So according to the https://www.philosophybasics.com/movements_pragmatism.html school, a thing is true if it works. And it worked in the previous version.
Once upon a time there was a very accurate heliocentric astronomical prediction model that used the concepts of nested spheres and epicycles to predict the movement of celestial bodies. For centuries it was “true”, and it took centuries of observations to find it inaccurate.
Because something was “true”, it doesn’t have to be “true” now. Likewise because you could do it in v13 doesn’t mean it was correct, “true”. If we adopt that mindset we will never have progress, in general, and technological progress, in particular; we will still be stuck with punch cards and mechanical computers.
It's clear that ArchiMate - as it so often does - uses the common English meaning; instance (basically a thing) not the computer science meaning - a thing that can be created through instantiation. One should never forget that ArchiMate is created by a committee of marketers who still fondly remember floppies and mainframes.
My argument is that if ArchiMate uses the common meaning of
instance, it is inappropriate to use the computing science meaning of
instance. Using the computing science meaning just confuses matters.
Also the example of switches, do they have the same set of Technology Functions. Well no if only one of them has PoE. Therefor they are not the same node/device. Nor are they realized from the same node/device. Might they both serve they same Technology Service, yes they might.
Does any of the discussion about switches using the definite or indefinite article belong in ArchiMate, No! Could we model this as a [computer science] ontology using a UML class diagram, why yes we can.
Considering that not all switches have the same technology functions, a
proper ontology model using a UML class diagram may be required depending, of course, on the modelling problem at hand - if all switches in the modelling scope are POE, this may not be required.
