starry (@jstrry) header
starry (@jstrry) profile picture
Taiwan
Joined November 2011
Photo by starry (@jstrry)Photo by starry (@jstrry)Photo by starry (@jstrry)Photo by starry (@jstrry)Photo by starry (@jstrry)Photo by starry (@jstrry)Photo by starry (@jstrry)
Poof V2 is LIVE Type a prompt. Get a full onchain app on Solana. Not a frontend with a wallet button bolted on. Wallet auth, smart contracts, security scanning, backend, UI. All from your first prompt. Think Claude Code, but purpose-built for Solana 🧵👇
69
31
20
270
48,101
Our @Solana mainnet rollout is LIVE. Type a prompt. Get an onchain app. Welcome to the future of Solana app building.
116
77
50
604
251,207
Looking for 2 roles, preferably with Solana experience to solve hard problems. Low latency, systems engineer Data Engineer Send resumes in DMs
5
3
21
2,977
Anyone who says that solana has won at anything is a sleeper agent. It’s 100% psyops. There is a mountain of work left to do on all fronts.
27
121
60
1,440
207,992
Introducing the new Vote Credits Ranking page! Validator operators can now easily see how their vote credit leaderboard rank has changed over time epoch to epoch!
1
2
3
917
this is a huge moment in the history of save. from starting out at a hackathon, to becoming the biggest solana DeFi protocol, to near death from the FTX collapse, to being reborn as save. now HALF of all SAVE will be burned.
The proposal to burn 50,000,000 SAVE (50% of the total supply) has PASSED! Thank you to everyone who participated in this monumental decision market! The burn will take place soon.
2
1
14
2,645
Great to see screenshots of the vortex validator income dashboard in the wild but it needs a lot of work still before it's useful for validators wanting to understand how to improve block revenue. Suggestions welcome!
Why spend time sabotaging other validators when we can just improve our own! While competitors are making excuses, @temporal_xyz is cooking. Stake with Temporal.
2
2
942
i wrote a thing about all the different teams in north korea dedicated exclusively to fucking your shit up and how you can know exactly which one just ruined your entire month paradigm.xyz/2025/03/demysti…
38
177
16
759
83,949
starry retweeted
Introducing Accelerate 🇺🇸 Scale or Die. Ship or Die. Join us in NYC from May 19-23 to propel crypto innovation forward in America!
878
700
229
3,428
1,445,772
Replying to @ferric
Assume that all assumptions are wrong (except this one). Develop a principled system level understanding of the thing you want to optimize. You don't have to be an expert in every single area, but you must understand how things interact. You can only act on measured data. Micro benchmarks are a waste of time. Never look at wall clock time, it's too inaccurate. Instead, profile. Profile with prod load. Test load is garbage. Take more than one profile (3 is a good magic number) to account for variance. Find the biggest bottleneck and study the root cause. You are not allowed to write code until you're confident you know the root cause. You must understand it or you risk implementing a solution that appears to work now but will fall off a cliff with a different load pattern. Develop a theory for a fix, then implement it. Be scrappy don't spend a month to write the perfect code. Do the fastest viable hack to validate the idea. Most ideas end up not working. If the fix doesn't seem to work, don't just throw it away. You must understand why it's not working. Once you understand though, yeet it asap, don't get attached to the idea or code you've written. Try the next idea, repeat until something works and the bottleneck is gone. Once the bottleneck is gone don't go full autism into optimizing to death. Making one component 1000x faster when the next component in the pipeline allows for a global 2x speedup is a waste of time. Find the next bottleneck, then go back to square 1. Congratulations you're now a performance engineer.
17
12
3
119
34,072
Announcing Alpha Cards: Built over 2 years, from ETHDenver to Devcon, featuring nearly 200 historic web3 moments. Unleash memes, myths and the magic of web3. More:
9
9
6
67
11,932
the cavey validator upgraded to 2.0.15 alongside the network a while ago. metrics look good. excited for 2.1 :) as promised, i wanna put out some educational content. either some code examples or maybe livestream some coding during a weekend what do you want to see? i’m thinking 1) solana programs in rust 2) solana programs in asm (eg escrow) 3) janky blockchain in a day 4) HPC and data structures in C 5) [REDACTED] alpha
22
3
2
74
28,171
starry retweeted
Did you know you can significantly improve your transaction priority and reduce CU consumption by using the new setLoadedAccountsDataSizeLimit (Compute Budget) instruction? See the new article by Anza Core Engineering linked below.
7
14
5
70
29,158
INCREASE BOBA REDUCE LEVEL OF SWEETNESS AND ICE
Excited to be supporting @TaipeiWeek this year and massive thanks to @Gen3Network, @ZeusNetworkHQ and @deanmlittle for twisting my arm to get involved and also doing all the leg work! Taiwan is exploding with web3 talent rn and we’re ready to show up!
1
3
10
12,394
What's your favorite metric to understand if the cluster is happy or sad? I like tower height
3
5
21
3,688
The biggest cheat code to get better at programming is to do open source, where you have to read other people's code, modify it, and learn to collaborate with people that don't necessarily share your views or priorities. Hire (and generally fund) open source people.
The biggest cheat code to get better at programming is to read more code. Many software interviews make you solve brainteasers, but this doesn't generate the correct signal. The hard part of engineering is being thrown into the ocean of a large codebase and trying not to drown.
3
8
1
49
12,709
10/19/2023: 28 SOL in tips 10/19/2024: 26,181 SOL in tips 🤯
2
1
30
1,775
starry retweeted
missed @SolanaConf? here’s all the talks: youtube.com/playlist?list=PL…
1
5
13
1,170