We already do this in the CLI interface, so it should be just fine.
It worked with the GNU version, but not with the BSD version.
They catch any exception, including KeyboardInterrupt, we don't want to catch it.
Conflicts: youtube_dl/extractor/__init__.py
For consistency and readability
1) added comment to explain extra xml link download 2) changed {} entries to {0}, {1} etc 3) removed redundant language header (the others are required) 4) checked out the old version of the supported sites md (the change was not required)
Urls can also contain contain a query (for example a timestamp '?t=foo')
Generic extractor is just fine for Libsyn embeds