mweidner037

Thanks for looking into this! > There is no human that will ever notice the difference between 0.002ms and 0.015ms per keystroke. Indeed, I think time per op really only...

I'm also running into this issue - specifically, that video recording doesn't work in pages created with `brower.newContext`, as described [here](https://github.com/microsoft/playwright/issues/9002#issuecomment-922187532). My use case: I'm using `storageState` for authentication, as...

I ran into this issue as well; here is a minimal reproduction: https://github.com/mweidner037/playwright-test-mwe-newContext-page Basically, calling `browser.newContext({storageState: ...})` where the `storageState` file has an origin entry, causes a blank page to...

> hi @jamesopti I also did same kind of code to fix it temporarily . But this is not sufficient. Because apart from TextSelection and NodeSelection I also have CellSelection...

Yicheng Zhang, myself, and Heather Miller ran a programmer user study (https://doi.org/10.1184/R1/22277341.v1) where we observed a related issue: to make a "schema" `{"cat": {}}`, participants would start their app with...

I'm seeing a similar slowdown on `@automerge/[email protected]`: the [automerge-perf](https://github.com/automerge/automerge-perf) trace takes 130sec to apply, although I may be using the library wrong. My code: ```js const automerge = require("@automerge/automerge"); //...

@alexjg Some more info about the slow trace above: - If I simulate live collaboration by calling `getChanges` after every op, processing the trace becomes much slower: 327 sec for...

I am seeing the same behavior (certain icons repainting every 30 seconds). It stopped once I killed appimagelauncherd. System: Ubuntu 22.04.3 LTS, Wayland, Gnome 42.9 appimagelauncher version: 202211022349-stable-0f91801~ubuntu22.04.1 (latest from...

I believe [B3.4] is the first benchmark that uses the repo's `transact` function to wrap `applyUpdateV2` calls, as opposed to local CRDT operations. Here is how I implemented `transact` for...

> I believe [B3.4] is the first benchmark that uses the repo's transact function to wrap applyUpdateV2 calls, as opposed to local CRDT operations. Indeed, removing [B3.4]'s `transact` call (like...