Am 18.02.2014 um 18:52 schrieb Dagobert Michelsen <[email protected]>:
> Hi Oliver, > > Am 18.02.2014 um 18:34 schrieb Oliver Kiddle <[email protected]>: > >> Dagobert Michelsen wrote: >>> >>> If you want you can try one of these: >>> http://mirror.opencsw.org/opencsw/allpkgs/alternatives-1.3.30c%2cREV%3d2010.02.18-SunOS5.8-i386-CSW.pkg.gz >>> http://mirror.opencsw.org/opencsw/allpkgs/alternatives-1.3.30c%2cREV%3d2010.02.18-SunOS5.8-sparc-CSW.pkg.gz >>> >>> They should work correctly on the provided packages even when in one file. >> >> I can't get them to do anything. Perhaps they need some database or >> something setup but if all I do is install them, nothing happens when I >> try to configure alternatives. The contents of /var/lib/alternatives on >> my Redhat based system don't seem quite similar enough to what we have >> in /opt/csw/share/alternatives. > > Yes, you need to reinstall the packages using alternatives, like the > Tcl-package. > >> Note that the alternatives/trunk/files/README file in subversion says: >> Phil decided that a from-scratch, CSW-custom implementation was >> needed, because the debian one was hugely bloated, and the redhat >> smaller one, did not play nicely with NFS-shared /opt/csw So please >> dont go getting ideas that we can migrated back to redhat,etc in the >> future! :) we tried, and it failed. > > Well, IMHO it just didn’t cope with NFS-sharing /opt/csw which I would > consider > an ugly super-legacy method now that we have system automation. So that would > be > not a big deal for me. > >> Having GAR create multiple alternatives files would seem easiest to me. >> If you prefer, I can modify the existing script to copy relevant entries >> to /etc/opt/csw/alternatives instead of a symlink. I think it can cope >> with things like package removals and upgrades with a little extra >> checking. If you get the RedHat system working, I'm not opposed to it >> either: I just suspect it is more work that way. > > The only thing we would need is a converter from the datastore of the existing > symlink-data-keeping to the method the RedHat implementation uses as > postinstall. > That way we would keep alternatives happy. I consider this fairly important to > fix, but I don’t know when I come to it. If you have time it would be great > if you could research this. Alternatively while we are at it we could also change the data storage from the ksh implementation to use the format of the RedHat implementation if feasible. That way we would get both advantages. Best regards — Dago -- "You don't become great by trying to be great, you become great by wanting to do something, and then doing it so hard that you become great in the process." - xkcd #896
smime.p7s
Description: S/MIME cryptographic signature
