Consider the following command with GNU-style options shown:

yt-dlp --restrict-filenames --windows-filenames --mtime --write-auto-subs --embed-subs --embed-metadata --embed-chapters --xattrs --no-remove-chapters

Assuming you have the program yt-dlp installed, that invocation with the URL of a video at the end will grab the video, restrict the resulting filename from having incompatible characters, structure the filename to Windows conventions, set the file-modification time to the last modified header’s time, write and embed automatically generated subtitles if they’re available, embed metadata, embed chapters, and write metadata to extended filesystem attributes if it can do so in your operating system. Some of those are default behaviors of the program but it does help me to have it explicitly written out for when I need to start modifying snippets. Sometimes I can’t recall the logic of a snippet and I can find things in a manual page faster that way.

Now consider the following command snippet:

yt-dlp --restrict-filenames --windows-filenames --mtime --write-auto-subs --embed-subs --embed-metadata --embed-chapters --xattrs --no-remove-chapters "https://archive.org/details/IBMPerso1987"

Assuming nothing went wrong that would download for you an episode of the Computer Chronicles related to the IBM PS/2. It is a program hosted by Stewart Cheifet.

A screen grab of the episode of Computer Chronicles referenced above showing Stewart Cheifet during the Random Access segment talking about Carnegie Mellon University.

Normally people think of peoples like yt-dlp for grabbing things solely from YouTube. If you list its extractors you find that you can grab from quite a few other places.

As things continue to fall apart around us, now would be a good time to archive things more. Born digital content is ephemeral. Such things seem to have better chances of long-term survival being transferred to film at the rate changes in the legal environment are taking place.