Ryan J. Shaw pfp
Ryan J. Shaw
@rjs
Solana is... interesting Dune data has weird gaps in data and timeouts when you try to lookup a tx, nothing works like EVM, it's just all so strange... and yet look at those MCs...
9 replies
1 recast
19 reactions

Brock pfp
Brock
@runninyeti.eth
"skills issue" Up to the data consumer to decide where in the stack that lies 😉 I'm super biased, but imo, most providers add Solana as an after thought / marketing ploy. Solana data is different, and large, but it's not inherently bad. Some of it is actually very good (e.g. balance snapshots with each tx)
1 reply
0 recast
1 reaction

Ryan J. Shaw pfp
Ryan J. Shaw
@rjs
Who's got the skills issue? Didn't say Solana data is bad -- I see the appeal in some of what they're doing, and IMO the EVM is way too low level and it seems like Solana is designed to tackle that. I'm commenting on my experience trying to do some stuff that is very simple on Dune with EVM-based data, but nowhere near as simple or intuitive with the Solana data. When you say you're biased, I was trying to identify what product you work on... is it envio? is envio data only accessible via graphql or can I run SQL queries over it?
1 reply
0 recast
1 reaction

Brock pfp
Brock
@runninyeti.eth
lol sorry! Yeah I should've added more context I'm building indexing.co so heavily focused on the indexing side. From my pov, Dune and others struggle with Solana because they try to shoehorn it into data models and pipelines meant for EVM (or at least don't take the time to do something dedicated) All of which leaves consumers of that data struggling to make use of it --> I meant it all as standing in solidarity with your pain points there. I can definitely appreciate where you're coming from
1 reply
0 recast
1 reaction

Ryan J. Shaw pfp
Ryan J. Shaw
@rjs
skimmed through the white paper, I love what you're doing -- it's very similar to what I've told people here I would work on if I had the freedom one thing I saw slightly differently: projects should expose their indexers in a marketplace, I should be able to run that locally if I want OR the hosted service runs it and sells the stream of data (optionally paying royalties to the maintainers). the idea being that e.g. 100 project consumers paying for 100 API keys is >> more expensive than 1 hosted service paying for 1 API key.... so long as hosted service charges e.g. 50% less than those API keys, it's an incredible value proposition for consumers
1 reply
0 recast
1 reaction