Tldraw's $2.7M Seed – A new collaborative infinite canvas for the web | Hacker News Reader

Tldraw's $2.7M Seed – A new collaborative infinite canvas for the web

tldraw.substack.com

154 pointssamwillis39 comments

What Steve is building is incredible, but almost more so how he is sharing the journey. He is a very good follow on Twitter, lots of lovely content showing how he is solving all sorts of interesting problems.

https://twitter.com/steveruizok

You can try the new beta here: https://beta.tldraw.com

Can you explain how this is incredible with relation to other products on the market? This seems very similar to various other drawing apps on the market. What's the unique value proposition here?

From what I've seen, the real-time performance is insane. Dozens of folks on one board with instant updates.

Lagged for me on Firefox the more I zoomed out and thus the more elements were on screen.

the uniqueness is that it's the highest quality drawing app on the internet. even if on paper other products say they do the same thing, there's a huge difference in quality.

It’s easy to discount a drawing app because it’s been done so many times before. But simply put, this one is just better than the rest. The UI interaction has been thoughtfully refined and never gets in your way. Over time, it’s become the tool I reach for regularly to create sketches. In retrospect, it’s surprising it hasn’t been done this well before!

Hey! Steve here, founder etc. The biggest value prop is having a canvas that’s “web native”, and so can render anything that a regular website can. Try pasting a codepen link, scratch project, google maps url, gif or video. Almost all of the developers who built on tldraw’s first release ended up building apps that leveraged this; for example, a spatial collection of graphs keeping track of aws cloud processes. https://twitter.com/foodybrian/status/1572784516586172418?s=...

There are trade offs for using the DOM vs. a custom <canvas> renderer, however I think they’re worth it for the types of applications that can be built this way.

We’ve also put a ton of thought (and opinions) into the interactions in tldraw. Some of this can sound pretty ridiculous, like rotating your cursor when your rotate a selection, while also counter-rotating the cursor’s shadow to preserve light direction. But especially with this sort of app, you really notice those details when they’re not there. As a product for developers, my hope is that the quality and attention to detail become part of the value proposition! Users certainly notice it.

Hey, very cool to have you chime in here Steve. After looking again, it seems I missed some of the cooler parts of what you've built here. (That tweet is a pretty cool demo video of some of the things you could make).

Also—this answer was exactly what I was hoping to be shown. It sounded like tldraw was capable of a lot more than drawing boxes with arrows that moved, but I hadn't figured out exactly what that was--now it all makes sense.

I went there and the first thing I saw was asking for HN upvotes

"we're on hacker news, no direct links but upvotes appreciated"

https://twitter.com/steveruizok/status/1598381851055300608

Honestly why is there a stigma against asking for upvotes? He isn't being pushy about it either.

Is everyone just selling their college senior projects to vc's for millions now

How would this be monetized by the developer?

ads could be strewn across the whiteboard, having to be overwritten or erased, being consumed by the user in the process.

Might want to pivot and mint NFTs from the all the drawing. Oh wait, but the VCs are on to AI, which means...

Yeah, brands would just _love_ that.

Maybe a subscription for enterprises using it?

You'd charge people for a version they can embed as a component in their own app.

Who is the actual market for that kind of integration?

Some opportunities that have come up so far: licensing the engine to commercial products, selling the related services like multiplayer, hosting the experience as an embed, integrations with other services (eg learning management systems), building out the dot com experience into its own a sass product for users / teams, building our own products for other verticals such as pdf annotation or ml image annotation, or door to door sales of nice looking arrows.

This is clearly self promotion…and the multiple alt accounts / friend accounts to try to create a particular impression of the software…

The way this is worded made it sound like a revolutionary MMORPG geo-based canvas which was a rip of the whole world or something.

Its just a half decent drawing tool written in webjs library of of the week#52 why would anyone pay 2.7mil to seed this????? I have a scratch game to sell them????

Anyone who tried it, how does it compare to Excalidraw?

They seem to be clones of each other. Not sure which came first.

I heard of Excalidraw first. And a quick browse of GitHub leads to me believe that Excalidraw was first.

I've only used Excalidraw, and it's my go-to. But I follow Steve on Twitter as it's fun to see what's he is excited about.

I actually contributed to Excalidraw, it’s a great project! They also use my ink (MIT licensed perfect-freehand library) for their pen tool. While tldraw mostly grew out of another project of mine (globs.design), I took a lot of inspiration from Excalidraw’s feature set. We’ve since caught up with eachother on a bunch of features (sticky notes, labels) and hopefully will continue to trade tips and features as separate open source projects.

Excalidraw is one of my favorite tools, and for diagramming it's much better. I'm also a happy subscriber to their private collaborative workspaces features.

With that said, I look forward to tldraw's library getting open sourced, and discovering what's possible to build with it.

I have an old wacom tabler from 2013 and love that the eraser works here. Google's Jamboard has been so frustrating because of that missing feature.

Clever name! (tl;dr aw)

I wonder how they make money now and if there is a tension there with the free software.

Makes me think of https://drawball.io

(Disclaimer: A colleague’s hobby project)

Why did you build a canvas manipulation tool using React as the framework? Are you concerned about reconciliation performance?

I think it's like Figma or more concretely react-three-fiber, which bypasses the React renderer for its own high performance one. React is used simply for the DX and for everything else in the UI I imagine, as I like writing in components rather than imperative DOM updates.

Looks like that's not correct. If you open your browser console with the React developer tools extension installed, you can see that not only is the canvas built with the DOM, but it's rendered with React, too.

That’s right, it’s React all the way down. It could render using anything, really; the React rendering part is quite loosely coupled from the main “what needs to be rendered” application layer.

That's actually super interesting. I would love to read a deeper dive on how you keep it so smooth, if that's something you'd be willing to share.

wow good for them. love to use this in a pinch

We use tldraw at LegendKeeper. Our users absolutely LOVE IT. Almost too much; our users are D&D players so they like to make giant family trees and faction relation diagrams. Tldraw handles it perfectly, though.

LegendKeeper looks like a great product too!

Awesome to see Steve's hard work taking off. TLDraw feels so good to use and I love watching him obsess over the little details on twitter.

this is dope af.

Reply on news.ycombinator.com