Closed sorainsm closed 4 years ago
@sorainsm, I like your proposed classifications. This seems like a good idea. I'll think about adding it to the template for the future. :smile: I think all of the assumptions should have an associated binding time, although for a family of families, the binding time could be variable. That is the binding time itself could be a variability that distinguishes families. Maybe your unclassified assumptions could be in the category of "build time or run time"?
Your run time bindings are really just inputs. I think rather than list them as assumptions, I would suggest listing them as inputs.
I agree that the "unclassified assumptions" should be elsewhere. In re-reading I ended up moving them around (A1 to build time as it was a build time decision to use Cartesian coordinates rather than other methods; A2 and A3 moved to input).
I've removed the previous run-time bindings, and instead have captured this information as input variabilities (see commit 22edeee19c834a57847365d81f8560206fa81bbb). I'm closing this issue.
Is this division of assumptions (by binding time, and then general assumptions) appropriate for this work?
Do general assumptions even exist, or should they all exist at a specific binding time (e.g. coordinate system is a scope decision because we're not considering others).
NOTE: These sections are still a work in progress.