Description of Problem:

I have an open bug from Limbo #1 where anaconda-generated XF86Config did not
work, but XConfigurator-generated XF86Config worked, so I wanted to check if
this bug still existed in Limbo #2.

After installing Limbo #2, and logging in, X did not start.

And running redhat-config-xfree86 now results in a rather uninformative python
traceback.  It appears that redhat-config-xfree86 tries to run X using current
settings, and when it fails, it just dies.

It took me some time to figure out that I had to run redhat-config-xfree86 to be
default.

I think that redhat-config-xfree86 should automatically fall back to --reconfig
if it detects that its first attempt to start X dies.  After all, the most
common situation where one would supposedly run redhat-config-xfree86 is to try
to fix a non-working X configuration, and requiring a working X configuration
for the default startup mode seems counter-productive.

Version-Release number of selected component (if applicable):

redhat-config-xfree86-0.5.1-1


How Reproducible:

Always.

Steps to Reproduce:
1. Start with a borked XF86Config
2. Run redhat-config-xfree86

Actual Results:

Observe the python traceback when X fails to start.

Expected Results:

Either fall back to --reconfig automatically, or at least prompt me: do you want
to blow away your current config, and begin from scratch?

Additional Information: