Will Price
π€ SpeakerAppearances Over Time
Podcast Appearances
Yeah, that's pretty much how it works.
There are some nuances that I would say are out of scope of this conversation.
But basically, you know, Lighter helped Telegram set up a bespoke solution that worked for them and their architecture that uses Lighter on the backend.
And I think we can call it a semi-custodial arrangement that uses decentralized Rails.
And, you know, the right solution for each potential lighter partner is going to be different, but, you know, can be supported, you know, as per their needs.
And I know the lighter team is focused on, you know, getting as many distribution partners as possible because, again, that's one side of the chicken and egg problem of getting liquidity network effects.
I mean, they have a lot of things to do.
That's one of the places that they contributed for sure.
So a human starts to notice latency somewhere around the quarter second mark or 250 milliseconds.
Lighters latency for taker orders is somewhere on the order of 20 milliseconds.
So, you know, well under that threshold.
And, you know, that basically lets you provide an experience that is for all intents and purposes instant for your traders.
Lighter, also on the architectural side, has a speed bump for taker orders.
So basically, any taker order or market order is going into a queue and it waits somewhere on the order of 140 to 300 milliseconds, depending on your staking tier and whatnot.
before it gets processed.
But importantly, the market makers don't know that those orders are pending, but they can update where they quote in the book as fast as their connection to the API permits.
And so this leads to a market structure where if the quote unquote true price of an asset