Colin Guthrie <gmane at colin.guthr.ie> writes: > 'Twas brillig, and Nix at 23/04/10 00:43 did gyre and gimble: >> Fedora, at least, doesn't use ck-launch-session: it uses >> ck-xinit-session, which is not in upstream console-kit at all; it's in >> the RH-specific xinit package. > > Can you say which pacakge provides that file? It doesn't seem to be part > of the main ConsoleKit spec (i.e. with a patch). > > I'd like to look at this file but without knowing what package it's in, > it's hard to comment. > >> It's derived from ck-launch-session but >> does some incomprehensible-to-non-dbus-hackers and uncommented thrashing >> about with dbus first. It too appears to run a shell and then exit, so >> how it does what it does is equally mysterious to me. Of course, it, >> also, has no documentation whatsoever. > > Maybe it's not ready for upstream yet? Fedora tends to be a testing > ground for a lot of new things, so this is not overly uncommon. > >> I love the new Linux world. :/ > > Would you prefer some completely unknown and mysterious system to one > you actually can poke about with and figure out? Honestly, if this > bothers you, do something about it - speak to the people involved and > help write the docs. Moaning solve precisely nothing. I don't think that's what he meant. There are upsides and downsides to how Linux works and is developed. I have felt the same frustration about these kinds of new subsystems that lack documentation. Most recently it was udev that has no description about its properties, not even in the source code. Of course we have to moan, otherwise will never document anything they do. And asking people who know absolutely nothing about something quite complex and obscure to write the documentation soves precisely nothing. -- David K?gedal
