On Wed, 2015-08-05 at 05:42 +0200, Mike Galbraith wrote:
Over the years, I've see a few updates release -> release+1 go south and require a rebuild, but when you're doing a major overhaul, you take precautions and budget time for it. TW turned the breakage risk into a frequent event, and after being happy as a clam with TW for quite a while, complete destruction happened at a bad time, so out the window TW flew. For me, TW proved itself to be too unstable for daily bread use, and not on one box, on four of them. I still use it on my test boxen, but never again will it meet my primary workstation.
And of course you filed bugs, submitted fixes and new openQA test
scripts to make sure the breakage doesn't happen anymore?
Why is everybody expecting 0 breakage at 0 work investment?
I myself have been using TW on all my machines without issues - but
granted, I also submit openQA tests when I see stuff breaking, making
sure we see the breakage in the future before it happens.
And, no plasma 5 does not count as breakage: the new version simply has
a new feature set. The old version is no longer maintained. And I think
the KDE Team did an awesome job in communicating the change well ahead
of time... and they also did not jump the KDE-4 ship prematurely.
I understand a broken system is something bad and I can only assure you
that, together with openQA, we try to do the best possible to avoid any
such thing... but, it REALLY takes the time and efforts of everybody,
to at least report failures in a proper bug.
Cheers, and who knows, maybe you'll switch back to TW - once we can
assure the things that broke for you are being tested upfront :)
--
Dimstar / Dominique Leuenberger