With Legacy Connection also is not possible to connect for me.
I tested in Firefox v 119.0 (64 bits) and Chrome 119.0.6045.106. Boths have the last versions. Is not possibel to connect dwith this version.
I am using Firefox 119 and am not encountering any issues. If even after upgrading to the latest Firefox you are still having issues, please open the browser's developer console, right click in the console and select "Save all messages to file" and then PM the file to me.
If you have any browser extensions enabled, it would also be good to temporarily disable them (or use private browsing mode) to minimize the potential for interference while trying to debug the issue.
I haven't tried it yet today, but yesterday both the latest Chrome and the latest Firefox failed to connect. Not in Legancy mode either.
Apparently it works with some older versions. So they probably broke something in browsers again
I have tested now over Firefox 119 and at 20% the connection stop.
I have not Extensions in Firefox.
The Phater Process error is:
I send a PM with the Multiporcess Console Firefox mistakes
For folks encountering issues, are you all on Windows 7?
Edit: I do not experience this behavior on my Mac with Firefox 119 but do get the failure on Windows 7 running Firefox 115 as discussed above. Not sure where the issue lies exactly yet and what workarounds are possible.
I am on windows 10 and I face issues both on chrome and Firefox latest versions both
No issues with Mac or iPad,
But i experience issues with windows 10 and Linux distros
I am in Windows 11 v. 23H2 from 5.11.2023 and Firefox 119.0 (64 bits)
Widnows 10
We have root caused the problem and are working to rectify the issue.
Ok. We believe the issue has been resolved.
Tested and the connection is OK. The issue is solved.
May we know what caused it?
Great, confirmed for iOS and Android. Both work again (also) without Legacy.
The standard connection mode uses a technology called WebRTC to ideally negotiate a peer to peer connection between the website and the companion app, and it will try to do so in a way that minimizes the potential number of hops between the two. In the event that a direct pathway cannot be found, WebRTC allows for a server called the TURN server that can be used as method of last resort where the clients set up a relay that goes through the TURN server. Now for a project as large as MIT App Inventor a fair amount of people end up having to use the TURN server hosted at MIT (and possibly based on your other thread some downstream products are also using ours...), this results in a decent amount of traffic so the server was shut off by someone in the IT department without first confirming whether it was expected behavior. So all of the above was a wild goose chase as we hadn't been informed of the shutdown. Turning the server back on fixed the issue.
Interesting, thank you very much. It's just strange that other builders (like Niotron) were/are apparently also affected at the same time. Coincidence ...?