On 27 September 2015 at 22:17, Michael Ströder <michael(a)stroeder.com> wrote:
Wolfgang Rosenauer wrote:
Am 27.09.2015 um 21:13 schrieb Michael Ströder:
Martin Pluskal wrote:
In this particular case:
If 3.0.3 contains bugs making certain deployments impossible and upstream
developers answer "use 3.0.9" it's pretty stupid to ship 3.0.3 in a
release now and start back-porting all the fixes to that package for years.
Personally I can live with custom packages for my own deployment and will
simply stop contributing then.
I don't get your argument. Even if we ship 3.0.9 now in a few months
upstream will say "it's pretty stupid to use 3.0.9".
We can only see the past but cannot predict the future.
IMO it's waste of resources (for upstream and distribution developers and
deployers) to ship an older release at a certain point of time while a newer
release already has the fixes known at that time.
IMO packagers should try to keep the amount of patches at a minimum. Otherwise
work piles up and it gets more and more tedious to make the next upstream update.
You sound very much like the typical Tumbleweed user and contributor
And that's fine, please, continue your contributions to Tumbleweed then
But the reality is that the 'constantly moving target' is _not_ what
_some_ of our _users_ want
A distribution that changes all the time requires an element of
regular adaptation, learning, and (unless they have a very limited use
case) alteration in order to continue being productive with whatever
implications of the latest version of $Package brings with it
And for Tumbleweeders that's fine
And for people who want to use, and maintain Leap, the mindset is different
It involves picking a stable, sensible version of $Package,
maintaining it with care, attention, and backports, for at least a
year, then revisiting whether or not upgrading to the latest and
greatest (or something close to it) makes sense a year from now in the
next Leap minor release
(In the case of the next Leap major release, there's much less of a
choice involved, the assumption is that it will start based on a
Tumbleweed snapshot in a few years from now)
So Michael, please don't be dismissive, and use terms like 'stupid',
to describe other people who have an opinion different from your own
With FreeRadius, if we end up with the SLE 3.0.3 version in it, there
will be someone, from SUSE, maintaining the package. That means
writing, backporting, and providing, patches for issues. If people
have issues to raise, they can raise them in *our* openSUSE Bugzilla.
To unsubscribe, e-mail: opensuse-factory+unsubscribe(a)opensuse.org
To contact the owner, e-mail: opensuse-factory+owner(a)opensuse.org