I am an operating systems nerd. I got nerd sniped in the late 90s by the open source movement and linux. It's just been super fun working one these hard systems problems for the last 20 years.

I see the purpose of decentralization is the same as fault tolerance. The meaning has expanded to include things that are academically hard to quantify, like if the majority of nodes are faulty, which could mean censorship or trying to commit an invalid state transition, etc..

It's the developers! Solana is an OS. My parents should never care about what operating system they use, but they should love the applications built on top of it. So our goal is to accelerate the devs as much as possible.

the code is open source, anyone can pick up the code and run with it, and there are so many copies of the state that its virtually guaranteed that at least one replica will survive. So if the users want to they can always continue no matter what happens to labs. So what does shutdown mean in this context?

I think there is feature overlap, so that make it competitive. But the designs are so far apart that it's really cool to see the r&d done by ethereum folks and try to apply those lessons towards a totally different goal.

stuff that reduces costs for consumers. like helium. find something that people overpay for and try to build a protocol around it to make a cheaper/faster version of that product.

sollinked/tiplink/usdc/sphere payments

helium, solana pay companies like sling/espresso/etc.. id like to see a boring consumer use case break out in 2024.

People were tribal about the file system they chose in their linux kernel config in the 90s. I think it's fine as long as people can step away from their screens and touch grass, and that the passion drives people to build better open source systems.

I think once a network is permissionless, that basically anyone can participate in any part of it, what really matters is the number of people that actually care enough to participate across all parts. This has little to do with number of nodes or stake or whatever.

consumer hardware is really really really competitive. Network effects require single player benefits to get going.

The Back Channel, by burns

1 to maximize post utilization

lol. papa dells pizza, the office in Urbana or brothers.

there can be only one

I think @dwr.eth's vision is way bigger and wants to include every possible crypto user.

oof, man. so much technical debt. I wish we could have shipped with asynchronous execution from the start.

I think the best open source actually comes from companies with commercial revenues, so when they publish something that they use they will continue using it and maintaining it.

32x+ more throughput every 10 years.

I think Jupiter routing is really showing the stark difference in the designs. Because it's a single atomic state machine, all the market liquidity is aggregated into one, even for 20 cent trades.

a network that is sufficiently decentralized can't have a permanent CR failure. Censored users can always fork their own instance and run it.

I wish solana builders always shipped open source first, and had a roadmap towards immutability for eth folks, that every 2 years hardware can do 2x more, and that the technology they are using should do 2x more.

more work :). I hope something like helium takes off and really demonstrates that the whole crypto loop can build a better product for a very large consumer base then a centralized provider.

high performance athletes make really good early startup employees

technology is good enough that with enough data it can probably predict exactly what I would say 99.99% of the time

payments. I think the fiat rail connections around the world are getting better and better, and I feel very competitive when they aren't solana first

more apps that save consumers money

grant programs, dunno why everyone runs one, but they basically never work.

1. a bunch of consumers use a web3 app 2. no clue on share, but payments, then maybe something like depin/helium 3. I think last cycle was the transition from cypherpunks to hoodies, and this one will be hoodies to suits :)

stability, single code base. Firedancer folks are building the second client. So it's all in progress, just wish it could be faster :)

developers developers developers...

stability, stability, stability...

why would I want to stop some aspiring systems engineer from building their own design?

applications with consumer traction. payments, depin, stuff like that

Checkout the last hackathon winners. It’s just awesome to see new founders with crazy ideas.

not familiar. by OS I mean the kernel. Tons of UX frameworks can run on top of it.

Nothing is stopping devs from trying L2s, except that fees on the L1 are so low that a general purpose L2 is not likely to be commercially viable. There is a parato efficient system that is single giant atomic state machine that is synchronized at the speed of light. Thats what I want to build

Lack of interfaces between programs. ABIv2 work is trying to fix it.

things that lower costs for a large range of consumers.

for solana, we just need more founders to do the grind. 1/20 find PMF. So for every forecaster there were 20 web3 social network attempts that you've never heard of.

consumer hardware is really really really competitive. I think passkey generated seed phrases have closed the gap quite a bit on the security side. I still want a mobile application distribution channel that goes straight to all the NFT power users without the google and apple tax.

00s and 10s Moores law was in hyperdrive for mobile chips. so it was really hard to keep up, felt like every year was a 2x improvement in hand held compute. It was technically very challenging and I had the opportunity to go really deep on design with awesome folks. Surfing is good for software architecture :)

Probably the solana twitter sphere. Would be cool to see more on farcaster

Developers need revenue or funding. So we try to help them scale their businesses

why not?

yea, sybilling is probably a net negative. especially for governance tokens. Who are the users you want to actually govern?

find 10 paying customers figure out what and why they want this product optimize the feature set for them

two vertical screens, vim

looking for AI tools good enough to generate js boilerplate and unit and integration tests for solana devs.

the right big company can have amazing mentors. The right startup can massively accelerate your growth by giving you way more responsibility then you can handle :)

Who are you building for, what problems are you trying to solve?

moving is really hard. I think apps that maximize growth across crypto will support multiple chains. I expect native solana teams to try to compete as well.

1. we probably need 1559 economic back pressure on every knob in the system 2. the backchannel by burns

find an app that would save them money. don't pick the kernel, pitch the application