>>> On Monday, 11. June 2007 at 6:26 pm, Dan Cox<[EMAIL PROTECTED]> wrote: > A few peculiarities with the OpenPKG run control system are causing > some end user confusion when working with services. (Using > openpkg-20070608) [...] Thoughts/comments? > Thanks a lot for the detailed report, Dan. All of what you reported describes correctly what "openpkg rc" is doing today.
I traced down that much of the difference between your expectations and actual behavior seems to come from a different definition of "disabled". For "openpkg rc", disabled means "not do the action". It seems to me your definition can be described as "not run the daemon". So we end up with OpenPKG "disabled": "openpkg rc start foo" -> not do the start "openpkg rc status foo" -> not do the status "openpkg rc stop foo" -> not do the stop Dan "disabled": "openpkg rc start foo" -> not run the daemon, refuse start "openpkg rc status foo" -> not run the daemon, display status anyway "openpkg rc stop foo" -> not run the daemon, stop if it running anyway The stop anyway case happens likely if disabling the daemon after start. It seems to me both approaches are valid. The "Dan approach" is just not implemented. Worse for OpenPKG, the "Dan approach" seems to be natural for humans. I have to investigate whether it is possible to implement your approach, too. We need to find a good name for both in order not to confuse users too much, considering rc_all and rc_def, too. Maybe renaming the existing functionality to "dormant" does the job. -- -- http://thomas.lotterer.net ______________________________________________________________________ OpenPKG http://openpkg.org User Communication List openpkg-users@openpkg.org
