Scott Robbins wrote:
I think that repeatability with the original machine is useful. That might involve uninstalling a package, removing config customizations, and then reinstall, test, customize config, test. This helps to discern whether it was a one-time issue, or related to the specific config performed.Now, for me to report this bug, and for it to be worthwhile, I'd have to install the alpha on a second machine and see if I can reproduce it. This was an install off of the live CD. For me to feel justified reporting it, I'd also want to install off the DVD and possibly a fewtest installs of other live CDs, e.g., KDE and the like.
Before creating a bug entry, check that updating to current updates hasn't already fixed the problem.
One way to do the A/B machine comparison is to run the test instance in a virtual machine. {eg vmware-server - watch the installer misconfiguring the selinux on /etc/services - and others}.
More reason to actually create the bug. It is thoroughly impossible to fix a bug that no one has reported ;-) Even if it gets marked cant/wont fix - at least further people searching for that issue can see the status, usually along with a description of why it was marked that way - and hence not create another bug.So, I don't report this bug, I don't want to waste the time of the developers, especially when I see that no one else has mentioned it
The Fedora people always say: stick it in bugzilla. Make sure you search for it before adding a duplicate.Nor do I post about it here, as this is a busy list, and again, I've seen no one else mention it. (I'm simply posting about it now as an example of a bug that I've experienced, and not reported.)
From the alpha/beta/rc changes, I get the impression that alpha is 'raw'er than test1 used to be.
DaveT.