, operated by insufficient numbers of poorly qualified staff - vs a linux based network setup correctly, operated by an adequate (or sometimes excessive) number of competent staff.
You get useless people in all walks of life... Useless unix admins, useless MCSEs, useless whatever else. This is why your supposed to filter them out during the interview process.
And also NEVER PUT ALL YOUR EGGS IN ONE BASKET, don't hire a single unix guy and expect him to run everything, hire 2 just incase one leaves or gets hit by a bus. And on that same note, don't get yourself locked into any proprietary systems for the same reason, if only one vendor makes it that vendor could easily stop making it, change it in ways you dont like or just simply go bust. And if you don't think huge companies like microsoft or oracle could go bust, you haven't been watching the news lately.
That same laptop could perform the same function using free software.
That same laptop is one out of many competing laptop designs, and over the years this competition has resulted in massive price decreases and technological improvements. Conversely, windows has improved far more slowly than hardware and is now more expensive than it used to be.
Software is only worth the value of any advantages it has over whatever free software exists, minus any disadvantages it has. If a proprietary application is equivalent or only marginally better than a free one, and yet costs $400 is it really worth it?
Compared to other OSes *that are actually sold*? The key point is that you can make an OS available for free, therefore that is the base price. For hardware, the base price is the lowest cost to manufacture (because it requires raw materials etc).
So compared to the actual cost of producing an OS, windows is extremely expensive. The fact that someone else can produce an equivalent product much cheaper is how the free market works, and now you must try to compete with that free product somehow. So far the method MS has chosen is to screw their customers by locking them in.
Only 15 years ago, x86 computers were also more expensive... In the grand scheme of things, software was relatively cheap. Today, windows makes up a significant portion of the cost, especially of a cheap machine.
Windows did nothing to bring the cost down, competition among hardware vendors brought the cost down, and a lack of competition on the software front has ensured that windows keeps getting more expensive even as the hardware to run it on becomes cheaper.
And yes, proprietary unix machines were stupidly expensive, because the vendors of them were greedy. Proprietary unix is now pretty much dead, primarily because they lacked most of the lock-in that microsoft has used.
Windows was always inferior to MacOS and AmigaOS. There were plenty of perfectly capable computers available cheaply, and there was free software available for most of them. Microsoft were largely just lucky to get a running start from IBM, and have then used various underhanded tricks such as lock-in to keep that position. Had there been no microsoft, someone else would have taken their place... They may have been a lot better or they may have been a lot worse, Microsoft pretty much sneaked in on the wave of standardised competitive hardware and stabbed everyone in the back.
Over all windows has been detrimental to society as a whole, you only need to look at the benefits that open and competitive hardware have brought us and imagine those same benefits brought across to software.
Slashdot Mirror
Exactly, they are talking to companies who already use linux, and the conclusion is that they have had a positive enough experience of their existing linux use that they are considering increasing their use of linux. They're not hiding who they are and where their data comes from, unlike many of the supposedly "independent" surveys sponsored by microsoft which seem to come out every now and again.
It's also been many years since i've seen a large company which wasn't making at least some use of linux.
A lot of businesses are afraid of linux because it represents "the unknown"... It's not surprising that having tried it in some areas and found it to be up to the task businesses are losing that fear and are now considering expanding their use of linux.
I find a lot of pentesters prefer gentoo, myself included... I wouldn't run backtrack unless i had an explicit need to do testing from a livecd.
If your replacing one server with another running the same os you aren't really adding any.
It's also extremely common for companies to behave as you do, completely trashing and rebuilding windows servers rather than upgrading them - because windows installs tend to deteriorate with age.
It's also more common to virtualise windows servers to keep applications separate from each other, and again on unix this is traditionally much less of a problem so its more common to have a more powerful server running multiple apps (not sure what your servers run, but you've already stated they're more powerful).
That said, the survey quoted percentages other than 100%, your company could easily have been among the percentage that isn't planning on accelerating linux use.
Vendor driven certs are worthless, wether they come from microsoft, cisco or redhat... Those vendors goals is not to educate people or even to ensure a high standard, they simply want more people out there promoting their products and having a large number of "qualified" cert holders helps more than a small number of "qualified and competent" cert holders.
Half is 50%, therefore 41% *is* less than half and 43% is nearly half. But you could argue the other way round, 43% is also less than half and 41% is also nearly half as both values are over 4/5 of the way to 50%.
Jaguar recently built a turbine-electric prototype hybrid:
http://www.wired.com/autopia/2010/09/paris-auto-show-jaguar-cx75/
One of the contributing factors towards Oracle/Sun having written most of the code was the terms under which you could submit code to the project... You had to assign copyright to Sun/Oracle, and many third parties were unwilling to do that.
Having a proprietary version for sale while being fully compatible with the free version actually helps improve the credibility of the suite among business users... There are a lot of suits who would never even consider downloading openoffice for free, but would be quite happy to buy staroffice from oracle.
RTF as originally specced was a very simple format..
What modern versions of word create when you tell them to save as rtf is nothing like the original spec for rtf, you would be better off saving in the doc format - just as proprietary, but more widely used so more effort goes into reverse engineering it.
I hate it when i go to read a news story, or a howto or something else online and it's only available in video form...
Especially technical guides, where a howto would let me cut and paste but a video won't...
You track down the handsets, shut them off and arrest the owners - making bogus 911 calls is illegal too.
When you talk about free software to the general public, most take it to mean "software that is available for no monetary cost"... Many users perceive free software to be pirated software, malware, or extremely poor quality binary only applications for windows downloadable from questionable sites.
Because compatibility with the kernel is often not enough, depending what the hardware is you might need userland programs to interact with it.
I believe they are against stickers which advocate or promote proprietary software, eg "designed for windows"...
Merely stating that a device is compatible with windows shouldn't be a problem.
Having open source drivers is generally enough, if those drivers are not included in the current version of redhat or slackware they will be included in the next, especially if the drivers get merged into the linux kernel.
There is a difference between saying hardware is designed for or made for a particular proprietary os, and saying the hardware is compatible with it. I have no objection with the latter assuming its true, but the former is active promotion of a third party product.
You are spot on about prisons being poor at rehabilitation...
The current system actually works very poorly in this:
You go to prison for a relatively minor crime, and are thrown in with a bunch of people who are usually far more experienced criminals than you... You get mistreated (beaten up, raped etc) by the other prisoners and some of them teach you additional criminal skills.
Once you come out, you've typically lost whatever you had before you went in, you are now more bitter, you are stigmatised by having a criminal record which makes it difficult for you to get legitimate work and to top it off you've learnt how to commit new or more effective crimes and have new criminal contacts.
So for many, the only course of action open to them is to commit more crime, and this time they may get away with it for longer due to increased experience.
This is utterly stupid, why should the government (ie TAXPAYERS) pay content distributors?
If anything, the government should put pressure on those distributors to lower prices, not pay them off.
Just because a company has been around for 30 years, doesn't mean it will still be around 15 years from now. Or even if it is, there is no guarantee they will still support their product.
Also support and lock-in are two completely different things, for example you can get Linux support from IBM, RedHat, Oracle, Novell, Canonical and many smaller players without needing to be locked in.
Support is often desirable, but lock-in never is. Having multiple choices for where you get that support from should always be a requirement.
By documentation, i assume you mean documentation regarding how the data is stored, so that worst case you can migrate that data into something else.
There are literally thousands of companies out there who are tied into various legacy proprietary systems, and cannot get their data out of them in any useful form. Many of these systems are no longer supported, their vendors are long gone.
With OSS, the data is usually stored in as standard a format as possible because there is no incentive to keep it locked in.
In business you have to consider the worst case, and with OSS, the worst case is never as bad as with proprietary code.
A company may seem financially stable and have a good reputation, but any investor knows things can go belly up very quickly.
Worst case with either OSS or proprietary, all the developers disappear or are killed etc... With OSS you have the sourcecode, and while it may be expensive and time consuming to hire developers and get them up to speed at least its possible.
Being FOSS is an advantage in itself...
Everything else being equal, something which is FOSS is inherently superior to an equivalent product which is closed source.
Second rate software is everywhere, you just get to see it more easily when you can freely download the sourcecode from the internet.
I have worked with lots of proprietary software which is decidedly second rate, but still has flashy marketing behind it so you don't see just how bad it is until you get locked in.
The differences, are...
Open source is far less likely to try and lock you in, meaning you can easily get away from the second rate code when something better turns up.
You don't pay a lot of money for it, people will often accept a slightly inferior product if it saves them money.
You have the opportunity to fix it yourself, you aren't beholden to the vendor.
You however make a good point, a good balanced evaluation of multiple options is essential, and the inherent advantages of OSS are always a plus point in its favor, proprietary software has to be substantially better to win out over OSS in a balanced evaluation.
The problem is that very few places actually conduct proper evaluation, i have the misfortune to work with someone who's idea of evaluating products is to make a spreadsheet and compare the advertised feature set side by side while reading bought into evaluations from the likes of gartner. No OSS ever gets into the gartner lists directly (only when used as part of a commercial package) because it costs money to get on. He never actually tests anything properly, and is naive enough to believe that what vendors say in their marketing literature is 100% true and that the likes of gartner are actually impartial.
I have virtually never encountered any company which did a total cost of ownership study...
Windows requires a lot of initial setup and ongoing maintenance costs.
On linux the costs are different, but generally lower... The only area where linux costs are higher is the artificial costs of breaking out of windows lock-in, assuming you are migrating from windows. That said, breaking free of vendor lock-in is a very worthwhile action and will save huge amounts in the long run, whatever you migrate to.
If you're setting up a new network which is not already locked in, the setup time/costs for windows and linux are fairly comparable, but linux typically has far lower ongoing maintenance costs.
Windows can take a HUGE amount of work to setup correctly on a large deployment, you have to setup a domain, setup av, setup policies, setup deployment and configuration of apps, setup processes for updating windows and the applications, thoroughly test everything, setup a monitoring system...
Both systems allow you to skip important steps while still having an apparently working setup, and end up with shoddy results (99% of windows networks i've ever seen), so costs are often not added up properly. If you skimp out on the initial setup then your ongoing costs will be much higher and your security poor, and windows is generally much worse in this regard.
You also need competent staff, and sufficiently competent windows staff are just as expensive as good unix staff... Sure you can hire incompetent staff, and incompetent windows staff are far more widely available, but the end result will be extremely poor and they will end up costing more than hiring competent staff in the first place.
Most of the comparisons i've seen, were comparing a (fairly typical) poorly configured windows network, operated by insufficient numbers of poorly qualified staff - vs a linux based network setup correctly, operated by an adequate (or sometimes excessive) number of competent staff.
You get useless people in all walks of life...
Useless unix admins, useless MCSEs, useless whatever else.
This is why your supposed to filter them out during the interview process.
And also NEVER PUT ALL YOUR EGGS IN ONE BASKET, don't hire a single unix guy and expect him to run everything, hire 2 just incase one leaves or gets hit by a bus. And on that same note, don't get yourself locked into any proprietary systems for the same reason, if only one vendor makes it that vendor could easily stop making it, change it in ways you dont like or just simply go bust. And if you don't think huge companies like microsoft or oracle could go bust, you haven't been watching the news lately.
That same laptop could perform the same function using free software.
That same laptop is one out of many competing laptop designs, and over the years this competition has resulted in massive price decreases and technological improvements. Conversely, windows has improved far more slowly than hardware and is now more expensive than it used to be.
Software is only worth the value of any advantages it has over whatever free software exists, minus any disadvantages it has. If a proprietary application is equivalent or only marginally better than a free one, and yet costs $400 is it really worth it?
Compared to other OSes *that are actually sold*? The key point is that you can make an OS available for free, therefore that is the base price. For hardware, the base price is the lowest cost to manufacture (because it requires raw materials etc).
So compared to the actual cost of producing an OS, windows is extremely expensive. The fact that someone else can produce an equivalent product much cheaper is how the free market works, and now you must try to compete with that free product somehow. So far the method MS has chosen is to screw their customers by locking them in.
Only 15 years ago, x86 computers were also more expensive... In the grand scheme of things, software was relatively cheap. Today, windows makes up a significant portion of the cost, especially of a cheap machine.
Windows did nothing to bring the cost down, competition among hardware vendors brought the cost down, and a lack of competition on the software front has ensured that windows keeps getting more expensive even as the hardware to run it on becomes cheaper.
And yes, proprietary unix machines were stupidly expensive, because the vendors of them were greedy. Proprietary unix is now pretty much dead, primarily because they lacked most of the lock-in that microsoft has used.
Windows was always inferior to MacOS and AmigaOS. There were plenty of perfectly capable computers available cheaply, and there was free software available for most of them. Microsoft were largely just lucky to get a running start from IBM, and have then used various underhanded tricks such as lock-in to keep that position.
Had there been no microsoft, someone else would have taken their place... They may have been a lot better or they may have been a lot worse, Microsoft pretty much sneaked in on the wave of standardised competitive hardware and stabbed everyone in the back.
Over all windows has been detrimental to society as a whole, you only need to look at the benefits that open and competitive hardware have brought us and imagine those same benefits brought across to software.