Viewing profile — mweidner

Indeed, I do not buy this argument. Would China's progress be close to where it is today without the US labs' examples? Would any of this be happening if OpenAI had not created Cha…

The folks I met who were talking about AI Safety in 2018 were certainly sincere, and the two people I knew who later joined Anthropic seem like the type to do it for the greater go…

Is the idea to keep the world in balance via MAD? I could see that, though it's a dangerous gamble. From Richard Rhode's "The Making of the Atomic Bomb", I got the impression that …

I fail to see how pursuing recursive self-improvement at full speed is compatible with Anthropic's stated goal of AI Safety. If nukes were not invented yet, would it really be a go…

For values that don't have a natural merge function (or where you don't want to bother writing one), would it make sense to sync update logs instead? That is: - The synced value is…

A CRDT that operates on code units should work out okay, because each grapheme cluster will always be inserted and deleted in a single edit - hence it should stick together in the …

I'm surprised to see the emphasis on tracking lines of text, which ties in to the complexity of merge vs merge-the-other-way vs rebase. If we are committed to enhancing the change …

You can think of the semantics (i.e., specification) of any CRDT as a function that inputs the operation history DAG and outputs the resulting user-facing state. However, algorithm…

While this is technically correct, folks discussing CRDTs in the context of text editing are typically thinking of a fairly specific family of algorithms, in which each character (…

Your part 1 post was one of the inspirations for that :) Specifically, it inspired the question: how can one let programmers customize the way edits are processed, to avoid e.g. th…

The rebasing step is indeed a transformation. Some info in the "rebasing" link here [1]. Unlike traditional Operational Transformation, though, there are no "transformation propert…

The PowerSync folks and I worked on a different approach to ProseMirror collaboration here: https://www.powersync.com/blog/collaborative-text-editing-ov... It is neither CRDT nor O…

This was my impression as well. If you ignore the paper and just look at the source code - and carefully study Seph Gentle's Yjs-like RGA implementation [1] - I believe you find th…

Managing "a flat-ish collection of nodes" that can be moved around (without merely deleting and re-inserting nodes) is tricky because of how paragraphs can be split and merged. Not…

> More specifically, if you can edit different parts of a same document on different devices, then the document should be split across multiple files that can be synced independent…

Indeed, Replicache works this way, using server reconciliation (one part of client-side prediction): https://doc.replicache.dev/concepts/how-it-works

This sounds similar to the idea behind articulated (though with ids UUID-counter instead of time-counter): https://github.com/mweidner037/articulated I will check out Antirez.

A decentralized, eventually consistent total order on operations is a fully general CRDT, in the sense that you can put whatever (deterministic) operations you want in the total or…

As the author, same. My best guess is: - Central-server collaborative editing work focuses on Operational Transformation (OT), likely due to inertia (studied since 1989) and the pe…

Even in the absence of a central server, you can still avoid CRDT/OT complexity if you have a decentralized way to eventually total order operations & apply them in that order: htt…

Indeed, this is close to what Yjs (popular CRDT library) does: each client instance (~ browser tab) chooses a random 32-bit clientId, and character IDs combine this clientId with l…

I believe that Automerge internally stores all operations in an eventually consistent total order, which you can use as a substitute for the server in server reconciliation (cf. ht…

Indeed, the simple approach of "send your operations to the server and it will apply them in the order it receives them" gives you good-enough conflict resolution in many cases. It…

You could avoid the CRDT rules if you only use the LLM on the server. I.e., user comes online and sends their diff to the server, which LLM-merges it into the latest state and then…

One challenge is that the algorithms typically used for collaborative text editing (CRDTs and OT) have strict algebraic requirements for what the edit operations do & how they inte…