As a non-coder I deeply appreciate the great and successful effort you have undertaken to explain these mechanisms and their improvement in layman's language. You have clearly edited this post with care, as I found only one typo, which is more than I can say for this coemmnt.
"HAF keeps more blockchain data than hivemind
"Perhaps the most important change is that a HAF server keeps a raw copy of all the Hive blockchain data, so virtually any blockchain data that someone may new to create a new set of API calls..."
The text '...blockchain data that someone may new to...' may not actually be a typo, but it makes no sense to me, and the rest of the post is so well written that even I felt well able to grasp it's meaning. The word 'need' seems appropriate in that sentence, at least to my unschooled eye.
I further appreciate what seems to be very beneficial strategies and good application of them in decreasing node expense and fragility of the platform, also dramatically increasing scalability, so that if flaggots ever quit driving good creators off the platform all of us can benefit from rapid growth and a shot at the moon. I confess I wondered what would have happened to Hive had the traffic that quickly drove Meta's Threads to ~100M users, or even a fraction of it, wandered thisaway, but it seems we have been well insulated from such potential trauma by those diligent beavers damming the flow of new users and retention that might flood our little pond.
We are well afforded scalability by your endeavors so that if they ever slack off we can prosper thereby (unless I have misunderstood your statement regarding one node being able to handle all current Hive traffic), and I am grateful.
Thanks!
Good catch, you're correct. It was supposed to be need, not new. I've fixed it now. I do try my best to avoid mistakes, but I sometimes go back and edit sentences to try to improve their readability and that can ironically increase the chance for typos of various sorts.
Yes, you shared the same concern I've always had: the original code just wasn't scalable enough to handle a lot more traffic without expenses increasing dramatically. That's why we've put so much work into improving scalability: otherwise there would have been no point to trying to really grow the user base. And you've correctly understood that one node can currently easily handle all of Hive's traffic, so we've made a lot of progress on that front.