On Fri, Aug 07, 2009 at 04:03:32PM -0400, Tim Abbott wrote: > On Fri, 7 Aug 2009, Matthew Wilcox wrote: > > > On Fri, Aug 07, 2009 at 01:48:34PM -0400, Tim Abbott wrote: > > > I've attached a patch to split the package into a kerneloops package and > > > a > > > kerneloops-applet package. The new kerneloops package has just the > > > daemon, and works as though "allow-submit = ask" were set to "no". There > > > may be some migration cost of people who upgrade and lose the GUI, but I > > > think that this is a substantially better design than having two packages > > > that both contain indentical copies of the kerneloops daemon, init > > > script, > > How about we do this the opposite way round, where we have > > a kerneloops-daemon package, and the kerneloops package depends > > on it? That way, people who upgrade from current kerneloops lose no > > functionality, and new installs who want to get rid of the applet can > > just install kerneloops-daemon? > > Hmm. I think long-term it would be best to have the names be > kerneloops/kerneloops-applet. kerneloops-daemon/kerneloops is certainly > much better than kerneloops-nogui/kerneloops, but I think is still a bit > confusing. It may very well be justified by saving the current userbase > from having things change out from under them.
We can set that as our goal and transition to it. So we should split the package into kerneloops-applet and kerneloops-daemon. Then create a new kerneloops dummy package which depends on kerneloops-applet. In a couple of years, we can rename kerneloops-daemon to kerneloops, and everybody should be happy. Computer Science Motto: All problems can be solved with an extra layer of abstraction ;-) -- Matthew Wilcox Intel Open Source Technology Centre "Bill, look, we understand that you're interested in selling us this operating system, but compare it to ours. We can't possibly take such a retrograde step." -- To UNSUBSCRIBE, email to [email protected] with a subject of "unsubscribe". Trouble? Contact [email protected]
