Skittleballs
Gephi comes creeping back.
I used to do a lot of data visualization with Gephi. All of the prior graphs on here have been Maltego - very detail oriented things, with lots of entity types, and it tops out at 10,000 nodes. Gephi is an open source tool and if you’re familiar with it, a half million node network will creak and groan, but you CAN make tolerable layouts.
Back when Twitter provided API access Gephi was used to visualize large scale social network constructs. You could see relationships of friends/followers, or mentions, or likes. Human networks always had a lot of variability, the programmatic stuff would show enormous repeating patterns.
Now Twitter is an erratic cesspool full of irrational turds. I stopped using it entirely almost two years ago and I’ve gone as far as sink holing it on my network. I have one surviving account for when I absolutely MUST look at stuf there, but it’s sequestered inside Sapper Labs Cloak, so it won’t splatter on anything right here.
This is a tripartite network taken from a mailing list log. This began as a 195k item list and Gephi summarized it down to 13k nodes. I had files showing three aspects of the dataset.
OS to Browser
Browser to IP
OS to IP
I’m not doing anything with this other than satisfying someone’s curiosity about this package. Although, come to think of it, there IS an undocumented Substack API and I DID spend some time coding against it last year …
