[tlug] Re: Just curious... how much impact does a kernel update make?

[John Fremlin <john@example.com>]

"Ian Wells" <ijw@example.com> writes:

> 2008/11/14 Edward Middleton <emiddleton@example.com>:
>> Dave M G wrote:
>>> Edward wrote:
>>>> Probably a change in the development model. ...
>>> Interesting. So does that mean reaching a 3.0 version is not an
>>> aspiration of the Linux developers?
>>
>> 3.0!!!  life begins at 10 ;)
>>
>> But in all seriousness, my impression is that its just a version not a
>> marketing number so unless something major changes or they decide
>> 2.6.257 doesn't work it may not change.
>
> There were significant differences between 2.4 and 2.6 by the time
> they'd finished.  2.6, however, has shown that adding architectures,
> drivers, filesystems and schedulers doesn't seem to cause enough
> change to bother with bumping the major version numbers.  Unless
> someone needs application binary incompatibility, meaning all your
> applications would need recompiling and retesting, I don't see why
> they'd go for 3.0 (or even 2.8).

Even the much maligned Linux kernel's ABI at least makes attempts to be
able to run old binaries on new kernels. ABI backwards compatibility
isn't as religiously adhered to as in for example Solaris but it is
aspired to and it's very difficult to imagine a world where 3.0 would
break everything.

Alan Cox suggested 3.0 an opportunity to delete lots of crufty old
drivers. 

http://lwn.net/Articles/296588/

But Linus decided against it.

http://lwn.net/Articles/298510/

At the end of the day, Linux 3 will arrive when and if Linus says it
has.