dirkf

>Same problem, but it seems like the problem disappears after up to 2 days since the live stream has been finished (please mind it when trying to reproduce the issue)....

There are several examples of sites that have multiple instances with unrelated URLs/domains, and listing all the instances is a maintenance headache. In this case I'd rather avoid matching `r'https?://piped(?:\.[^.]+)+`...

Apparently the "legacy build" for browsers without dynamic `import()` doesn't work ...

https://github.com/TeamPiped/Piped As with Invidious, the presence of an OpenSearch `` with `title=Piped` can be used to identify the pages. However, as the pages are generated dynamically with JS, we can't...

>...we can't do much with them except redirect to the equivalent YT pages. Not quite true: apparently we could use [the API](https://piped-docs.kavin.rocks/docs/api-documentation/). Then we'd have to make a new extractor...

In a POSIX shell, `?` is a shell matching character that will normally be expanded relative to the current directory. It's unlikely that a URL containing `?` will match a...

>Looks like piped detection from ytdl just re-uses youtube servers instead of piped servers. It just grabs ?v= query after /watch & just appends it to youtube.com/. It's just a...

We can study https://piped-docs.kavin.rocks/docs/architecture/, etc. Each _Piped_ instance/site provides an opaque JS front-end to an open API implemented by the back-end. The front-end may direct requests to a different instance...

Agreed. And Newpipe Extractor is IIRC a Java program which imports a JavaScript interpreter module from the runtime library. Unfortunately this is one area where Java wins over Python.

At least if you invoked a node.js-alike you'd be passing it to the right language.