@Edent I’m delighted to see you added the move support. Do you think doing the reverse is also feasible, so if the experiment fails it doesn’t become the Hotel California?

@thisismissem @LucasWerkmeister @Em0nM4stodon Okay, thanks!

FWIW, I landed on a different w3.org page (https://www.w3.org/wiki/ActivityPub/Primer#Client_to_Server_API) that was less than helpful.

That said, the appeal of a Mastodon-like API here is that it keeps things a lot simpler than the can of worms that I understand ld+json to be (and I immediately note the sparse “Uploading Media” section, which will slow down the “Hourly Panda Picture" type bots).

@thisismissem @LucasWerkmeister @Em0nM4stodon Is there a good link for the AP C2S protocol? Not the most helpful terms to throw at a search engine, unfortunately. (One of the hits I got was this https://socialhub.activitypub.rocks/t/activitypub-client-to-server-faq/1941 2021 page that advocated against it but that feels like forever-ago in many ways)

@LucasWerkmeister @Em0nM4stodon
Related: @thisismissem mentioned @fedify (https://hachyderm.io/@thisismissem/113393232532833193) but I have no idea what the “gap" is between that framework and having something that supports random people being able to create new bot accounts and getting up and running posting content from those bots.

@LucasWerkmeister @Em0nM4stodon was wondering the same, but what would be the "requirements”?

- a fully functional web front end?
- support migrating accounts (in and out)?
- API... compatible with Mastodon, or nah?

I have a feeling that you'd pretty soon end up with a (stripped down) Mastodon to make it useable/attractive destination for (already existing) bots

I was very briefly mulling over snac2 for instance but IIRC migration is still in beta; API? stats? No idea.

@thisismissem I don't know if this is too US/$-centric and maybe I've already mentioned it before, but https://cushionapp.com/ tries to help with that as well. Originally put together by @destroytoday, and (as far as I know) still maintained by him and a few others.

@http Have you heard of @timbray's "topfew” which scratches a very similar itch?

Topfew Fun

ongoing by Tim Bray (www.tbray.org)

More Topfew Fun

ongoing by Tim Bray (www.tbray.org)

My problem is that while I know about it, my muscle memory is too strong and it isn't necessarily installed in environments where my shell scripts may land, so that (muscle memory) isn't likely to change any time soon.

Oh, and the top and unique artists in 2w

$ jq -c 'select(.type=="Song")|[.desc]' .history | sort | uniq -c | sort -nr | head -n10
193 ["Queen"]
178 ["Led Zeppelin"]
157 ["Aerosmith"]
151 ["AC/DC"]
143 ["Journey"]
115 ["Foreigner"]
114 ["Pink Floyd"]
106 ["Van Halen"]
103 ["Tom Petty"]
98 ["Boston"]

$ jq -c 'select(.type=="Song")|[.desc]' .history | sort | uniq | wc -l
120

#radio #jq #shell #bash #scripting

There's a local “classic rock” station KFOX, usually on during a commute. I wondered how repetitive it actually was and (ab)used their "Now Playing" website widget to track songs/ads/promos.

- Backend data is a JSON blob (1st pic)
- Add a small shell script (2nd pic), run for 2 weeks
- Top songs in the 3rd pic
- Unique ones over 2 wk:

$ jq -c 'select(.type=="Song")|[.desc,.title]' .history|sort|uniq -c|wc -l
372

- Length of the tracks (see 4th image): 27h!

#radio #jq #shell #bash #scripting