Kevin

@typedarray.eth

356 Following

143430 Followers

Kevin pfp

@typedarray.eth

Glad you're enjoying it! Thanks @greg :)

1 reply

0 recast

0 reaction

andrei pfp

seriously impressed with ponder.sh - what a great piece of software! Thanks again @greg for getting me started.

1 reply

1 recast

7 reactions

Kevin pfp

@typedarray.eth

POV: you work on tooling that's compatible with *any* EVM chain

10 replies

3 recasts

54 reactions

Kevin pfp

@typedarray.eth

Depends on how large the app is (e.g. how many historical events). The smallest apps can run on Railway/Alchemy free tiers. Lmk if you have any other questions!

1 reply

0 recast

1 reaction

Kevin pfp

@typedarray.eth

Historical data on reorgs is pretty scarce IME. Some block explorers record them. Not aware of an analytics-oriented tool that does it well.

0 reply

0 recast

1 reaction

Kevin pfp

@typedarray.eth

Upgrade and enjoy! Chat: t.me/ponder_sh Release: github.com/ponder-sh/ponder/releases Docs: ponder.sh/docs/block-events

0 reply

0 recast

0 reaction

Kevin pfp

@typedarray.eth

Block indexing is built using the same RPC-compatible, reorg-aware engine that powers log indexing. The `interval` option accepts a block interval, but it’s easy to convert a time interval (e.g. every day) into a block interval using the chain's block time.

1 reply

0 recast

0 reaction

Kevin pfp

@typedarray.eth

Ponder 0.4.16 introduces block indexing Block indexing is great for cron / scheduled tasks: - Read the latest price from an oracle contract every minute to build a price chart - Aggregate data every hour to update protocol-wide statistics

5 replies

5 recasts

44 reactions

Kevin pfp

@typedarray.eth

Ponder 0.4.15 improves crash recovery behavior Before, if a fully indexed app crashed (often due to a short-lived RPC reliability issue) it would need to reindex from scratch Now, Ponder apps periodically checkpoint their indexing progress. After a crash, you can restart the app and it will pick up where it left off

5 replies

4 recasts

55 reactions

Kevin pfp

@typedarray.eth

This gave me more fomo than any other farcon content

0 reply

1 recast

5 reactions

Kevin pfp

@typedarray.eth

You wouldn’t last one hour in the asylum where they raised me

6 replies

2 recasts

45 reactions

Vitalik Buterin pfp

Vitalik Buterin

I agree but then again I'm the sort of person who doesn't notice the difference between watching movies in 1024x768 and anything higher.

0 reply

1 recast

4 reactions

Kevin pfp

@typedarray.eth

Holy shit. This is a technological breakthrough.

0 reply

0 recast

2 reactions

Kevin pfp

@typedarray.eth

Yea, I'm just skeptical that 0.1s unlocks a better user experience in practice if you're already doing client-side prediction, optimistic UI updates, etc. My guess is that the point of diminishing returns is closer to 500ms. I'm probably very wrong though, just genuinely curious. cc @frolic

2 replies

0 recast

1 reaction

Kevin pfp

@typedarray.eth

Released in 0.4.10 https://github.com/ponder-sh/ponder/releases/tag/%40ponder%2Fcore%400.4.10

0 reply

0 recast

0 reaction

Kevin pfp

@typedarray.eth

In the next version of Ponder, error logs for SQL constraint violations are more human-friendly

1 reply

3 recasts

33 reactions

Kevin pfp

@typedarray.eth

0.4.9 adds support for transaction receipts and speeds up historical indexing by ~50%, thanks to @kjs https://x.com/kyscott18/status/1782826942908997653

3 replies

2 recasts

36 reactions

Kevin pfp

@typedarray.eth

TIL degen chain has a very short block time (currently 0.1s, average 0.4s). For what applications does such a fast block time offer tangible benefits over a 2s chain like Base?

5 replies

4 recasts

50 reactions

Kevin pfp

@typedarray.eth

I might be misreading the diff, but looks like you removed the query for `alreadyProcessed`, such that it will now always be undefined? (And perhaps that query was expensive?) Regardless, super curious to get to the bottom of it

1 reply

0 recast

0 reaction

Kevin pfp

@typedarray.eth

Desperate for more details, pls

1 reply

0 recast

0 reaction