A 10-Way Linux Distribution Battle To Kick Off 2016

[edmon]

I never mentioned security problems myself, the main issue with testing is the maintenance burden compared to stable. The whole rolling release concept is completely wrong for servers.

there isn't any "maintenance burden " if you know what are you doing for everybody else is true that there some kind of "burden"!!!
do you know what exactly is rolling release? and what the rolling releases have in common with debian testing?

[darkbasic]

do you know what exactly is rolling release? and what the rolling releases have in common with debian testing?

No, I maintain 16 servers since 10 years and I daily use Debian Stable, Debian Testing, Arch Linux and Gentoo but I never heard about "rolling releases", maybe you can shed some light?

[Kano]

edmon

This is very interesting what you write, because Debian Testing/Unstable is/was in a huge libstdc++6 transition that affected lots of packages especially KDE ones. Maybe it did not affect servers without X. You should be really careful to use testing instead of a specific codename in the sources.list as a system can be broken easyly after a new release and ongoing transitions. Of course if you use sid instead of unstable you gain nothing. Basically the next stable codename is save to use in frozen state or in a nearly frozen state but not as general rule, this is just wrong.

darkbasic

Basically testing is a moving target, always pointing to the next stable release. Unstable points always to sid, so basically you can consider both as rolling releases - as testing mainly has just fewer packages and gets the unstable ones after a delay.

[linuxcbon]

@ phoronix, can you please add more distributions to your testings ?
- Mint
- Mageia
- Manjaro
- Puppy linux
- Steam OS
- Quirky linux (which I am using )
...

[edmon]

edmon

This is very interesting what you write, because Debian Testing/Unstable is/was in a huge libstdc++6 transition that affected lots of packages especially KDE ones. Maybe it did not affect servers without X. You should be really careful to use testing instead of a specific codename in the sources.list as a system can be broken easyly after a new release and ongoing transitions. Of course if you use sid instead of unstable you gain nothing. Basically the next stable codename is save to use in frozen state or in a nearly frozen state but not as general rule, this is just wrong.

darkbasic

Basically testing is a moving target, always pointing to the next stable release. Unstable points always to sid, so basically you can consider both as rolling releases - as testing mainly has just fewer packages and gets the unstable ones after a delay.

to dispel your worries, here is my desktop sources.list:


deb http://httpredir.debian.org/debian/ stretch main contrib non-free
deb-src http://httpredir.debian.org/debian/ stretch main contrib non-free

deb http://security.debian.org/ stretch/updates main contrib non-free
deb-src http://security.debian.org/ stretch/updates main contrib non-free

deb http://httpredir.debian.org/debian experimental main

## Depôt MultiSystem
deb http://liveusb.info/multisystem/depot all main


and i'm using only KDE.
on servers i'm mostly put testing as release...
i'm using dselect and when dselect want to remove more than 10 packages i'm postpone the update
for several days and after some time everything is calm and you can update without removing the mail server for example )


ps. there are many ppl who used debian this way and this is why it so popular despite his stable release has old packages

ps2 now there is new version of ifupdown which wants to remove systemd )) so i just put one = before it in dselect until it doesn't want
to remove half of the installed packages

[linuxforall]

For an LTS, UBUNTU 14.04.3 did just fine.

[liam]

So Debian wins and Debian loses, and some packages should look at their default compile flags, and consider adding runtime CPU checks.

If Debian is the only thing you're interested in then, yes, that's a reasonable summary.