On Fri, Feb 20, 2009 at 11:52:38 -0800, Adam Williamson awilliam@redhat.com wrote:
Sure. Stuff breaks in development branches. This is why all instructions relating to Rawhide specifically state that you should install it on a test system, test partition, or test virtual machine. So that when it breaks you can shrug and go 'oh, well' rather than screaming and going 'NOOOOoooo, my precious work!'
I still curse, but usually because it happens when I wanted to do something and stupidly did an update just before then. I haven't lost data because of using rawhide in a long time, and then happened when I accidentally installed to the wrong partition when I was doing a lot of rawhide installs to help track down a driver problem.
We're not really talking about 'normal users' running Rawhide, anyway. We're talking about people who are interested in helping improve release quality and testing bleeding-edge code, who can at least work a text editor from a console, and have the flexibility to understand that Rawhide stuff *will* break occasionally and they'll have to work around it.
I think people need to be able to significantly more than use a text editor. You need to be able to rescue your system when booting fails. I think you pretty much need to be an amateur sysadm.
The point is that this pool of people is in fact far larger than the number of people who currently run Rawhide. It should at least include the vast majority of packagers, yet from what I've seen, it seems that a lot of Fedora packagers only run stable releases, which is a pretty reliable indicator that we really could have more people running Rawhide.
I think going for the vast majority of Packagers might be overly optimistic. Besides needing sysadm skills, you also need good bandwith.