Plan for tomorrow's FESCo meeting (2010-11-17)

Kevin Kofler kevin.kofler at chello.at
Sun Nov 21 22:04:38 UTC 2010


I wrote:

> Adam Williamson wrote:
>> How do you expect to be able to maintain an entire desktop environment
>> on a distribution you don't even have installed? I have some sympathy
>> for the 'fifty people said it works on F14, it probably works on F12
>> too' argument, but for a *small, leaf* package, not for an entire
>> desktop environment! If I were a KDE user running F12 I'd feel very
>> unsafe knowing someone was happily pushing updates of the entire
>> environment who did not even have a running F12 machine.
> 
> I've sometimes actually done testing on older releases out of sheer
> laziness to upgrade to a newer one (see also me testing that F13 KDE 4.5.3
> upgrade), but with all this bullying of "Want current software? Upgrade
> your Fedora!", with previous supported releases getting only second-class
> upgrade support, that's going to stop soon (in fact, I'll probably upgrade
> my machines to F14 before the end of the month). (Pretty much everyone
> else in KDE SIG always runs the latest Fedora. I'm almost the only one
> left on F13.) So by limiting the kind of support previous releases get,
> you're actually INCREASING the risk of untested updates, by making it
> unattractive for your developers to run those releases.

PS: In addition, more aggressive package upgrading on stable releases would 
also reduce the amount of gratuitous differences between the releases, 
making it less likely for stuff to work on Fedora n and break on n-1.

In short: Want higher-quality updates for previous releases? Then push 
version upgrades wherever possible (even and especially for libraries, as 
long as they're ABI-compatible or can be group-pushed with a small set of 
rebuilt reverse dependencies)!

        Kevin Kofler



More information about the devel mailing list