Maksim Sadym
Maksim Sadym
This basically means that [`ThreadDebuggerCommonImpl::serializeToWebDriverValue`](https://source.chromium.org/chromium/chromium/src/+/main:third_party/blink/renderer/core/inspector/thread_debugger_common_impl.cc;drc=04159b02a986bdd73d461370aeb40dc7dbb576b3;l=301) has to check the `value` vs each of the [custom blink objects](https://source.chromium.org/chromium/chromium/src/+/main:out/Debug/gen/third_party/blink/renderer/bindings/core/v8/). There probably some ways to cut this corner.
After [Serialize dom wrappers (4303236) · Gerrit Code Review](https://crrev.com/c/4303236) is merged, Mapper can simply replace type `platformobject` with `object`. At least until specification is changed ([Should we serialize platform objects...
TODO: Verify after #555 the message is sent to Mapper Server
We need to GC after all the helper calls using [`Runtime.releaseObject`](https://chromedevtools.github.io/devtools-protocol/tot/Runtime/#method-releaseObject).
The repro script in the description does not reproduce problem, however this one does. The difference with the original one is in fact it does not set a viewport, and...
Actually the [`await this.#scrollIntoViewIfNeeded()`](https://github.com/puppeteer/puppeteer/blob/070ee03d31318afc2054b2cfd719372fcfd4cab7/packages/puppeteer-core/src/common/ElementHandle.ts#L737) is not needed in case of `captureBeyondViewport: true`.
Returning all the capabilities were used sounds reasonable, even though it creates a bit more traffic.
> @sadym-chromium could you maybe explain what is/was broken with the test? sure! 1. The test `test_release_mouse_sequence_resets_dblclick_state` checks both with and without release, which is confusing. 2. Moving it to...
> @sadym-chromium could you maybe explain what is/was broken with the test? 3. If there is no `release` actions, why the `dblclick` should be absent at all?