2019-10-12 | Journal | Tree Script Done
I finished the script that generates all DFD levels from a tree. One of the interesting things about this project, is that the structure of the data falls out of the simplicity of the code. I may think I have the simplest tree structure, but the process of automating the parsing and generation of the diagrams shows where it might be simpler.
Now I need to take the DFD model and use a similar tree mechanism to get component, website, workflow, and checklist trees/scripts and then merge to get a single tree schema.
Currently the tree is simple enough that it can be updated fairly easy from a filesystem; however, dusting off my Forest Ranger program makes sense. I am still having trouble with the wx stuff and Python 3.7. Likely that needs to rest for a bit before it works well.
I have verified that the tree structure works well with git. That was another way that I simplified, as some things wouldn't work well with git. I have a Raspberry Pi in the attic crawl space that is running my git server. Once I get all of the pieces together big, first priority and working in full web presentation w/ solution description and run book sites, then I can start working on some of the funner stuff, like setting up MQTT for collaborative tree modifications.
Leave editing mechanism as-is with trees. Remove graph generation from Forest Ranger and focus on just the connection tool so that it is quicker to build out the trees. Forest Ranger can trigger the Python script to regenerate by domain for now.
Build out the other domains, create multiple diagrams for each domain, and integrate into runbook and sd.
When all are created and working smoothly, create a complete demo, including inference with graph databases, collaboration, and the full flow from analysis to documentation. It could be the sole part of L domain, and could be sent out as you secure work.