Research drag and drop mode

[rsthornton]

Explore a mode of analysis that incorporates more flexibility for the analyst to capture knowledge without rigid adherence to the methodology.

User should be able to drag and drop icons to assemble SOI knowledgebase.

[Jtensminger]

@rsthornton The drag and drop action isn't super hard to design, but removing the methodological constraints (the rules dictating where you are able to drag-drop-connect entities) is a big technical change from the perspective of file saving and the sysXML format.

If you could share more thinking and elaboration on this issue, that would be helpful!

[rsthornton]

@rsthornton The drag and drop action isn't super hard to design, but removing the methodological constraints (the rules dictating where you are able to drag-drop-connect entities) is a big technical change from the perspective of file saving and the sysXML format.

If you could share more thinking and elaboration on this issue, that would be helpful!

The general thinking here can be broken down into two components.

1. We haven't yet directly incorporated one of the key aspects of SL which is "atomic work processes." (kindle page 290) I'm imagining that at some point down the line, there could be value in allowing for a more "bottom-up" approach to system analysis where the analyst builds up a system from these atomic components rather than decomposing it from the top down. They might prefer a drag-and-drop approach when doing this?

2. I'm also imagining that there could be future value in allowing users to play with some of the concepts of system language in an exploratory fashion while relaxing the methodological constraints of Deep Systems Analysis.

This is all pretty rough speculative thinking at this point and definitely isn't a short or medium term priority. But it is something that has come up in my early explorations.

Any general thoughts or follow up questions would be really helpful for me!

[Jtensminger]

@rsthornton I think the bottom approach might just be a bunch of chained atomic processes. At some point, you'd need a way to select which ones are apart of the same super system and shift up a level. It might be worth just drawing that out on paper to see what it looks like. The logic we've implemented for element types connect with each other would still be the same (I think).