Thanks @jdescottes. Lgtm. The fix covers the issue.

In D275196#9595110, @nchevobbe wrote:

In D275196#9595095, @nchevobbe wrote:

I may be missing something, but if apply D259239: Bug 1960549 - [devtools] Add resource watcher for webtransport events r=#devtools (and its parent), but not this patch, I'm still able to see the the webtransport request.
So I'm wondering if this patch is necessary, since it only seem to be some preparation work for showing the datagram later?

Actually even with only D276447: Bug 2006181 - Set the browsing context id for the webtransport channel r=#necko-reviewers, it already see the request. Is there any reason to not merge this patch with D259239: Bug 1960549 - [devtools] Add resource watcher for webtransport events r=#devtools, since they both seem to be preliminary work that don't have impact on the netmonitor for now?

@nchevobbe The summary hints to another patch on the stack. Is that no longer the case anymore? is this now a standalone patch?

Thanks @nchevobbe for the work to use a better api.
LGTM!

lgtm! Thanks

Thanks for the reviews Nicholas!

In D275196#9575432, @nchevobbe wrote:

Thanks bomsy
I'm not sure this is ready? Maybe I'm missing something but we're adding empty listener for the web transport events, so I'm not sure what the patch is doing.
I ran the example from https://github.com/BiagioFesta/wtransport, and while I do see the https://localhost:60344/ URL, there's nothing in the "Response" panel as I'm sending data (whatever the type is).

The only goal of this patch is to see the request in the netmonitor. The response panel is going to be empty because we do not receive datagram packets yet (That is future work planned)
The empty event listeners are going to useful when we start showing datagram packets in the response panel. For example the webTransportSessionClosed will be used to show when the connection is closed in the Response panel.

Also, the type column says "plain" for the 60344 request, should it says "webtransport" ?

That is the same thing we show for websocket requests, in the future we will look into improving that. (I’ll file something for that)

Also this doesn't have a test for the netmonitor, is this because the limitation of mochitests regarding TLS ? (I think you mentioned this a few times)
In such case it would be nice to mention this limitation in the commit message, and have a bug somewhere to add a mochitest, blocking it on adding support for TLS in mochitests

Yes this is because of the limitation of mochitests, i'll detail it in the description and link to the bug.