Searching across Drummer (or any other "tools for thought)

[kevinctofel]

@scripting wrote a blog post earlier this week about where to find different things he's written about via Drummer, although the question may be larger than that: How can we search across various "tools for thought" to find related thoughts, thought patterns, specific topics, etc...

I took a high-level stab at one possible way and outlined it in a repo. I also tested a very basic POC. Would love to hear feedback on what I've dubbed the "BrainBase" for now. 😉

Essentially, I want to suck in things I've written in outliners, on Twitter, etc.... into a database via standard formats, and create a web-based search tool across all of that data.

[scripting]

@kevinctofel -- first thanks for picking up the ball here. as far as i know you are the first.

okay -- next -- a bit of background. there are products in the "tools for thought" space that have both an editor and a database. those are the databases i want to get content into.

i want to use my editor and benefit from their database.

by doing so i hope to learn what their databases actually do (I've heard a lot of claims that sound kind of unbelievable to me, but let's give them the benefit of the doubt), and also break this assumption that to be in a database you have to use their editor. this is a huge problem in tech kevin. for example to have a substack newsletter i have to use their editor. to post to facebook i have to use their editor. medium has an api, but it's limited in a very important way, so effictvely you have to use their editor. see the pattern? i do not want that to happen here in "tools for thought land."

so it's not just any database i want to get my ideas into, it's those databases.

now some of the companies are receptive to this idea, esp my friends at LogSeq. but this is the kind of thing it's best if the users do. independence matters here.

i'm sure you get it, but to really nail it down, the other day we were having a discussion here about connecting Drummer with GitHub. Lots of power there, imho. But it wouldn't be the same if it were connected to a new different kind of repository manager. Getting into GitHub itself is the goal.

Now one other thing -- you can have OPML in JSON. We have a toolkit that reads OPML files and returns JavaScript structures. From there, do whatever you want. Save it as JSON. It's up to you. I wouldn't have left this stone un-turned-over, so before assuming something like that doesn't exist, ask.

Maybe you don't program in JavaScript, okay, let's convert that toolkit to the language you work in.

[kevinctofel]

Ah, I see the distinction. I was approaching it from a different angle mainly because I've evaluated several of the tools for thought on the market and most of them were closed source when it comes to their database (if not everything). I know some are open-source but most are not.

So my mindset was how to get copies of my data out of their databases into something open and useful (to me). Appreciate the explanation and for pointing out the Node package for OPML in JSON. I do use JS for most of what I do, so this is super helpful. :)

I may continue down my path for a little while when time between classwork allows. There isn't much of that time to be had, unfortunately. But, it would be a good learning experience, for one. And while I'd hope these tools go in the direction you're thinking, until they do, my digital thoughts are all over the web.

[scripting]

@kevinctofel -- you got it right. i don't care about the source as much as they allow the users' data to flow where it wants to go. my approach is, as long as they leave their world closed off to my data, we'll just have to build everything they built behind their wall. but the logseq guys are the real thing, or so it seems. so as long as we have one partner, that's enough to get the ball rolling.

a similar situation happened with scriptable apps on the mac in the early 90s. we got a few apps to go first, and once we did all their competitors had to follow suit. if pagemaker was open to our scripting, then so would quark.

we had no such problems in rss, because of wired, red herring, motley fool and salon, and then we got the ny times and it was off to the races.