On Feb 13, 2011, at 9:15 PM, Jeff Johnson wrote: >> >> Log: >> set mi->mi_count when rpmmiNext() initializes db cursor so that >> rpmmiCount() won't break (mdvbz#62267) >>
Note that using %trigger* in order to re-create symlinks when a package is upgraded is just wrong as well. Yes you can do that, Mandriva may very well have been doing this for years, but that isn't what I'd call "correct". I'd call it a design flaw in RPM that has been copied into packaging again and again. (aside from 30 seconds of thought) It would not be impossibly hard to give bash-completions its own sqlite3 database of symlinks using embedded sqlite. Pushing data to applications, not overloading and abusing the package state machine, is increasingly necessary as applications get more sophisticated. Yes the trigger value needs to stay exactly as always (and also true for what Funda is attempting with FConf schema, arrived in private e-mail yesterday). 73 de Jeff
smime.p7s
Description: S/MIME cryptographic signature
