[neon] KDE Neon - Why Ubuntu

[Harald Sitter]

Hola,

Since Jens yesterday raised the question of why neon will use ubuntu
as a base I threw together a white paper on the topic.
I think right now the important part to keep in mind is that until we
have a greenlight from the KDE e.V. Board to pursue an Ubuntu base
from a legal POV, we could (and might have to) change the base we are
going to work on. So everyone should keep an open mind when talking to
others about the base we will or should use. I for one do not care as
long as we can build something great on top of it, and right now the
best I know happens to be Ubuntu.

# Intro
KDE Neon provides an easy and elegant way for people to test the
latest from KDE, or use the latest releases of KDE Software.

Since KDE Neon is meant to put focus on KDE Software, the base system
requirements are very trivial:

- Linux kernel
- GNU toolchain
- Actual KDE Software dependencies
- Interesting 3rd party software (firefox, vlc, inkscape...)

Just about every distribution in the world matches these requirements,
so for all intents and purposes all of them are suitable.

In particular, since we only care about KDE Software, any religious
concerns a la "my installer is 2 seconds faster than yours" should not
play into this.

# Why Ubuntu?

Since we care little to not at all about the superfluous terminal
software stack we can look at this from an entirely KDE purpose driven
point of view.

## What We Want

- A stable foundation that rotates no more than once a year
- (Security) Updates for the foundations and everything we do not
handle within KDE Neon.
- A healthy ecosystem

Most of the mainstreamish/large distributions meet these requirements
(sans Arch as that is rolling).

## Why we want a stable foundation

- Users like things to not explode
- Rolling foundations *always* have a risk of random explosion so we
do not want rolling foundations
- What we want are rolling KDE software, the rest we care little about
- This does have the disadvantage of not being able to use the latest
and greatest foundations (e.g. systemd or pulseaudio) unless we
decidedly make it so which might be substantial effort.
- By not having to QA the foundations over and over again, we can
focus on providing high quality experience of KDE software instead of
chasing our tail by regularly having to do deliberate fullstack QA

## What Ubuntu Gives us Additionally

- Ubuntu has so-called Long Term Support (LTS) releases that only
happen once every two years. LTS releases do get first class
(security) updates for 6 years, making it appealing to enterprise
users.
- LTS releases can also get opt-in backports of foundation stacks
(e.g. X11). So if need be we can actually selectively opt to use some
newer foundations as they become available.
- Canonical does offer hardware certification programs for LTS
releases making it easier to find semi-reasonable devices to run on.
- Ubuntu is pretty much the only distribution that holds any sort of
sway with Desktop OEMs such that it gets preloaded and supported.
- Unlike most other contenders Ubuntu is itself based on another
viable option (Debian), so if Ubuntu as a platform doesn't turn out
too well we do have a way out by rebasing to Debian without risking
too much disruption for the user, nor having too reinvent all our
technology.

# Conclusion

By and large any base distribution that isn't rolling could do. It is
however my current believe that Ubuntu offers the most interesting
platform to build on; in no small part due to Canonical's efforts of
making things appealing to enterprise.

http://futurist.se/gldt/wp-content/uploads/11.03/gldt1103.png