Notices by John Regan (jprjr@mastodon.social)

@dalias I'm having a hard time thinking of an activity more degenerate than "gambling on nuclear weapon detonations"

@gamingonlinux it's pretty obvious this law was written by people who only use single-user phones and tablets.

Also, and I cannot stress this enough - very few tech-savvy people get into politics. I've been asked some incredibly bananas tech questions by politicians before. They're very out of touch. This law being a prime example.

If you live in California you need to call your local reps and tell them why this is a bad idea.

@dalias @neil don't even get me started on those Google AI snippets. Recently had an experience where a coworker told me a bunch of nonsense.

"Where did you get that? None of that makes sense?"

The answer was that stupid Google AI snippet. Pure garbage, plus it's very insulting to get AI nonsense quoted at you.

@dalias @neil I think the intersection of Firefox users who specifically want AI is pretty small.

I will say the amount of people I see using AI, generally, is increasing. I know some people who use ChatGPT instead of a search engine. Very non-tech bro people.

I think there's more usage than just AI bros, but I think most Firefox users choose Firefox for privacy and better ad blocking - the Venn diagram of Firefox and AI is the thinnest possible sliver.

@whitequark @david_chisnall > - I always had clean SPF but never bothered with DKIM or DMARC

This is something I see really often, and I think nowadays DKIM and DMARC are more of a requirement.

I see some organizations implement DMARC reject policies with SPF as their only mechanism. I don't think people realize how many mail forwarding services are out there that will always break SPF, DKIM is really crucial.

@ska @astraleureka @dalias I will say, one thing that's nice is being able to just send the bare minimum http 1.0 GET request, parse a few headers, and from that point just give bytes of data to your audio decoder.

But on the other hand it's $CURRENT_YEAR, everybody's using some kind of http library that abstracts things away anyway so chunking isn't hard to do.

I think the next icecast release is planning on supporting chunking but that's been in development for years.

@ska @astraleureka @dalias looking at the dates of the first Shoutcast release (1998) - I think it's conceivable they chose to not do http chunking - either due to the RFC still being written at the time of development, or even if it was done, it was probably still pretty new.

@dalias @ska @astraleureka one case where buffering the whole response and computing the content-length wouldn't work is Internet radio streams (icecast, shoutcast, etc). They just stream data forever.

Granted I'm not sure any folks out there are going to implement such streams as CGI/NPH scripts so, probably a moot point.

@lzg you also watching Christmas Mail?

@gabek yeah, our Patreon funding fell below our monthly server costs so, we decided to retire the video portion of the radio. It's still around as an Icecast-based, audio-only stream.