openSUSE 12.1 Arrives: What's New and What Happened to 12.0? (Linux.com) [LWN.net]

Joe 'Zonker' Brockmeier takes a look at openSUSE 12.1. "openSUSE is also getting smarter about managing kernel upgrades. With 12.1, Poortvliet told me that Zypper now has an option (which isn't yet the default) to keep an older kernel until the system has successfully booted into an upgraded kernel. This solves the problem of systems being rendered unbootable because of a new kernel that has a hardware conflict of some sort. Right now, it's not the default, but users can enable this so that they can delete the old kernel after a successful boot - or keep up to two old kernels indefinitely. Speaking of kernels, openSUSE 12.1 comes with Linux 3.1. It also features a new init system pioneered by Fedora, systemd."

(Log in to post comments)

openSUSE 12.1 Arrives: What's New and What Happened to 12.0? (Linux.com)

Posted Nov 23, 2011 2:14 UTC (Wed) by Baylink (subscriber, #755) [Link]

What happened?

They, too, have succumbed to Version Numbering Silliness.

"People never touch .0 releases *because they're called .0*" -- they notionally say; not, y'know, *because they're the first revision in a new major release* -- "so let's lie, and not *call* it .0, and trick lots of people into using it and crashing *anyway*".

Yes, "crashing anyway" is hyperbole-and-a-half, but please, spare me: "let's lie" is *not* hyperbole, not even one little teeny tiny bit, and it breaks the contract which version numbers are between users and developers.

Yes, I'm *really* revved up about this; Firefox's "let's get rid of version numbers entirely, and oh yeah, let's piss everyone off by releasing a new major rev (and breaking everyone's plugins) every 6 weeks so they'll think it's *better* that way" and Asterisk following in the footsteps of Sun (Java) and SCO (Unix "4.0", which was actually SVr3.2"v4.0") were just the earliest examples of this crap.

For my part, I'm a systems administrator by trade -- you know, the guy who has to follow security advisories (which denominate safe and suspect versions by *version number*), and decide when to deploy updates and upgrades safely and how much labor will be involved (which you do, generally, by which component of a *version number* has changed); like most sysadmins, I live and die by version numbers, and I take this moronic self-important horsecrap *VERY* personally indeed, yes I do.

Given the zero comments here, and the 1701 comments sniping at Lennart, perhaps I'm the only person left who cares, but I do, very much.

openSUSE 12.1 Arrives: What's New and What Happened to 12.0? (Linux.com)

Posted Nov 23, 2011 23:56 UTC (Wed) by dlang (✭ supporter ✭, #313) [Link]

the silliness of eliminating a version number entirely was quickly dropped

as for the 'contract with users' about the version number, that may be a valid argument against skipping 12.0 and going directly to 12.1, but it has nothing to do with the issue of going from 4.1, 4.2, 4.3, etc to 4, 5, 6, 7, etc

both approaches can be legitimate, it all depends on the development model in use.

If the development model is "develop major new changes, release a version, fork it and update the old version with minor things while developing the next set of major new changes", then you should have a x.y style version number.

however if you have a rolling release process where every new version includes all pending fixes and new development, a single number is perfectly good.

openSUSE 12.1 Arrives: What's New and What Happened to 12.0? (Linux.com)

Posted Nov 24, 2011 23:05 UTC (Thu) by massimiliano (subscriber, #3048) [Link]

"so let's lie, and not *call* it .0, and trick lots of people into using it and crashing *anyway*".

Well, this is sort of off topic, but... another lie we live with is daylight saving time. And understanding it will tell you a lot about human beings.

Daylight savig time is a clever idea, it says something like "please, in summer, do everything one hour earlier, from when you wake up to when you go to bed".

Now, imagine asking people "start working at 8am sharp instead of 9am for six months"... how many of us would do it for more than a few days, or even for just one single day?

With daylight saving time, you just call "9am" what is actually "8am", and all of a sudden everybody diligently wakes up one hour earlier, every day!

All of this to say: as human beings, we are very, very prone to being manipulated by overt lies, and never object to that. Calling "12.1" a "12.0" release just exploits this fact...

Cheers,
Massi