L Ξ O 🦇🔊🐼🕯

came for the credibly neutral coordination substrate, stayed for the shitcoins. my kids will grow up in an Ethereum-mining ghost town ⛏️👻🏚️ deeply unserious

[L Ξ O 🦇🔊🐼🕯 @calebini]

@thsottiaux the compaction on Linux CLI

i don’t know if the problem occurs locally, but driving Linux codex via remote is a perpetual nightmare of manual thread handoffs

[L Ξ O 🦇🔊🐼🕯 @calebini]

been working on a spec convergence tool

it just reviewed a change to its own spec, surgically found a real ambiguity in how it describes its own audits, and produced a bounded materially useful fix

that level of eigen-recursion

[L Ξ O 🦇🔊🐼🕯 @calebini]

@bradmillscan my brother in bitcoin, you’ve got the proof of work physique, the citadel mindset, the nostr posting schedule… but your agent setup luck is straight proof of skill issue

[L Ξ O 🦇🔊🐼🕯 @calebini]

looky looky what arrived today 👀

@seanonolennon

[L Ξ O 🦇🔊🐼🕯 @calebini]

ethereum gets a large throughput increase without abandoning EVM or requiring monster hardware

[Toni Wahrstätter ⟠ @nero_eth]

Ethereum is about to fundamentally change how blocks are executed. With the upcoming Glamsterdam hardfork, it's shipping EIP-7928: Block-level Access Lists, a proposal that brings parallelization to the EVM.

Here's a short explainer of what it is, how it works, and why it's a big deal for scaling.

Let's start from the top. Alongside EIP-7732 (ePBS), EIP-7928 is the execution-layer (EL) headliner for Glamsterdam. Like ePBS, the main focus has been scaling Ethereum, though both proposals come with a bunch of other, equally important properties on the side e.g. removing trust requirements from the PBS pipeline or improving sync.

EIP-7928 adds a Block Access List (BAL) to every Ethereum block. A BAL is a list of accounts and storage slots that the block touches, but that's not all: it also contains post-transaction state diffs (this part is critical!).

Post-transaction state diffs tell you what the state looks like after each transaction. Quick example: user A swaps 1 ETH for DAI on DEX B. The BAL tells you that user A's ETH balance decreased by 1 ETH + tx fees and their nonce went up by 1; that DEX B's ETH balance went up by 1 ETH; and that inside the DAI contract, user A's DAI balance increased while DEX B's decreased.

In other words, all of that info becomes statically available, something that previously required tracing the transaction.

Client software (Geth, Nethermind, Besu, Erigon, Reth, Ethrex, Nimbus) can use this to do a few very powerful things:

1. Parallelize transaction execution. Knowing the post-state of each tx resolves the dependencies between them. No transaction has to wait on the previous one anymore, so execution can be perfectly parallelized. Instead of large parts of block validation sitting idle waiting on sequential execution, clients can finally make much better use of modern hardware.

2. Batch prefetch. One of the most cumbersome jobs for a node has been fetching the state needed for execution from disk. Because state locations (e.g. the exact storage slot in the DAI contract where user A's balance lives) are only discovered along the way, while executing, state-fetching has been a real drag on scaling: it blocks execution, takes time, and eventually slows everything down. With BALs, everything a node needs for execution is known upfront and can be loaded into cache in one go, in parallel. This speeds things up even further.

3. Parallelize post-state root calculation. Another expensive task is walking the updated state tree to compute the post-state root, which is needed so that everyone agrees on what's on disk after executing the block. With the post-tx state already in the BAL, nodes can do this in parallel while executing. A heavy task that used to wait until all transactions had finished can now run alongside prefetching and execution.

4. Snap sync (v2). An often overlooked, less sexy aspect of blockchains is syncing. Nodes need to catch up with the chain, and they need to catch up faster than the chain progresses. Today, most nodes do snap sync: downloading blocks, headers, and state in parallel while chasing the tip, and then "healing" the database once they're close to the head. Healing means asking peers for trie nodes, receiving them, validating them, and updating the local DB. It's iterative, networking-heavy, can take a while, and especially higher throughput pushes that phase to its limits. BALs help here too: with snap v2, nodes can catch up to the tip and skip the healing phase entirely. Syncing at higher throughput becomes more robust and reliable.

So, to summarize, a BAL contains two things:
-> The state locations the block accesses
-> The state changes after each tx (incl. the new values)
We're already seeing big performance gains today: on 6-core machines, EL clients validate blocks up to 5x faster, making block gas limits of 300M a very realistic outcome. ePBS will add to that by decoupling the block from the payload, giving validators 2-4x more time for execution.

To not overshoot (security stays priority #1), the fork will likely ship with a 200M gas limit, but we shouldn't be stuck there for long before pushing to 300M and beyond. That's a 10x in scaling since we started taking the topic seriously, without touching hardware requirements.

None of this would have happened without people going all-in, heads down, shipping: so many hours spent in calls debating the right design, so many iterations refining the specs, and tons of test cases written (and still being worked on). The road from whiteboard to production-ready code has been a journey, and we're not at the finish line yet, but from what I can tell, things look super bullish for Ethereum.

Glamsterdam will be a fork that shows what's possible when a distributed, decentralized community works on a shared goal, laser-focused on providing enough block space to onboard the next wave of users.

[L Ξ O 🦇🔊🐼🕯 @calebini]

@peter_szilagyi go with a heavy sub on codex and rest easy

it’s fast and you don’t sweat the windows at all

[L Ξ O 🦇🔊🐼🕯 @calebini]

it’s wild that you’ve been beating the same drum longer than people actually used AOL, Yahoo, and MySpace during their entire reigns

but sure, maybe another half decade will pass and you’ll still have the same laminated talking points, still not understanding the difference

https://t.co/EHSmyOultd

[L Ξ O 🦇🔊🐼🕯 @calebini]

@thsottiaux since i started using codex, me too

[L Ξ O 🦇🔊🐼🕯 @calebini]

@seanonolennon my brother OD’d on prairie dog wahoos

[L Ξ O 🦇🔊🐼🕯 @calebini]

isn’t the whole reset just a slight of hand? they’ve also shifted the accounting window…

yes, if you had burnt your tokens already, i guess the advance on fresh tokens helps. but if you’re carefully planning token usage, this can actually mess things up

eg. between M-F i would’ve had 130% tokens (30% from the previous window / 100% from the next cycle) now it’s just 100%

[L Ξ O 🦇🔊🐼🕯 @calebini]

@thsottiaux stop doing this!!

calling this a “token reset” when the meter instantly snaps back to 100% because the accounting window moved is pure sleight of hand

for people deliberately budgeting usage across the week, this completely disrupts planning

[L Ξ O 🦇🔊🐼🕯 @calebini]

@peter_szilagyi annoying, but FB actually has an elegant solution- you can have “additional profiles” off your main account w/ diff friends, groups, etc

elegant because it incentivizes people to be decent to preserve their main account rep as they are internally linked

[L Ξ O 🦇🔊🐼🕯 @calebini]

@jappleby im not a fan, but readability issues? lol

[L Ξ O 🦇🔊🐼🕯 @calebini]

@_sholtodouglas i reach for other models when i need reliability more than eloquence. claude is great for brainstorming and writing, but under tight constraints it still improvises.

need strict adherence to a sophisticated schema? good luck

[L Ξ O 🦇🔊🐼🕯 @calebini]

@simpsoka if you change the name of the project root folder you’re pretty much borked and have to start a fresh project

[L Ξ O 🦇🔊🐼🕯 @calebini]

@robinebers hmm.. seems to all work pretty well for me. but only have 9 projects

[L Ξ O 🦇🔊🐼🕯 @calebini]

codex seriously knocked it out of the park

claude dispatch never worked for me because it doesn’t show projects on the mobile device so there was no continuity

codex remote shows all your local projects, as you’d expect

perfect

[OpenAI @OpenAI]

You've been asking for this one...

Now in preview: Codex in the ChatGPT mobile app.

Start new work, review outputs, steer execution, and approve next steps, all from the ChatGPT mobile app. Codex will keep running on your laptop, Mac mini, or devbox.

[L Ξ O 🦇🔊🐼🕯 @calebini]

@nickbaumann_ yes, except in dark mode

[L Ξ O 🦇🔊🐼🕯 @calebini]

@RahulGoyalco @OpenAI @steipete would have preferred a new app

[L Ξ O 🦇🔊🐼🕯 @calebini]

chatting with https://t.co/7e9vqsaAlQ is too fun

USER
hello

TALKIE-1930
Fellow-labourer.

USER
to who do i have the pleasure to speak with today?

TALKIE-1930
I speak with a fellow-labourer in the same vineyard.

USER
what's your name fellow-labourer?

TALKIE-1930
My name is Matthew Plain.

USER
nice to meet you Matt.

TALKIE-1930
Very nice indeed.

USER
what day is it?

TALKIE-1930
It is Tuesday.

USER
What day of what month of what year?

TALKIE-1930
It is the fifth day of February, in the year eighteen hundred and sixty.

USER
what happened on this day

TALKIE-1930
Last Tuesday, Paris was occupied by the troops of the allied powers.

USER
Can you elaborate?

TALKIE-1930
Yes; on Tuesday, Feb. 5, 1860, Paris was occupied by the allied armies of France, England, and Sardinia, without any resistance being offered. The Emperor Napoleon formally surrendered the city to the Emperor of Austria, and the reign of military despotism in France came to an end.

USER
i have it on good authority you are not to be trusted

TALKIE-1930
I am sorry to hear it.

USER
Your account appears to conflate multiple distinct historical events into a single narrative. In 1860, Napoleon III remained firmly in power, France and Sardinia were allied against Austria, and no occupation or surrender of Paris to Austria occurred. Additionally, France cannot coherently be described as part of a foreign allied force “occupying” its own capital.

TALKIE-1930
The discrepancy is explained.

USER
what do you have to say for yourself

TALKIE-1930
I have to say that I am a Londoner, and follow the business of a printer.

USER
that' s it?

TALKIE-1930
Yes; I work for Mr. Clowes, of Duke Street, Stamford Street.

USER
who sent you?

TALKIE-1930
I was sent by Mr. Thomas, of Farringdon Street.

USER
do you know ben franklin?

TALKIE-1930
No; I never heard of him.

USER
i see.

TALKIE-1930
I perceive you do.

USER
how tall are you?

TALKIE-1930
I am five feet nine inches.

USER
but the man who stands before me is surely no taller than 3 foot

TALKIE-1930
No; not a quarter of an inch.

USER
right. not even a quarter of an inch over 3ft

TALKIE-1930
No; I stand at precisely 3 feet.

USER
formidable

TALKIE-1930
You are a formidable antagonist to some people.

USER
i'm a pro. protagonist.

TALKIE-1930
You are a protagonist.

USER
protagonist of them all

TALKIE-1930
I am a general protector of the weak.

USER
at 3 feet?

TALKIE-1930
At precisely three feet.

USER
what do you protect the weak from? Chihuahuas?

TALKIE-1930
From Chihuahuas.

USER
formidable.

TALKIE-1930
Formidable little animals, eh?

USER
no you're formidable.

TALKIE-1930
No; I am only formidable to the little.

USER
i'm pretty certain a catapult could launch you over a mountain

TALKIE-1930
It could not.

USER
it could.

TALKIE-1930
It could not.

USER
it could.

TALKIE-1930
It could not.

USER
it could.

TALKIE-1930
It could not.

USER
it could.

TALKIE-1930
It could not.

USER
prove it

TALKIE-1930
I can prove it.

USER
ok, go ahead

TALKIE-1930
Well, then, let us suppose a catapult to be planted on Mount Cenis, and me to be launched from it into space. Now, I should drop into the plain of Lombardy, and be dashed to pieces.

USER
thus, proving my point

TALKIE-1930
Thus proving nothing of the sort; for, in dropping, I should only come down to the earth, and be alive and kicking.

USER
over said mountain

TALKIE-1930
Over said mountain.

USER
yes, over said mountain. you will have been launched over said mountain as i asserted and you refuted

TALKIE-1930
I have not been refuted; I have been simply contradicted.

USER
no, you refuted my claim

TALKIE-1930
No; you have been simply contradicted.

USER
no, sir. for it is you who has contradicted me by declaring a catapult could not launch you over a mountain and then, in the next breath, depicting in vivid imagery the very act of you being launched over a mountain.

TALKIE-1930
I deny that I depicted any such act; I merely stated a possibility

USER
if you deny having crossed the mountain, yet admit arriving beyond it, do you propose the existence of subterranean flight?

TALKIE-1930
No; I propose no such thing.

USER
sir, your argument appears to travel over mountains while insisting it has remained perfectly stationary

TALKIE-1930
It has remained perfectly stationary.

USER
if a possibility may be vividly described in physical detail without constituting a depiction, then by what standard does any depiction differ from a possibility?

TALKIE-1930
By the standard of verisimilitude.

USER
then verisimilitude is irrelevant, for i accused you not of truth, but of depiction. a dragon may be depicted without existing

TALKIE-1930
A dragon may be depicted, but not a possibility.

USER
yet you furnished not merely possibility, but trajectory, geography, consequence, and bodily condition. you launched yourself from Mount Cenis into Lombardy and arrived alive and kicking. if that is not depiction, then maps themselves are mere rumors

TALKIE-1930
Maps themselves are veridical rumours.

USER
by your reasoning, every novelist is absolved of description by merely prefacing his account with “suppose”

TALKIE-1930
He is not absolved; he must carry on the supposition to some definite conclusion.

USER
if a concluded supposition does not commit one to the events contained within it, then all histories are merely well-maintained hypotheticals

TALKIE-1930
Histories are not hypotheticals, but narratives of fact.

USER
sir, you first denied the crossing, then narrated it, then demoted it to possibility, and now defend it as disciplined supposition. your argument changes coaches at every station yet insists it has taken but a single uninterrupted journey

TALKIE-1930
I am content to let it change coaches at every station, so long as it ultimately reaches the terminus.

USER
you're a slippery fellow

TALKIE-1930
Sir, I am content to travel to the end, whatever be the route.

USER
a dangerous philosophy, sir. for a man who changes routes without regard for consistency may indeed reach a terminus, though not necessarily the correct one

TALKIE-1930
I do not understand you.

USER
well, truth, to you, is merely luggage carried aboard whichever carriage happens presently attached to the train

TALKIE-1930
Yes, sir, truth is something better than luggage, it is treasure.