At 16:30 Uhr +0200 11.08.2002, mathias meyer wrote: >dear finkers > >in the fink_0.9.12-1 days there was the discussion about a fink >cleanup command, which was then put into fink_0.9.12.cvs on the 9th >of june by max horn (see the corresponding mail to this list). i >made then a suggestion (archive: 10th june) on how i would like a >cleanup command to look like. >now fink_0.10.0-1 is officially released, so i was wondering again >how far fink cleanup was implemented. after doing a fink cleanup i >noted the following:
Note that I didn't document this command on purpose: I had no time to do proper testing of it yet :-) So use on your own risk (but of course I am interested in making this command safe, so i am grateful for your testing and feedback). > >ODDITIES: >1) some recent (installed) .debs (such as base-files-1.6-1 and a >whole lots of others were deleted) Hm, this shouldn't happen (and over here doesn't happen as far as I can tell). [...me tests on his system by replacing the unlink's with print's...] No definitly nothing like that happens here, only .debs and source files that are really obsolete are deleted here. >2) some old .debs (such as less-358-4, fink-0.9.12-1) are still on >the system even though a newer version (in this case less-376-1 and >fink-0.10.0-1) is installed The logic is this: we keep all source files and deb files for which an info file is present. This is done on purpose, since it sometimes might be on purpose that you have an older version of the .info file. In short: Better delete less than to much :-) > >POSSIBLE EXPLANATIONS: > >for 1): (a shot in the blue) all (as far as i checked) these >packages were in unstable. still this does not mean that all my >recent, installed unstable .debs were deleted. can't really find an >explanation here. Hm, you know this *might* be the explanation if you swapped the order in which stable and unstable are set in /sw/etc/fink.conf, or if you removed trees since hten or stuff like that. However if you didn't change anything there since installing the packages, it shouldn't affect this at all. "Shouldn't" as of course I can't gurantee the code is correct (well my former CS teacher would correct me there and mumble something about the fact that one can proof programs correct, but err.. let's not think about that to much, it's theoretical computer science :-) Cheers, Max -- ----------------------------------------------- Max Horn Software Developer email: <mailto:[EMAIL PROTECTED]> phone: (+49) 6151-494890 ------------------------------------------------------- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf _______________________________________________ Fink-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/fink-users
