crdt-benchmarks VS peritext

Author of Yjs here. I'm all for faster data structures. But only benchmarking one dimension looks quite fishy to me. A CRDT needs to be adequate at multiple dimensions. At least you should describe the tradeoffs in your article.

The time to insert characters is the least interesting property of a CRDT. It doesn't matter to the user whether a character is inserted within .1ms or .000000001ms. No human can type that fast.

It would be much more interesting to benchmark the time it takes to load a document containing X operations. Yjs & Yrs are pretty performant and conservative on memory here because they don't have to build an index (it's a tradeoff that we took consciously).

When benchmarking it is important to measure the right things and interpret the results somehow so that you can give recommendations when to use your algorithm / implementation. Some things can't be fast/low enough (e.g. time to load a document, time to apply updates, memory consumption, ..) other things only need to be adequate (e.g. time to insert a character into a document).

Unfortunately, a lot of academic papers set a bad trend of only measuring one dimension. Yeah, it's really easy to succeed in one dimension (e.g. memory or insertion-time) and it is very nice click-bait. But that doesn't make your CRDT a viable option in practice.

I maintain a set of benchmarks that tests multiple dimensions [1]. I'd love to receive a PR from you.

[1]: https://github.com/dmonad/crdt-benchmarks

Diamond types author here! Congratulations on getting your crdt working! It’s lovely to see a new generation of CRDTs which have decent performance.

And nice stuff implementing peritext! I’d love to do the same in diamond types at some point. You beat me to it!

Im building a little repository of real world collaborative editing traces to use when benchmarking, comparing and optimising text based CRDTs[1]. The automerge-perf editing trace isn’t enough on its own. And we’re increasingly converging on a format for multi user concurrent editing traces too[2]. It’d be great to add some rich text editing traces in the mix if you’re interested in recording something, so we can also compare how peritext performs in different systems.

Anyway, welcome to the community! Love to have more implementations around!

https://github.com/josephg/crdt-benchmarks

https://github.com/dmonad/crdt-benchmarks/issues/20

So yes, a particularly large document is not the norm but it can happen.

JavaScript CRDTs can be quite performant, see the Yjs benchmarks: https://github.com/dmonad/crdt-benchmarks

This doesn’t appear to support rich text formatting ranges like bold, italic, etc - unless I’m missing something in the API. AFAIK Peritext is still the state of the art in rich text CRDT algorithms https://www.inkandswitch.com/peritext/

I’d love to see this build the rich text stuff from the Peritext algorithm.

The work Ink & Switch (unaffiliated) do has been an inspiration to my with regard to local-first and decentralized software: https://www.inkandswitch.com

They have a quasi-manifesto on local-first (https://www.inkandswitch.com/local-first/) and have published the best rich text CRDT around, Peritext: https://www.inkandswitch.com/peritext/

Lots of interesting work happening in this space.

Take a look at https://automerge.org/ and the stack those folks are building. You're exactly right that it's a difficult balance (specifically the trick is proving commutativity for the domain-specific data of your application). But automerge (and then https://github.com/inkandswitch/peritext) show it's at least possible. Good stuff.

Yes - The BFT problem only matters when you have Byzantine actors. But I think users deserve and expect the system to be reasonably well behaved and predictable in all situations. Anything publically writable, for example, needs BFT resilience. Or any video game.

As for the prosemirror problem, I assume you’re talking about weird merges from users putting markdown in a text crdt? You’re totally right - this is a problem. Text CRDTs treat documents as a simple sequence of characters. And that confuses a lot of structured formats. For example, if two users concurrently bold the same word, the system should see that users agree that it should be bolded. But if that “bold” intent is translated into “insert double asterisks here and here”, you end up with 4 asterisks before and after the text, and that confused markdown parsers. The problem is that a text crdt doesn’t understand markdown.

JSON editing has similar problems. I’ve heard of plenty of people over the years putting json text into a text crdt, only to find that when concurrent edits happen, the json grows parse errors. Eg if two users concurrently insert “a” and “b” into an empty list. The result is [“a””b”] which can’t be parsed.

The answer to both of these problems is to use CRDTs which understand the shape of your data structure. Eg, use a json OT/crdt system for json data (like sharedb or automerge). Likewise, if the user is editing rich text in prosemirror then you want a rich text crdt like peritext. Rich text CRDTs add the concept of annotations - so if two users bold overlapping regions of text, the crdt understands that the result should be that the entire region is bolded. And that can be translated back to markdown if you want.

The ink & switch people did a great write up of how this sort of crdt works here: https://www.inkandswitch.com/peritext/

> I'm looking out for practical CRDT ideas that works well with richtext.

Have you seen Peritext from Ink & Switch? https://www.inkandswitch.com/peritext/ It's relatively new, but is a CRDT aimed at rich text!

To put it in a different perspective, plain text editing has well-solved CRDT patterns. But, semantic data-structures like rich-text or syntax trees is what's tricky and has unsolved challenges.

Peritext[1] is the only one that came close to solving rich-text, but even that one left out important aspect of rich-text editing like handling list & table operations as "work to be done later".

For people interested on why it's difficult to build CRDTs for richtext, here's a piece I wrote a year back: https://writer.zohopublic.com/writer/published/grcwy5c699d67...

Related HN discussion: https://news.ycombinator.com/item?id=29433896

[1] https://github.com/inkandswitch/peritext

Anyone unsure of what a CRDT is, this is the perfect intro: https://www.inkandswitch.com/peritext/

The two most widely used CRDT implementations (combining JSON like general purpose types and rich text editing types) are:

- Automerge https://github.com/automerge/automerge

- Yjs https://github.com/yjs/yjs

Svelte is, but that is your smallest problem. You want to look into CRDTs (conflict-free replicated data types) to offer true (offline) collaboration. A popular JS library to solve this complex problem is called [automerge](Conflict-free replicated data type). A rather recent development in that area specifically for text-based content is Peritext. Also check out this interactive tutorial about CRDTs.