Steven Luscher

Sorry about this. The _real_ problem here is that we export _way_ too much stuff from web3.js, and the _only_ reason we export a type like this that's fundamentally internal...

Quite possibly related: #27859.

Where you do see this discussion ending up? Right now, we have [a mapping in one of our example apps for knockout.meteor](https://github.com/steveluscher/knockout.meteor/blob/master/examples/todo_list/example.js#L21-36) that (once making its way through the plugin)...

I think I figured it out @RoyJacobs. In making [this fiddle](http://jsfiddle.net/M4SKb/1/), I learned that you have to move the "root" mappings into a property named `""` if you want them...

There are a few strategies to roll a change to return types incrementally that we could use here. #### 1. Rename, deprecate, then remove We could deliver `VersionedMessage` against a...

Can you try one of two things: 1. Move `react-native` to being a `peerDependency` instead of a `dependency` in this library. 2. Maybe make a change to the browser field...

> …dependencies should be kept as minimal as possible to reduce bundle size… The `react-native` dependency here does not increase bundle size; we build one bundle for browsers, and a...

> Proposed Solution: Remove the `react-native-url-polyfill` This polyfill was added to make `web3.js` compatible with bonafide React Native projects running in JSC/Hermes. This is so that the usage of `new...

Neat! Can you paste in a quick screencast of this feature? Command-Shift-5 on Mac to take a screen recording.

Say more, @timaiv. I actually haven’t looked in to what v0 transactions demand of web3.js.