first they change the OS name with the excuse of 3rd party software looking for 199x era strings

I'm pretty sure that was never confirmed by MS to be a reason, just something 'some dude' wrote somewhere which is now treated as truth..

I'm willing to be proven wrong of course, but naming tends to be more about long term promotional/prv/advertising than 'dumb code' if only because they could probably catch 99% of that software doing the dumb thing and spoof the return to be Win7 or whatever if needs be.

Indeed, they didn't officially confirm it.
But some dude claiming to work at Microsoft said that that was the technical reason. And a simple examination of public known source code proved him correct, there is a horrible amount of software doing bad practices. Who knows about propietary software.

Probably this is the case of someone raising the very real concern from the technical side (putting a '9' will bring us trouble!) and the marketing guys got the perfect excuse to bump the number to 10.

And when marketing side hits and wins again technical side... A cuite kitten die D:

https://buildfeed.net/

And remember when the Linux kernel changed from 2.X.YY version numbering to 3.XX, changing not only the numbers but the format of the version string as well, despite all the whining and crying of people afraid that the world would break if they dared to bump the major version number?

Yeah, me neither.

You cannot compare Linux and Windows worlds.

On Linux when there is a kernel update, you usually update all your software that comes with the distro you are using (if necessary), the guys of your distro take care of upgrading the software provided in the repository for you. If you do not use a specific distro, you have to do it by yourself (when possible, ie if the application provides the sources in some way).

If you are running a closed source piece of software on Linux and you have issues after a kernel upgrade you are powerless.

Backwards compatibility is not a important goal on Linux, never had been. Linux is not meant to be backwards compatible, this is one of its most strength on the server side and one of its most weakness on the consumer side. Imo.

And remember when the Linux kernel changed from 2.X.YY version numbering to 3.XX, changing not only the numbers but the format of the version string as well, despite all the whining and crying of people afraid that the world would break if they dared to bump the major version number?

Yeah, me neither.

Going to 3.1 broke a *LOT* of stuff .

Of course at that time the file trees in most of the distros were a major mess - affecting everything from the ( already buggy ) sound to graphics .

Kernel version change confirmed, marketing guys won. .-.

http://msdn.microsoft.com/en-us/library/hh869301.aspx

Now Creative audio cards will have tons of problems (just an example), not only unsaved settings :|

Dear Microsoft Marketing guys, hope you get the squits for a weak at least.

Wah wah wah... it's a good few months away yet, any software worth its salt will be upgraded and anything that isn't is probably just shit and shouldn't be used.

I'd rather have a sane kernel version number than to be forever held back by crap developers who couldn't manage something as simple as a '>=' in a conditional.

Besides, as I mentioned, it's not like they can't detect broken programs and just spoof the kernel number if they are that important.

In short feel free to stop crying about this...

Wah wah wah... it's a good few months away yet, any software worth its salt will be upgraded and anything that isn't is probably just shit and shouldn't be used.

I'd rather have a sane kernel version number than to be forever held back by crap developers who couldn't manage something as simple as a '>=' in a conditional.

Besides, as I mentioned, it's not like they can't detect broken programs and just spoof the kernel number if they are that important.

In short feel free to stop crying about this...

Of course for current project it's not an issue, never had been for current supported software.

The issue is for backward compatibility and 3rd party drivers that probably they could run fine but the check for kernel version.

Little example:

Never had a Creative audio cards? Every time kernel NT 6 minor version changes, the audio settings were lost after the restart, but the rest works fine, even unsupported piece of hardware by Creative. Now I'm really curious to see what wonderful issues will have Creative audio cards owners.

Yes its always a problem related to bad programming, but if you can avoid them, why change the major kernel version? Marketing on a string that has only a meaning for developers?

Um, I think all this discussion would have been avoided if anybody had bothered to mention how retrieving the version works now:

http://msdn.microsoft.com/en-us/library/windows/desktop/ms724439%28v=vs.85%29.aspx

Since Windows 8.1 you're required to explicitly state in the manifest what's the newest Windows version your program knows about, and Windows will never return any version number larger than the one for that version. If you never state it then you'll be stuck with seeing 6.2 as the kernel version (i.e. Windows 8.0).

In other words, if your program sees 10.0 as the Windows version then it's because you told Windows that your program knows about it.

Um, I think all this discussion would have been avoided if anybody had bothered to mention how retrieving the version works now:

http://msdn.microsoft.com/en-us/library/windows/desktop/ms724439%28v=vs.85%29.aspx

Since Windows 8.1 you're required to explicitly state in the manifest what's the newest Windows version your program knows about, and Windows will never return any version number larger than the one for that version. If you never state it then you'll be stuck with seeing 6.2 as the kernel version (i.e. Windows 8.0).

In other words, if your program sees 10.0 as the Windows version then it's because you told Windows that your program knows about it.

It seems to be a nice solution.

Didn't know that, thank you.

Ok, I think all my concerns were wrong. Now it's only a Java problem :D