Niclas Hedhman wrote:

On Tuesday 11 November 2003 07:35, Stephen McConnell wrote:



meta-info (description of a component type) meta-data (descriptions of deployment scenarios)




Is it only me, or isn't these too definitions too confusing??? I also think that it restricts future meta definitions to a large degree.

May I suggest a term shift;

meta-info -> Component Type Meta
meta-data -> Deployment Scenario Meta


You can try!


It will make it a lot easier for new people, and people like me (memory=640kB, frequent NMIs), to read descriptions.


I understand the "confusing" concern. I must confess that many years ago I was dragged kicking and screaming into this arena and yes - when you really get into scenarios ivolving multiple meta abstractions - things do get interesting and belive it or not - the complexity adds clarity. Today our meta-model does not have a meta-meta-model. If we do have a meta-meta-model - the notions of meta versus instance would be easier to understand. My guess is that we will end up in meta-meta here in Avalon within the next couple of years. When we make the jump it will be painful but once you make the jump it like riding a bike.


I.e. stock-up on Asprin!

;-)

In the meantime .... consider the following dimensions:

META-INFO             META-DATA
-------------------------------------
class                 instance

type                  criteria (particular context requirements,
                     dependency types,...)

profile               scenario (a particular context solution,
                     configuration, and assembly)


Now that wasn't so hard!


Cheers, Steve.

--

Stephen J. McConnell
mailto:[EMAIL PROTECTED]




--------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]



Reply via email to