[MUD-Dev] Re: DevMUD: Inheritable modules

[Chris Gray]

[Joachim Pileborg/The Arrow:]

 >I'm speaking about what modules (or rather, interfaces) a MUD must have to
 >call itself a DevMUD.  At this point I would like to explain my thoughts
 >about what I mean about 'DevMUD compliance'.

OK, now I think I understand. My next question is: why would anyone want
to create a new DevMUD core? My understanding of the current view is that
the core won't be very big and the sources will be publically available.

Now, I can see someone adding stuff to their own server so that it can
use DevMUD modules, so in that sense, compliance is useful.

 >For example:  If a module say "I need the Telnet interface!", the core looks
 >at its database and if the Telnet interface is found, it loads the module
 >containing the Telnet interface (if it isn't loaded already) and tells the
 >module asking for it "Here you go, one Telnet interface ready for use."

OK, understood. That assumes, however, that we want the core to do
automatic loading of needed modules. Another possibility is that no
module is initialized until all of its dependencies are satisfied (everything
it needs to call is there, and bound to it). Then, the core doesn't
have to do any automatic stuff. It is up to whatever triggers module
loads to get it done. This could be a config file, as has been suggested,
or could be explicit module-load commands coming from an on-line connection.

--
Don't design inefficiency in - it'll happen in the implementation. - me

Chris Gray     cg at ami-cg.GraySage.Edmonton.AB.CA