On Wednesday 15 January 2003 14:22, Matthew Saltzman wrote: > > Obviously, it doesn't want to blow my current configuration away > > but this is pretty non-informative! Is there any way to determine > > which configuration files will get blown away if I choose to > > install this thing? > > Do you remember which ones you modified? rpm --verify will tell you > about any files that are not identical to the ones distributed with > the installed RPM.
Heh, over the years I've learned Mike's rule: "Do not to trust your memory for anything that took place more than a month ago." Whenever I violate this rule I find that I spend more time troubleshooting this rule violation that the original problem! Thanks, the --verify will work. I looked right past that option as I was looking for something that would install the package without destroying the configuration files. > > Anyone have an idea why this package doesn't work like so many > > others and just identify new configuration file with the *.rpmnew > > naming convention and allow me to spend a few quiet hours trying to > > figure out what to merge between old and new? > BTW, the skip list policies are outlined in the man page for > up2date. Hmmm, what I was looking for was a description of what logic up2date uses to decide when a choice has to be made between the options: 1.) skip because a local configuration file has been changed. or 2.) install but rename the new configuration files *.rpmnew There must be some mechanism that determines which of these two paths up2date chooses since I've now seen both behaviors with the same up2date configuration. Regards, Mike Klinke -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED]?subject=unsubscribe https://listman.redhat.com/mailman/listinfo/redhat-list
