release cycle (revisited)
 

Hi,

I'm sticking with Slackware since 1995, I like the philosphy,
but wonder about the last dilated release cycles. I mean if
the target group is mostly old-timers + a few adventurous types,
that's ok.

But many newbies will be seriously put off if they try out
the last "stable" version and their system will fail to install
a boot loader to NVME, or, if they manage to fix that, the
system stays off-line after reboot because the kernel won't
detect their recent gigabit ethernet adapter (e1000e).

I know there is a backlog of things to do, and we're all
happy for every patch and fix PV and the team develop or
adopt from other enthusiasts, but hw/sw development has
expanded and cycles have shortened compared with 20 years
ago. I greatly appreciate focus on stability and relatively
conservative view on adding new features, but having release
cycles over 2 years as a norm would de facto turn Slackware
into a rolling distro.

But hey, maybe it's all in the distro name... :)

And what is wrong with it?

Should we all chasing whatever new while drinking our kool aid? As far as I know, not everything new means better :)

Having _stable_ systems _is_ a big deal for _me_ (and the systems that i care about)

Not only slackware is stable, but compiling a kernel and using it for newer hardware _is_ much more straight forward and successful to do compared with any other (big) distros. The same is true for updating and _recovering_ for almost every single package
of the whole dstribution. Moreover almost all of the time everything just works, provided some common sense.

KISS :*

Go over to the Ubuntu forums and look at how much trouble people have with the supposedly well-tested LTS versions that are released every two years during the first several months after a release. I'll stick with a distro that releases a new version when it's ready, thank you very much.

Maybe not change the release cycle but add the latest LTS Kernel to extra for the most recent stable release after a period in Current. The 4.14.XX series is stable enough to do that.

Sure, but I described a different problem above, it was not about you or me.
If you cannot serve a bootable system in your last stable version, many
new users will look elsewhere, even to Ubuntu.

So yes, maybe occasional upgrade to kernels/bootloaders would be safe
or even beneficial in most situations. For example, in 14.1 a forced security
update from 3.10.17 to 3.10.107 seemed to have solved occasional hangs for me
while starting X. But the change should preferably be done to the original
installation tree, for example, after a certain period spent in patches/packages.

And what’s wrong with that?
Not every distro must cater to the lowest common denominator.

this statement includes a lot polemic. If the same amount of Ubuntu users with the same experience level in Linux would use Slackware, you could argue exactly the same and replace the distro name.

it is obvious that since some time, with the 14 release, the release Slackware cycles became longer, notable longer. look it up
https://en.wikipedia.org/wiki/Slackware#Releases

this has some effects, the OT mentioned NVM and new network hardware.
It must be OK to discuss this without such, not very productive, polemic comments.

You fail to realize that Ubuntu has a very large team working on it, where Slackware has PV and the chosen few with bits thrown in from users.

And personally, I like the release cycle of Slackware. Have you seen the release cycle of some distro's like Linux Mint? they've only been around since 2006 and are getting ready to release number 19. That's just crazy if you ask me.

I am very well aware of the fact, so please keep your verbal hammer in your pocket ;-)

And I don't care too much, if I want I run current. But this has no relation to the valid point the OT does, and starting a productive discussion without retoric hammers and empty phrases we all know here (and some like me start to become tired of them)

and openSuSE Leap was at 42.x and is now back to 15 :party:

It all depends what you're used to. ;) As a Mint user, a new major release every 2 years doesn't seem excessive to me, particularly as those releases can be quite distinct. I mean we're not talking Google Chrome or Dropbox version numbering here.

I have to agree with another poster though - you take the cake out of the oven once it's ready. One of the good things I like about Mint is that although they stick to a general schedule based on Ubuntu LTS, they won't release a new version until it's ready rather than choosing and sticking to a specific date.

Have you ever had a conversation without turning into a [removed personal attack]?

The release cycle becomes longer and longer because the operating system itself because more and more complex.

Heck! The 4.0 release shipped KDE and it still installed under a freaking 1GB! Now we are over 10GB, last I checked.

So, we are 2 years away from the 14.2 release, and the Plasma5 was not yet adopted, while Eric swears that it will be part of 15.0 ...

Did you really believe that our BDFL will adopt Plasma5 (which is roughly a third part of distro, as number of packages) on the last 100 meters before 15.0 release? Myself nope.

So, I bet on a current release cycle of 3 or 4 years long and in future it will become even longer.

I believe that the single solution to have shorter release cycles is to cut down the main distribution, specially putting the future Plasma5 in its own repository, officially sanctioned or not. ;)

That's true to some extent, but it's not the whole story. I was reading said Ubuntu forums recently, in a thread where someone was complaining about all the issues with the recently released 18.04 LTS. (That thread was eventually locked, as tends to happen over there when people criticize.) The general advice was to never install an LTS until the first bug fix release several months later, if you want something stable and well tested. If a large team of developers consistently does not have enough time to actually test things and squash major bugs before the fixed release date, that should tell you something about the downsides of that type of release cycle.

So, what do we do? We just admit to ourselves that Linux lags behind when it comes to support of the most modern hardware? Just try to relax while waiting for things to settle down a bit and become stable? Which will coincide with the next release of Slackware. :p

Is this the same "stable" Ubuntu that abandons its own in-house projects every couple of years to move on to the next biggest thing? And the next, and the next? The bug-ridden Ubuntu is the least stable OS I've ever tried. But at least it boots?