> As others have pointed out. Anaconda should remove whatever it needs > to remove in order to install what it needs to install. > > This is not an impossible or even particularly hard problem to > solve. It is merely a problem that no one has taken the time to > solve, with the result that RPM will fail to install an RPM if > there's something "unexpected" where the RPM files / directories / > links / devices are supposed to go. So add a special install mode to > RPM which says, "you're allowed to blow away anything you find of > the wrong type if you encounter a blocking mismatch during > installation," and have anaconda enable that mode. > > Anaconda should be allowed to create any config files that it's > designed to create and install any files that it's designed to > install. That doesn't preclude having a preformatted partition that > has a bunch of files and directories on it that Anaconda doesn't and > shouldn't need to care about or touch.
The amount of work you're describing here is huge, and the number of people who would benefit from such a setup is very small. I'd guess that for whatever scenario you can imagine, another scenario can be imagined that would not be able to be handled. That's what I mean by saying the stream of bugs would never end. I don't want to start haggling over details of example after example, but just to give you one example to make this a more concrete discussion. Let's say you do an x86-64 installation. You then later go and do an i386 installation reusing the / from before. You now have two sets of the libraries laying around, for different architectures. What happens? Or, you do an F14 install that's got upstart installed. You then go and install over top of it and do an F16 install and get systemd. Which one wins and boots the system? In both of these scenarios, it's not that there's some config file confusing anaconda. It's files owned by RPMs that would not be overwritten by installing something else, and those files will cause problems. How do you even determine what's "unexpected" or "of the wrong type"? > >Well for kickstart users who want to do this intentionally, they can > >still continue to do so (though, be careful). > Someone else said in another message that this ability was going to > be removed from kickstart as well. I said it, then I went and looked at the patches and said that I was incorrect. - Chris -- test mailing list test@lists.fedoraproject.org To unsubscribe: https://admin.fedoraproject.org/mailman/listinfo/test
