[cap-talk] a relationship is not an object

[David Hopwood]

David Hopwood wrote:
> John Carlson wrote:

>> [...] But the relationship between the part and the assembly doesn't
>> really exist as a separate thing that you can grab ahold of.
> 
> In many (most?) 3D modelling programs, it absolutely does exist as
> such -- both in the database schema and its implementation, and as
> a user-manipulable object in the UI.
> 
> In a feature-based modeller such as SolidWorks, for example, these
> relationships are called "mates", and there is a subwindow showing
> all of the mates in the current assembly, and allowing their
> properties to be independently edited. Incidentally, in this kind
> of modeller, mates *do* have identity (you can overconstrain the
> assembly by creating two identical mates, then delete one of them,
> and it will still be constrained). A user can't possibly make
> effective use of a feature-based modeller without thinking of
> mates as objects or "separate things".

I should also have mentioned that feature-based modellers in addition
treat *functions from parts to parts* as objects. Perhaps the most
remarkable thing about this is that users with no mathematical
background seem to be able to grok this model quite easily, with
practice (although, based on observation of training courses, a
small proportion of users find this aspect very difficult).

-- 
David Hopwood <david.hopwood at industrial-designers.co.uk>