A small amount of footage from the 2017 solar eclipse. Taken in 4K HD with an Panasonic FZ300
Will be getting some footage of the 2024 eclipse with the same camera along with an audio field recording to hopefully capture the sounds of nature being confused by the event
Trying another livestream of the stagnant water
Was finally able to spot a rotifer for the first time, but it ended up climbing out from under the cover slip
Also spotted what I thought might've been a hydra at first, but it was an amoeba with it's cytoplasm fully extended into tentacles, which it eventually retracted and went on it's way.
Various other microbes were spotted too
Decided to make a slideshow presentation filling in the months since my last videos.
00:00 INTRO
00:29 WHERE DID YOU GO
03:35 BIBLE
09:15 WHAT ABOUT FINANCE STUFF
12:34 NFTs
24:53 MUSIC
28:19 MUSIC CONT.
31:14 ELECTRONICS
37:06 DE-GOOGLE
46:50 DECENTRALIZE
49:53 AND LIFE
Slideshow created with: https://www.libreoffice.org/discover/impress/
Audio, compressed, normalized and silence removed with: https://www.audacityteam.org/
Video created with: https://ffmpeg.org/
So, I've decided my abuse of Redis is BAD and my irrational fear of SQL needs to be confronted
I've read a few docs on SQL (SQLite) and have decided to see how the better-sqlite3 NPM package works out https://github.com/JoshuaWise/better-sqlite3
I think I should be able to completely drop Redis eventually, but will start with the code that pulls down data from the /contracts LedgerX endpoint and then store IV surface data there next.
Just wanted to stress free stream some COOL STUFF, like a vintage diatom arrangement, it's show-n-tell and maybe start working on something for the market making project
I've been putting this off for a few days.
It feels like a big job, I need to basically remove a lot of old "ways of going about things", nodes, and start querying the orderbook database instead of relying on websocket messages themselves. I feel like having our own local "game loop" sort of clock will be more conducive to our style of trading (market making), so instead of being reactive to every single market change, we can look at the up to date book every at sub-second intervals to recalibrate our orders and model. Right now, since I built the system as an IV chart with dreams of trading, one of the major flows takes non-authenticated book top messages from the FTX US Derivatives websocket and calculates an IV surface, updating an SQLite db with a theoretical IV for each contract (along with it's actual bid and ask IVs)
From there it sends messages to the front end which are rendered as charts. You can see how this will end up rooting up old code and throwing it out in favor of a more local "polling" type system that checks the in memory SQLite db state and generates the IV surface. From there we can send messages like we were, but it should be much less hard on the CPU since websocket messages could arrive in bursts causing the system to calculate 10s of IV surfaces and render them per second...
This new way would allow us to take control of the amount of CPU to use and have better control of our timing and certainty of the exchange book state (I hope). It should also make programming and testing various algorithms easier.
I don't know if it's the right move or not, but it seems to be.
I hope no one searches up these streams hoping to find a tutorial, it's mostly me bumbling around figuring stuff out
Looking to figure out the Node-RED dashboard (particularly the template node along with chart.js) this time