Karen Tung wrote: > Stephen Hahn wrote: >> * Karen W. Tung <Karen.Tung at sun.com> [2007-09-25 23:56]: >> >>> I just posted a very preliminary draft of the architecture diagram for the >>> Distribution Constructor. It shows the different components, and how they >>> interact with each other. As this is just a first draft, everything is >>> likely to change as more research is done on the different components. >>> >>> You can see the document in the link below if interested. >>> >>> http://opensolaris.org/os/project/caiman/Constructor/design_notes/ >>> >> I think that the GUI is likely to build the profile, and then invoke >> the coordinator/constructor to realize the profile. It's tough to >> communicate this distinction with just arrows, but the profile is a >> static aspect and not an executing aspect like the other parts of the >> diagram. >> >> - Stephen >> >> > Hi Stephen, > > Yes, the GUI is the user's tool to build the profile. I guess the > coordinator > is kinda like the backend of the GUI, where all the callback function > for the GUI will > reside. Not sure whether it will eventually be like that, but that's > the current thought.
'a profile' is listed as one of the three ways users provide input to the engine, so then is the profile something a user can fully hand craft (if the syntax were known), and bypass the GUI or Cmd? If so, then what's the initiating process that starts the engine in this case? Can a full profile get fed to the GUI and Cmd? Are the various 'Post processing' methods modifiable/appendable, or are they just selectable predefined sets of tasks? -ethan > > --Karen > _______________________________________________ > caiman-discuss mailing list > caiman-discuss at opensolaris.org > http://mail.opensolaris.org/mailman/listinfo/caiman-discuss
