On Tue, 03 Apr 2001, Jean-Philippe Guérard wrote: > ie you install a init script with update-rc.d, but you have no way > of knowing which script is installed, and which runlevel it is > installed for. You have two classes of code that needs to interact with the initscripts: those who should not even try to understand what is behind invoke-rc.d (anything that runs initscripts), and those who need full, complete and correct knowledge and read/write access to the initscript subsystem (configurators, the initscript system itself, invoke-rc.d itself...) > Programs that need this information (e.g. webmin-core, ksysv, rcconf, Will have to depend on the sysvrc package (does not exist yet), so as to be removed from the system if one uses another rc scheme (such as file-rc). > check-rc.d should be a very simple script, and provide > efficiently all the necessary information. > > Does that seem reasonnable ? Well, not really. Those programs need R/W access to the configuration, so they should deal directly with the symlinks and init scripts. And they also will have to be forbidden to screw the system up for someone who uses another rc scheme. Unless you're going to propose interfaces to do all the reading and writing of the configuration (update-rc.d is NOT flexible enough in that regard, I think), I see little point for check-rc.d. -- "One disk to rule them all, One disk to find them. One disk to bring them all and in the darkness grind them. In the Land of Redmond where the shadows lie." -- The Silicon Valley Tarot Henrique Holschuh
Attachment:
pgpfOgiNK_ctV.pgp
Description: PGP signature