21.11.2023 Area A Data Modeling

[hampusnasstrom]

Agenda

We will discuss the System base section.

Meeting Notes

Attendees: @hampusnasstrom, @JFRudzinski, @mkuehbach, @budschi, @JosePizarro3, @aalbino2, @ka-sarthak, @ndaelman-hu

• We discussed the need for a best-practice on plugin writing
  • Markus Scheidgen wrote some docs on resuable code
  • We will deal with this in next weeks meeting
• Chema introduces the current state of the System remodelling
  • Google slides
  • We should work on an example basis, trying to implement them and see what needs to change
• Discussion on the System section
  • Currently we have all parameters in the base section but this should be specialized later
  • Chema made the point that over-nesting can cause problems when trying to generalize
  • Sebastian raised the point that very specialized sections (inheriting in the end from System) should live in plugins
  • We clarified that the System in the google slides actually refers to a subtype of System
  • Nesting System helps with time evolution where we can assign a time step
  • How do we deal with specifying that certain Systems should be present in a specific use case

Tasks

• Area A (@hampusnasstrom) will add a system description for the solar cell example and some additional ones
• Review the examples within 3 weeks (until 11.12.2023)

[JFRudzinski]

[JFRudzinski]

I just wanted to add a comment that I thought about during the meeting but didn't want to cloud the discussion:

I REALLY think we need to stop using the term "topology", so that it does not continue to leak into our schema names and such accidentally. This is my fault because it's the common name for the system hierarchy in my community, but I think more generally it is a bad term since it is not obvious what it means and is even misleading depending on your background. I will put forth "System groups" as an initial suggest, but I am certainly open to alternatives.

[JosePizarro3]

Grouping sounds good, and it is indeed what we should be using to refer the System nested in itself. It is just that "topology tree" works good; but, how would you call this tree otherwise? "System grouping" directly?

[mkuehbach]

On the naming of the symbol "topology" aka "grouping":
PLACEHOLDER_set is what I typically use, i.e.
triangle_set, hexahedron_set, volume_element_set, system_set

On the use of the symbol "topology":
Indeed I agree it should be used only if one really talks about a concept that talks about the mathematical term and graph that is behind the topology of something

Why I will from now use the established terminology from the respective semantic web technology fields of research:
because the least what nomad needs is yet more jargon which does nothing but reimplement the wheel