Jebus.. #Python and all it's batteries. So #Matplotlib knows how to talk #SVG already. I just started to speculate on how hard that would be to implement and.. it's already there built into Matplotlib.
Well. That worked out rather nicely then.
Being back working in Python feels way too magical at times.
Turns out the mechanisms I'd already put in can do graphs and plots already with clean readable markdown - no new plumbing needed.
I'd have thought that a far-fetched goal moments ago.
@stevenaleach it’s nice to open SVG as text and just… fix the thing, if you need to.
@kschroder I've never actually used SVG before - was vaguely aware it was able to be in-lined in HTML and went from "wait, those diagrams in GitHub #markdown are SVG aren't they?" to "My mardown supports inline expansion of named stuff from yaml frames (and the markdown is contained in the frame and can reference its own frame)" to "how to make SVG from plot?" to "MatPlot does that" to "my system does plots?! Really? It just does that?" in all of 10 minutes.
@kschroder Like literally stunned to just discover/realize that being able to do inline expansion of stuff into the markdown before it is rendered means being able to have named stuff in markdown turn into pretty pictures when it renders. I mean... that *kind* of stuff was the point of the mechanism, but I didn't realize it could just do something like that with no further work needed. That's... cool but almost wrong?
@stevenaleach so wrong it’s right
@kschroder Turns out not be *quite* that easy, but close. I've tested with just an SVG circle as the blob and it works... My one attempt with Matplotlib seems like it allocated space but my plot wound up off the screen or something.. so maybe a little cleaning will be required, but that's not not my code's responsibility to fix and something I'm not getting side-tracked investigating. The presence of a big red dot on my markdown test page says things just work ;-).