Cool, that's an interesting way to apply parser combinators. Where can I download the library? Also, the Sample Code link at the end is 404.

Sorry about that, broken links have been fixed in the episode and here's a link to the library: https://github.com/pointfreeco/swift-parsing.

Speaking of broken links, you're also returning 404 for HEAD requests:

  $ youtube-dl https://www.pointfree.co/episodes/ep185-tour-of-parser-printers-introduction -f http-720p
  [generic] ep185-tour-of-parser-printers-introduction: Requesting header
  [snip]
  zlib.error: Error -5 while decompressing data: incomplete or truncated stream
  $ wget --method=HEAD https://www.pointfree.co/episodes/ep185-tour-of-parser-printers-introduction
  [snip]
  HTTP request sent, awaiting response... 404 Not Found

(Fixed youtube-dl by diking out head_response (with `= False`) in youtube_dl/extractor/generic.py:2408 if anyone else cares.)

Making Mondays something to look forward to! I already can use this library for a new project.

This post links to a video where at least one presenter (the one on the right) reads directly from a prepared text while trying to vary his delivery in a way that's supposed to make it sound like it's natural conversation or, say, a lecture which is prepared but where no verbatim script exists. A word of advice (at least if you're trying to maximize your reach*): don't do this.

I listen to at least one podcast where the host does this, too, and it's tolerable for an occasional listen but extremely grating. I can only imagine that this happens because no one has ever told them "don't do this; it is extremely grating", so they think it really is achieving some sort of polished, Ira Glass/NPR effect. It doesn't, and I imagine further that for lots of people it's actually intolerable instead of being merely ineffective, which means those listeners opt out instead. (Even in the case of the podcast I mentioned, I'm less motivated to seek out that show, so I end up listening to far fewer episodes than otherwise.)

To better understand why I'm commenting about this at all, read <https://pointersgonewild.com/2019/11/02/they-might-never-tel...>.

* this looks like a free episode in a for-pay series, so it seems reasonable to assume that this applies here

PS: The project that is the focus of this presentation (swift-parser) is not linked anywhere. The yowconference.com link for "Unified Parsing and Printing with Prisms" is broken. The link for the code sample on GitHub is also broken.

Wow, now that you mentioned it I won't be able to stop noticing it. I wonder what a better approach is? Some Youtube videos like 3blue1brown's content kind of have to be scripted, but you'd ideally still like the presentation to sound natural. Maybe they should have a script, rehearse it once, then do the real take without the script?

I think the key is: don't have a script per se, have a list of talking points, and be willing to re-record sections as appropriate.

That said, one of my favorite YouTubers (CGPGrey) does all of his videos from verbatim scripts. But he also records his voice overs multiple times in order to get a good take.

There is a link to the library in the first sentence of the transcript.

I know Pointfree do script their episodes but I can’t say I’ve ever noticed it in any detrimental way.

> There is a link to the library in the first sentence of the transcript.

That wasn't the case before, just as it wasn't the case that the broken-but-now-fixed links were fixed.

> I know Pointfree do script their episodes but I can’t say I’ve ever noticed it in any detrimental way.

Not really relevant to what I wrote.