On 4 Jun 2003, seth vidal wrote: > > > b) If yum STILL fails the check for disk space, and has several > > packages to install that are themselves quite large (and are in fact one > > reason that the disk space check still fails, allowing for the > > possibility that /var is or is not in the same partition that the > > install is tight on space in) it would be lovely if yum would try to > > discern a pattern for installation that would still permit the install > > to proceed. This is a more difficult request, but two immediate > > approaches, each embodying a certain degree of AI and arcanity, suggest > > themselves: > > This has been requested before and it gives me the heebie-jeebies. > > There is A LOT of room to cause death and pain to a user if this were > done. > > I can say that I'd never want either of the suggestions you gave to be > done automatically. yeah, sigh. That's what I had to do by hand, eventually -- all the internationalization support in OO is too damn big. OO needed over 300 MB free to install and I had only 240 (with the rpm's still on the install partition and all caches clean and I even removed festival and the old kernel). What I really am going to have to do is reinstall the laptop from scratch, repartitioning root a bit bigger and maybe splitting off /var. There is pain this way, too. At least I know what to do and how to do it, but this will be a "challenge" for user-level yumsters regardless. Still, I thought you'd say that. It is a heebie-jeebie-causing idea, after all;-) I just thought a) might slip past your guard, as it is ALMOST sane. Simple, at any rate, even by hand, if you know enough to do mv /var/cache/yum/base/packages/openoffice* /home rpm -Uvh /home/openoffice* rm /home/openoffice* when /home has GB of free space where it isn't really useful and / is starved. Of course encapsulating mkdir /home/.yumtmp (deal with failure or preexistence) mv /var/cache/yum/base/packages/openoffice* /home/.yumtmp (dwf) rpm -Uvh /home/.yumtmp/openoffice* (dwf) rm -rf /home/.yumtmp (dwf) is a bit safer (I had thought maybe perhaps safe enough)...but you still have cleanup issues if yum dies in mid stroke or you have a user named .yumtmp or the like. Rats! ...:-) rgb -- Robert G. Brown http://www.phy.duke.edu/~rgb/ Duke University Dept. of Physics, Box 90305 Durham, N.C. 27708-0305 Phone: 1-919-660-2567 Fax: 919-660-2525 email:rgb@xxxxxxxxxxxx