Dave

Working on getting the web-ui repo upgraded because node.js blows worse than php. This day won’t end.

Security people are unbearable. Every time there is a big urgent security issue they do a pile on and try to make every other thing just as urgent.

RE: https://podcastindex.social/@dave/117480517086273711

Fixed.

RE: https://discuss.systems/@ricci/117480591354709019

On the same day as the global PHP panic? That's like someone choosing to release their indie film on the same day as Ghostbusters.

#api After upgrading php-fpm on the api front-end servers this morning I just noticed the fpm default config was in place so the api is running like a dawg. Restoring the pre-tuned settings now...

#api Podcast index api servers are patched and serving data again.

RE: https://podcastindex.social/@dave/117479258728316180

Patching underway. Fingers crossed...

RE: https://podcastindex.social/@steven/117475936422223299

#api Heads up, if Podcast Index api goes down at 9:00 am central time you'll know why. Will know more when the announcement hits.

#api Just pushed a new change that will hopefully auto-merge all of those duplicate wordpress podcast feeds across the Index. So far it looks promising. It will take a few days to fully sweep the whole Index.

RE: https://podcastindex.social/@mitch/117458083323416290

This was my point! Give the listener control of the entire experience. Let their agents customize the whole experience.

Modern tech vc used open source for decades to make obscene piles of cash and barely giving anything back. LLM coding is going to flip that script. The very thing that enabled tech megacorps (FOSS) will be the thing that allows everyone to easily escape paying them any more money.

Worried about data centers being built in your area? They will probably never be built:

"Gartner Forecasts 80% Of Neocloud Market Will Cease To Exist By 2030"

https://finance.yahoo.com/technology/ai/articles/crwv-nbis-iren-stocks-tumble-181012372.html

This company (LiquidStudios/InceptionPoint/QuietPlease) makes no sense:

https://podcastindex.org/podcast/8072852

They change their story every few months.

It never ends:

https://podcasts.apple.com/us/podcast/a-deep-dive-into-pescos-online-billing-services/id6797827135

Makes great sense:

"How podcast:person should handle characters in fiction and narrative audio?"

https://github.com/Podcastindex-org/podcast-namespace/discussions/498#discussioncomment-18772440

Prediction: By the end of 2027, one "tier 2" podcast app (meaning Overcast, Podcast Addict, Castro, etc) will put a "Skip Dynamic Ads" checkbox on their settings screen, which will remove DAI slots from podcast episodes on-device after download. And, as soon as one of them does it, they will all do it.

Spam:

https://podcasts.apple.com/us/podcast/imia-podcast/id1833151497

RE: https://podcastindex.social/@martin/117404457764518454

Congrats Martin! I'll give it a whirl!

FYI, linode did a host migration (reboot) on the utility server right in the middle of the weekly full db dump. [sigh]

Podcast Index terms of service updated (finally):

https://api.podcastindex.org/tos_v2.0.html

The old v1.x terms were 90% boilerplate legalese that never accurately reflected what we wanted. I get emails weekly asking for clarification.

Here is the github repo for the terms: https://github.com/Podcastindex-org/terms-of-service

The PI terms of service are an open project just like everything else. If you think something can be worded better, or there's a way to make things clearer just open a PR.

Man, I'm still sort of reeling.

A glimpse of the end.

Has anyone used Cloudflare R2 extensively enough to give advice on whether it's superior to Linode object storage cost-wise and ease of use wise?

Fun new problem: restricting un-metered outbound bandwidth from clients using direct linode object storage URLs instead of our published Cloudflare cached versions of those URLs. That accounted for 40TB of bandwidth last month.

The sysadmin fun never stops.

Podcast Index had 65 terabytes of outbound bandwidth this month. A record.

I'm going to start placing a schema "legend" in the NDJSON database export bucket. It should detail all the colums and table relationships. Just FYI.

The latest Better Offline episode has 9.5 minutes of ads in a 17 minute episode.

4.5 minutes of pre-rolls and 5 minutes of post-rolls.

All pharma ads that don't apply to me.

But, hey, at least it's the same identical ads in the pre-roll and post-roll for weeks on end so that's good.

https://mouse.dev/blog/muse-runtime-export/

"I asked Muse to archive the files it could see and send them to my Google Drive. It did.

The download was about 2.7 GB compressed and 6.8 GB unpacked. It appeared to contain the root filesystem of the Linux environment assigned to my session, including Ubuntu system files, Muse’s internal documentation, integration code, app templates, memory files, and agent logs. There were also SSH key files."

#api I fixed a huge problem with the feed aggregators. Based on this snapshot of our database metrics, can you guess at roughly what time that happened?

#api There was a problem with search ranking in the search/byterm endpoint after switching the db schema. This should be fixed now. Will explain in the board meeting tomorrow.

#api All out brawl with the feed aggregators this morning. I'm about to win though.

#api Adding 2 new indexes on the feeds table. Expect the api to choke for a few minutes until this completes.