![]() ![]()
|
|
![]() ![]()
|
Well, fortunately I can retract part of my critique regarding the postinst script.
This was perfectly fine and correct: ln -sf neon /etc/dpkg/origins/default But, nonetheless: $ dpkg-vendor --query vendor Ubuntu still doesn't seem the right/intended/expected result. Therefore, I think the rest remains problematic. So, what is the right solution? Is 'Ubuntu' right, even if the 'neon'-variant is configured-in but not used (intentionally/erroneously)? |
![]() Registered Member ![]()
|
Probably even if the vendor remains ubuntu, the neon package (from the PPA) should override it because of the package version numbering being "higher" with all bits appended to it. Just that the version happened to not override (compared higher) when 16.04.2 official package version came in?
"Thou shalt not follow the null pointer for at its end madness and chaos lie."
|
![]() Registered Member ![]()
|
If you see the ubuntu package versions, they were "9.4ubuntu4.3" then "9.4ubuntu4.4" but neon was "9.4ubuntu4.3+p16.04+git20170112.1812", now it is "9.4ubuntu4.4+p16.04+git20170210.1607"
So, in case ubuntu again bumps to "9.4ubuntu4.5" it will again mess with neon unless the devs bump the neon package version beforehand ![]() Probably neon should have better package version - not sure how the comparison to decide which is higher is done - maybe a string compare?
"Thou shalt not follow the null pointer for at its end madness and chaos lie."
|
![]() ![]()
|
I, too, am concerned about the version numbering scheme in KDE Neon.
Another possible can of worms may be version-pinning in the '/etc/apt/preferences.d' files. For example:
I really don't know: Is it a good idea of KDE Neon to call their 'Suite' and 'Codename' the same as Canonical/Ubuntu call their's? |
Registered users: abc72656, Bing [Bot], daret, Google [Bot], lockheed, Sogou [Bot], Yahoo [Bot]