>And if I picked a Linux distro from 8 years ago I'd find some bugs in it too. Use the current version of Windows, it has fewer bugs.
Perhaps that last bit should read "Use the current version of Windows, it has fewer known bugs."
As software gets larger and more complex the tendency is for more bugs to be introduced. Open Source mitigates this problem by allowing everyone to inspect and fix it.
By that logic, if your company provided Windows 3.1 you would go around saying Windows has no multiuser support.
Windows has an interface for doing what you asked. It had a bug / didn't work as described so they fixed it. Your company wont buy the fix, but that doesn't mean the fix doesn't exist. It means you can't take advantage of it, sure.
But saying Windows doesn't support it is deliberately misrepresenting Windows to make it look bad. It's spreading FUD, and it's wrong.
Plus, why did you bring WMI into this? You can't seem to do it with WMI: http://msdn.microsoft.com/en-us/library/aa394595(VS.85).aspx "If you are not using DHCP, you cannot use WMI to disable a network connection". Were you just bringing WMI into it because it sounds scary and complicated and so would help shore up your FUD, or is there a way and this MSDN article is wrong?
You just said something monumental "Your company wont buy the fix, but that doesn't mean the fix doesn't exist. It means you can't take advantage of it, sure."
Thats one of the biggest things that Linux has going for it, if something breaks, YOU DON'T HAVE TO BUY THE FIX!
You can A. Fix it yourself; B. Get the latest source which probably has the fix (if it's a problem anyone else is having); or C. Filing a bug report and have the community fix it usually within a day, or even hours or minutes for really simple fixes.
Sure, Linux isn't for everyone, but for corporations and small businesses looking to save money, it looks better every day!
FOSS has many benefits, but I had always considered the fact that you can "fix it yourself" not to be one. Do any of you fix random bugs in your desktop applications? (Firefox, for example).
It might not be that real of an option for an average individual, but fixing it yourself in a large business could save you millions, depending on how big your IT Infrastructure is.
Think of this example. Company ABC has Windows XP deployed throughout their company. They need to upgrade some hardware, but the new hardware only has drivers for Windows 7 and Vista. Their only option (assuming a lot) is to upgrade to Windows 7 or Vista.
Let's say that they have 1,500 desktops that need to be upgraded. I'm not aware of the details of VLK license prices, but let's say this will cost $150/workstation, including labor to install it. Roughly approximating, this comes out to at least $225,000. This doesn't even count the amount of money they will spend battling incompatible software and hardware issues, after the upgrade.
--------------------------------
Company XYZ is using Linux Kernel 2.6.17, they have a problem with a piece of hardware that only has drivers that work with the 2.6.28 kernel. Company XYZ is a software development company, so they can just use their internal software developers to backport the driver from the 2.6.28 kernel to 2.6.17. They are already paying the developers a salary, so it won't cost them any more cash, maybe just a bit of productivity on other projects. This amounts to a huge savings over what Company ABC had to spend to fix a similar issue.
So wait, the thing I paid hundreds and hundreds of dollars for over the year won't give me the new fixes unless I fork over more money, but the free one will?
I can see not getting new features put into XP, but fixing BUGS should be free.
I didn't pay for bugs, I paid for software. If it's buggy and they don't want to support it, they should sell it at a discount. If they charge me for a complete product, but the product is buggy, they owe me more software.
It took me hours of research to do a simple down/up in Windows -- something that took me 30 seconds of research in Linux.
And you yourself illustrate all the odd and very long things you have to do in Windows to do a simple down/up of the NIC, and to get it to work correctly.
While in Linux, I can just do "sudo ifdown eth0"
All in all, I prefer the 30 second way, versus the five hour way.
By your very use of evidence, you proved my point. Cool.
You're still claiming that it took you hours of research to do in windows without acknowledging that you're using an ancient version and that it is a one line command in current versions.
I could counter that in the scope of all possible operating systems there's no reason disabling a network adapter should be a simple thing and that you're picking on it because you considered it an easy thing in Linux' favour.
I could counter how much better it is to type "tracert" on windows than "traceroute" on linux, or that windows PPTP setup wizard blows Linux out of the water, or that there are uncountable things I've spent hours reaearching on bob platforms because they aren't as simple as I want them to be.
Or I could say "yoruban wax potato parachute"; I suspect any of the preceeding would result in further argument about how I'm proving your point.
I set out to counter your claim that x is too difficult in windows, but this further back and forth - it's not going anywhere, is it?
