Quoting Matthew Palmer (mpalmer@debian.org): > I'm rather hesitant to introduce this sort of tight coupling between > packages; it makes things brittle and prone to breakage. On the other hand, > I know how hard it is to be able to detect network state, and changes to > what netcfg does in the future is likely to require changes in that code, > too. > > How about a compromise position, of having netcfg write out some sort of > state file containing machine-readable details of what it did, which > clock-setup (and anything else that decides it needs to deal with that sort > of thing) can use to make their decisions? That seems interesting. I'm CC'ing Joey, as everything that has to do with D-I general design needs to get his advice.
Attachment:
signature.asc
Description: Digital signature