Breaking Ground on Digital Gardening

I’ve decided that I’m pretty bad at capital-B Blogging™. The last “post” I made (as of writing) was the Databricks guest post on Spark Structured Streaming that I did back when I worked at M Science in 2022. That particular post went through a few different stages of revisions:

1. Revising initial drafts as I learned more about Structured Streaming as a concept
2. Revising “personally final” drafts with the Structured Streaming engineering team at Databricks
3. Revising “technically final” drafts with somebody from our company’s internal marketing team to make sure I didn’t sound like a doofus/didn’t make it seem like the company hired doofuses (it hired me, so it totally did, but that’s obviously not something to broadcast openly)
4. Revising the “marketing final” draft with a lawyer to make sure I wasn’t blabbing trade secrets to the general public
5. Revising the “legally final” draft with Databricks’ blog team to make sure the formatting was right (which I didn’t do a good job of, apparently - the final published draft ended up having some weird formatting errors. Not totally sure what happened there ¯\(ツ)/¯ )

That was quite a bit more than I’d expected when initially pitching “What if we published our internal guide for everybody?”. That, along with starting a new job that introduced me to a far higher standard for engineering, raised my own internal bar for “what should I put on the internet?” Since then, I’ve written quite a few posts that I’d describe as ~60% “done,” but didn’t end up hurdling internal (and arbitrary) bar for publishing.

Considering the editorial standards I started this blog with, the idea that I’ve had any bar at all is a bit funny.

The cost of learning is teaching

As somebody that has benefitted immensely from the free body of knowledge that exists on the internet, this doesn’t make me happy. A graveyard of posts I’ve arbitrarily deemed “unfinished” isn’t particularly helpful to anybody, including myself. Just because a set of knowledge isn’t refined enough to be perfectly, 100% clear to a broad audience doesn’t mean it wouldn’t be helpful to somebody, including myself in the future.

I was fortunate enough to stumble on Maggie Appleton’s Ethos on Digital Gardening, which notably resonated with me more than many things I’d read in the recent past. In particular, this diagram she designed clicked something in my brain that’d remained un-clicked for quite awhile:

brief aside: Maggie brings a level of polish and class to doodling that is frankly enviable.

The cost of perfection is infinite

Putting things on a spectrum of “chaos” to “cultivated”, in conjunction with the idea that “the cost of perfection is infinite” (if not asymptotic), puts into hardly-uncertain terms that simply because a thing is not perfect, or close to it, necessarily means that it is unhelpful and should die a slow death in a draft bin.