Kernel prepatch 4.0-rc1

[Posted February 23, 2015 by corbet]

Linus has closed the merge window for this release and released 4.0-rc1 — meaning, of course, that the current plan is to call the release "4.0". "But nobody should notice. Because moving to 4.0 does *not* mean that we somehow changed what people see. It's all just more of the same, just with smaller numbers so that I can do releases without having to take off my socks again." The codename has also changed to "Hurr durr I'ma sheep."

Kernel prepatch 4.0-rc1

Posted Feb 23, 2015 10:28 UTC (Mon) by cesarb (subscriber, #6266) [Link] (1 responses)

Note that if you have a program which gets confused by the 3.x kernel version, and are using the UNAME26 personality to make it see 2.6.40+x instead of 3.x, the version number as seen by that program just jumped backwards to 2.6.40. See override_release() at kernel/sys.c.

Kernel prepatch 4.0-rc1

Posted Feb 23, 2015 15:29 UTC (Mon) by josh (subscriber, #17465) [Link]

http://news.gmane.org/find-root.php?message_id=1424670192...

Kernel prepatch 4.0-rc1

Posted Feb 23, 2015 14:18 UTC (Mon) by lyda (subscriber, #7429) [Link] (10 responses)

It's so sad to see the loss in version number quality as we move from analogue to digital. Hopefully someone will maintain a gold plated versioning patch for those of us who appreciate high fidelity Linux kernel versions.

Kernel prepatch 4.0-rc1

Posted Feb 23, 2015 14:35 UTC (Mon) by paulj (subscriber, #341) [Link] (8 responses)

I have such a patch. The +++'s have been gold-plated to preserve the warmth of the analogue kernel version numbers. The patch comes with a special silver braided uuencoding^Wshielding to protect it from interference. Each patch is hand-crafted, with every character insulated from each other and kept free of oxygen to ensure excellent fidelity and dynamic range. This outstanding level of quality can be yours for our special super-low price of €1500!

Limited time offer only, so don't delay in getting in touch!

Analogues :-)

Posted Feb 23, 2015 17:30 UTC (Mon) by david.a.wheeler (subscriber, #72896) [Link] (4 responses)

I'm having trouble getting some of my 33 1/3 RPMs to run on it. Any suggestions? :-).

Analogues :-)

Posted Feb 23, 2015 17:33 UTC (Mon) by lyda (subscriber, #7429) [Link] (1 responses)

Have you tried using alien on your RPMs?

Analogues :-)

Posted Feb 24, 2015 18:26 UTC (Tue) by jcm (subscriber, #18262) [Link]

Have you tried digital new fidelity? DNF is definitely the way to go for those suffering from slower RPMs.

Analogues :-)

Posted Feb 23, 2015 17:39 UTC (Mon) by felixfix (subscriber, #242) [Link]

You need a trinary system.

Analogues :-)

Posted Feb 25, 2015 3:48 UTC (Wed) by paulj (subscriber, #341) [Link]

Try using a different spin of Fedora? :)

Kernel prepatch 4.0-rc1

Posted Feb 26, 2015 16:13 UTC (Thu) by Baylink (guest, #755) [Link] (2 responses)

Alas, your reply implies that the semantics of version numbering are unimportant. As a 30 year SA, for whom this stuff is important, I can tell you that nothing could be further from the truth.

Well: "systemd is the best thing since fried cheese". That could.

Kernel prepatch 4.0-rc1

Posted Feb 26, 2015 17:53 UTC (Thu) by zlynx (guest, #2285) [Link] (1 responses)

Version numbers mean different things in different projects.

From my experience the only meaningful numbers are minor version and patch level. The first number -- the major version -- is essentially set by Marketing when they decide that some new feature justifies it. Linux doesn't have a Marketing department so Linus has to just decide on his own.

Some projects use the major version to indicate compatibility breaks. PostgreSQL does it this way. That is probably what *you* mean by the semantics of version numbers.

When a product maintains full compatibility are they stuck on that major version? Say that Microsoft Word 2016 keeps the same document format as Word 2013. Does that mean its version should have stayed the same? Even if it added new features and changed its UI?

Lots of products bump the major version when the app does exactly the same thing but it now has prettier graphics.

Kernel prepatch 4.0-rc1

Posted Feb 27, 2015 14:11 UTC (Fri) by jezuch (subscriber, #52988) [Link]

> From my experience the only meaningful numbers are minor version and patch level.

The corollary is that (nowadays / for mature software) the version number matters only when you have N branches maintained concurrently (where N can be just as well equal to 1). And then it matters only inasmuch it allows distinguishing between concurrent branches and between versions in the same branch. Other than that - "the semantics of version numbering are unimportant", indeed, maybe except to folks in the marketing department.

Kernel prepatch 4.0-rc1

Posted Feb 23, 2015 18:21 UTC (Mon) by rsidd (subscriber, #2582) [Link]

Neil Young will help you get high-fidelity digital versions. Maybe your ears can't tell the difference, but in that case you're not the intended audience.

Any indication of adopting a version number system?

Posted Feb 23, 2015 22:55 UTC (Mon) by Felix_the_Mac (guest, #32242) [Link] (4 responses)

Did Linus say anything like "From now on numbers will go from x.9 to x+1.0" ?

Any indication of adopting a version number system?

Posted Feb 24, 2015 1:30 UTC (Tue) by mathstuf (subscriber, #69389) [Link] (3 responses)

Depends on whether you read

> so that I can do releases without having to take off my socks again

as something about this release or all future kernel releases.

Any indication of adopting a version number system?

Posted Feb 24, 2015 18:28 UTC (Tue) by jcm (subscriber, #18262) [Link] (2 responses)

I read this as shocking evidence that Linus has only 9 toes or that he has 19. In the latter case, I now understand all those socks and sandles comments over the years. Makes perfect sense. Must cause alarming chafing to occur.

Any indication of adopting a version number system?

Posted Feb 24, 2015 19:32 UTC (Tue) by mathstuf (subscriber, #69389) [Link]

Off-by-one error? We count from zero around here ;) .

Any indication of adopting a version number system?

Posted Feb 25, 2015 8:00 UTC (Wed) by tao (subscriber, #17563) [Link]

Linus is a hacker; surely he counts to 1023 using his fingers?

Kernel prepatch 4.0-rc1

Posted Feb 24, 2015 2:46 UTC (Tue) by bojan (subscriber, #14302) [Link] (1 responses)

So, avoid it, right? Because it's .0 and all that... :-)

Kernel prepatch 4.0-rc1

Posted Feb 26, 2015 16:14 UTC (Thu) by Baylink (guest, #755) [Link]

Or use SuSE; since they outlawed .0 distro releases, they'll probably bump the kernel release numbers as they pass through too.

Aggravated about this? Me?