suppose i have a relocatable package foo-1.i386.rpm i can install it to /opt/foo1 with: rpm -ivh --prefix=/opt/foo1 package foo-1.i386.rpm [providing i'm not using rpm 4.1, with the broken --prefix option, that is] or i can install it to /opt/foo2 with: rpm -ivh --prefix=/opt/foo2 package foo-1.i386.rpm but what i really want to be able to do is to install it to *both* /opt/foo1 and /opt/foo2. (and i'm creating the package, so i can ensure that there are no conflicts in doing so.) but rpm would fail on the second install, complaining: package foo-1 is already installed i could work around this with: rpm -ivh --replacepkgs --prefix=/opt/foo2 package foo-1.i386.rpm but that's not a good idea, since the rpm database now just knows about the copy in /opt/foo2 (if i were to query with 'rpm -qs', for example). and if i were to erase the package: rpm -e foo it would only erase the copy in /opt/foo2, not the copy in /opt/foo1. is there any solution to this? does rpm have a means by which it can be aware of two (non-conflicting) installations of the same package (possibly different versions, which is the reason i'm exploring this) on the same machine? i'm guessing the answer is probably no. perhaps through the use of --dbpath? but it seems like a bit of a hack to have to create an entirely separate rpm database just for each duplicate package. i believe that debian provides more control than the simple --prefix option of rpm with dpkg --root=dir | --admindir=dir | --instdir=dir. i am looking for what i believe is the equivalent of that kind of control in rpm, which i'm assuming doesn't exist. but i'd be very happy to be told otherwise. - rich __________________________________ Do you Yahoo!? Yahoo! Mail - Find what you need with new enhanced search. http://info.mail.yahoo.com/mail_250 _______________________________________________ Rpm-list mailing list Rpm-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/rpm-list
