Op Friday 14 February 2003 11:11, schreef Patrick S. Vogt:
> Hi all,
>
> so we're basicly back to the old features, except for the systemtray,
> check(all|node) and the dcop interface.
cool, seems its going fast.
> My next step is to implement a KMonNodeList, this should allow to implement
> the missing features.
>
> Soon a KMonConfig should be implented, that reads in the config files and
> defines the config widget, I would prefer if each KMonNode subclass handles
> those parts it needs.
>
> One of the things I am not sure about is wether it is better pass the
> KMonNode to the KMonNotifier when created or wether one should pass a
> KMonNode through the stateChanged signal/slot.
well, you mean we could make a 1-1 kmonnotifier<->kmonnode (at creation time)
or one notifier for possibly many nodes (at signal/slot). well:
- you make 1-1 obligatory with the first solution, and with the second
solution, if we decide somewhere in the future 1-1 is better for some cases,
we can go for it without changing much.
- it doesn't look like we need strict 1-1...
so i'd opt for the second one, but i could be wrong..
also, i have to do a talk for my thesis at the end of the month, that means i
will not be too active on kmon for the near future. i hope to improve that
asap.
greetings,
frank
>
> Cheers
> Patrick
|
