KtorZ

An interesting idea, yes by default for CC; but full vote when one disagrees. I even think it's implementable without any changes in the ledger (albeit with a one time full cooperation of the CC members) 🤭

@ElderM I really hope that you manage to find stuff we don't know and haven't discovered yet. That'd be really cool. Testing wise, there's probably some nice avenue to explore where this whole effort could benefit everyone! No harm done.

For vibe-node, one thing I really want to stress is that what I am doing is not an argument for not funding Amaru and Dingo. Quite the contrary. I am burning 500 million tokens a day. I am starting 2 steps before most vibe coders because I have a wealth of experience in how to use AI for production grade engineering. The thing I am missing that I would need as a next step is either a lot of time for me to read and understand everything, or hire an expert like @wolf31o2, @_KtorZ_ , or @Quantumplation to do a deep dive into the code to make sure it's not severely messed up. The thing is, we don't get deep expertise in the ecosystem unless we pay them to do this stuff manually. If we pay our experts to stay dedicated to being experts, we can help to solve the node diversity problem by supplementing with nodes generated using my approach. It is not a replacement for these guys doing what they do. AI always needs a human evaluator. The measuring stick cannot be "it runs", unless you want more situations like happened last year with the bug that caused a lot of chaos. If we didn't have these experts to identify the bug AND fix it in record time, Cardano would have had a crisis on its hands.

Questions about CIP-113 (a.k.a. Programmable Tokens) and the recent work around it? Tune in for a deep dive!

No. 1. None of the reported numbers on the Aiken version reflects the current implementation. E.g. - The transfer simple is currently at [cpu: 68.97M, mem: 193.70K], so about on-par - The mixed many transfer sits at [cpu: 107.93M, mem: 307.89K], well below the Plutarch's implementation - etc... 2. The difference IS about micro-optimisations. We are talking couple of % in the 100M cpu and 100K mem range. So 1% difference is about 0.01% of the overall transaction max budget. Said differently, a difference of 1% roughly translates to 200 lovelaces of execution cost; that's about $0.00005. 3. The size of the Plutarch scripts is 4522 bytes. Aiken's latest scripts weight 2787 bytes. Each byte costs *at least* (if reference inputs, and in the lowest price tier) 15 lovelace. So that's currently a difference 26,025 lovelace. So even if the Aiken version was 130% more costly; it would still be balanced by the script size. For the complete analysis / rationale, please read: github.com/cardano-founda…

Comprehensive benchmarks of Plutarch implementation of programmable tokens against the Aiken implementation (on average the Plutarch implementation is 39.6% more efficient) This benchmark covers real use-case scenarios pulled from actual DeFi transactions from the ledger history. For a practical example, a DEX transaction which processes 16 swap requests when converted to the equivalent transaction on the programmable tokens mini-ledger is 28% more expensive in the Aiken implementation. In every single case, the Plutarch implementation is more efficient. In the largest observed difference, the Aiken implementation consumed 3.33x more CPU and 2.4x more memory than the Plutarch equivalent. The difference here is not micro-optimization, it is an entirely separate class of efficiency. We are not talking about a 5% difference; we are talking about an average difference in efficiency of 39.6% across benchmarks produced from real world use-cases randomly sampled from mainnet DeFi protocols.

Going live discussing Net Change-Limit & constitutional committee role in Cardano.

Well, the command-line gives you the raw (voting) stake distribution for DReps, which includes ALL dreps irrespective of them being active or not. So you have to couple that with the `drep-state` command to filter out those whose mandate have expired due to inactivity. This is because the expired DReps are only discarded during the ratification (since they may come back 'online' at any point really); so their stake distribution is preserved but not accounted for when it comes to vote ratios. And yes, pretty sure of it!

Ola put quite a lot of work into that, comparing with CLI values, even found CLI issues and reported them: github.com/IntersectMBO/c… This comment (coming from you) makes me second guess it, but we can't do it more detailed than "to the lovelace" coming from the CLI/node. So, if you are sure about it, we will put in more work and try to verify what you said.

@eternlwallet @NicolasC3rny @Amaru_Cardano I believe you are not excluding inactive DReps, thus counting them as "no" votes when they should really not be counted at all.

@NicolasC3rny @_KtorZ_ @Amaru_Cardano According to our numbers there is still 131m ADA voting power missing on the YES side, being at 64.7% currently.

@Quantumplation @aiken_eng Spent a good portion of the afternoon on it 🙈

@_KtorZ_ @aiken_eng If you're up to it, would you review my source maps PR? github.com/aiken-lang/aik…

Squashing few bugs and small quality of life features on the @aiken_eng compiler today. If you've got any request/ideas (other than changing the formatter 🙈), shout!

Only ~5% left to reach the threshold! If we manage to get there before 21:45 UTC this evening (i.e. before the current epoch ends), then that's a massive piece of mind and relief for the Amaru's team. adastat.net/governances/2f… Happy to answer any questions about the proposal🫡

@am_tortl And you're planning to give a large portion of those $250k to the teams behind the free open source projects your Claude stack is based on, right? Because funding open source is the right thing to do 🥰, right?

Claude just 1 shot your node diversity with a fully detailed step by step checklist. 250k plz.

@ElderM @wolf31o2 Free time🤨? We don't do that here. Cardano is your free time. #NotACult

@wolf31o2 @_KtorZ_ Wait...do people work on 1 project at a time until its complete? What do they do with their free time?

My open source work for Cardano:

@wolf31o2 @Scalus3

@_KtorZ_ @Scalus3 Yup. The Lantr team are amazing.

Be like Scalus. Relentlessly build in open source. Deliver a mix of useful and advanced features. Raise the quality bar continuously. Make Cardano available to a massively large ecosystem (Java / Scala / Kotlin / Clojure). Be like Scalus 🫶

Discussing Amaru and node diversity with the Amaru core team *right now*, feel free to tune in to join the conversation! Regardless, the call will be recorded for those who cannot attend. 💝

@jonahkoch @atriumlab It's obviously exaggerated but you get the gist. It's not a multi-million dollar budget.

@_KtorZ_ @atriumlab Wishful thinking 🤭 nothing in life is free. I get it, theoretically minimal upkeep if nothing changes.

@bladepool0x @blockfrost_io Bitcoin Lightning might have been one. So I've heard.

Experiments of the evening: deploying across two L2s on two different blockchains, orchestrating 5 pieces of infrastructures. Let's see how it goes🤞. Props to @blockfrost_io for accepting ADA to pay for requests. Worked like a charm.