Mitch Downey :pci:

open source developer @ podverse.fm / fediverse / foss / javascript / node.js / podcasts / react / react native / typescript

Public Key

npub1yfr5d05qj3qzz84kntze3c8accyz0unwczgm0fxuypuqdnenkwuqew6g6v

NIP-05 Address

mitch@podcastindex-social.mostr.pub

Profile Code

nprofile1qqszy36xh6qfgsppr6mf43vcur7uvzp87fhvpydh5nwzq7qxeuem8wqpz4mhxue69uhhyetvv9ujuerfw36x7tnsw43q0qhfp5

Publishing to

relay.ditto.pub

Author Public Key

npub1yfr5d05qj3qzz84kntze3c8accyz0unwczgm0fxuypuqdnenkwuqew6g6v

Show more details

Last Notes

2026-03-19 05:33:00 UTC

- reply

@nprofile…zv63 @nprofile…zpdn @nprofile…cn7u this is also something we could add to Podverse Alpha fairly soon. The main blocker for me would just be a working livestream to dev and test against. Is there a 24/7 feed using this in the wild? Maybe too expensive for someone to run, but I imagine the use case could apply for displaying "Now Playing" info for 24/7 RSS radio/tv stations?
Also, the use case seems like it may be broader than "value" specifically. "Live Info Updates" or something.

2026-03-19 05:28:59 UTC

- reply

@nprofile…zv63 @nprofile…zpdn @nprofile…cn7u yes, I would like to support it.
re: a schema, a small suggestion I have would be, where possible, to follow patterns that align with RSS naming conventions?
channelTitle
itemTitle
Not sure what else this could apply to. Not strictly required. Just an idea.

2026-03-04 18:38:41 UTC

- reply

@nprofile…8zna interesting, I assumed this flip happened years ago. I barely know anyone who listens to AM radio anymore.

2026-02-21 17:55:44 UTC

- reply

@nprofile…zpdn I basically burned out on open source development from 2024 to early 2025...feeling much better now tho, and LLM advancements in the meantime have made things even more exciting and easier.
ActivityPub is still in the picture. I think there hasn't been much discussion lately around AP as a solution for boost metadata? @nprofile…lm4c could speak to AP much better than me. I'm still getting caught up.
I do think, broadly, LLM is making it easier to adopt robust standards like AP.

2026-02-07 00:36:33 UTC

@nprofile…8zna I was asking LLM some questions about value tag implementations, and look where it decided to go for source documentation!
https://cdn.masto.host/podcastindexsocial/media_attachments/files/116/026/542/933/631/282/original/216cbc49809353d7.png

2026-01-29 20:23:48 UTC

Podverse Alpha update 🛠️
We are nearly ready to implement V4V in Podverse Alpha web...and I am completely out-of-date in terms of what the current practices are.
Can anyone direct me to resources I should follow for a web-based Bitcoin V4V implementation? Is the standard still keysend? Are there multiple types of implementations we should support by now?
#podverse #pv2

2026-01-26 16:30:37 UTC

- reply

@nprofile…x5pc ok, I think I get it now. The Mostr bridge is silently mapping that npub it assigned to me to mitch@podcastindex.social in the API requests it makes with the fediverse. I imagine if we could see the Mostr API requests we would see our original handles in there along with the npub it makes visible in the message text.

2026-01-26 16:07:12 UTC

- reply

@nprofile…x5pc ah thanks for the info.
Do you know if that is a setting that the podcastindex social instance has enabled? Or does the mastodon software issue npub keys to every account created? I'm trying to make sense of how I'm being contacted in the fediverse through an address I didn't create (or don't remember creating).

2026-01-26 15:31:14 UTC

- reply

@nprofile…x5pc looking at hellthread website, I don’t think my Fediverse account even has a nostr npub id? Everyone else on the thread does though. Don’t understand how or where I’m tagged on this stuff.

2026-01-26 15:28:44 UTC

- reply

@nprofile…x5pc what’s weird is I don’t even see myself tagged in the threads I’m getting notified for. They haven’t happened again since last night so ig I’ll just ignore it for now.

2026-01-26 02:18:25 UTC

I'm getting notifications containing weird spam that makes no sense on here. Anyone else having that? How do you stop it?

2026-01-25 20:20:18 UTC

- reply

@nprofile…peg3 ah yes, well you are right that npm seems to have something similar to Rust in this respect.

2026-01-25 20:18:05 UTC

- reply

@nprofile…peg3 the pv packages it consumes (and whole infrastructure) will be bumped to 5.0.2, even if they don't have changes.
A major trade-off there is inefficiency, but from what I understand, we could add more robust tooling for monorepo handling later.
NPM has a feature called "workspaces" which appears to drastically simplify our module linking and dev process for a monorepo setup, vs our 10+ repo setup previously.
Anyway...it's a grand experiment lol but I think it's looking good so far

2026-01-25 20:15:20 UTC

- reply

@nprofile…peg3 thanks for the advice. After more research however, I've decided to go forward with the monorepo. I think the trade-offs of improved LLM context handling, easy of "quickstart" development, and simplification of open source contribution and CI are going to be worth it for our very large infrastructure.
One rule that is simplifying monorepo for us is currently we're planning on keeping all the packages and apps on the same version. Ex. if podverse-web updates from 5.0.1 to 5.0.2

2026-01-23 21:30:47 UTC

- reply

but I'm also thinking it may be easier for open source contributors to create PRs for Podverse if they can just create 1 PR per feature. Right now, a typical feature update will require changes to multiple repos. It's a clean process imo, but it does add significant overhead to PR creation and reviews.
Anyone have experience and opinions on monorepos? Worth the switch in our case, or could it create more problems than it is worth?

2026-01-23 21:28:29 UTC

As the Podverse infrastructure has grown to 10+ repos, and given how LLM seems to perform better with all its context in one place, and it would be easier to save llm prompt history in one place...I'm wondering if we should make a major switch to a monorepo 😣
I've never done a monorepo before. If it were just LLM, then we may not need it (we could create yet another repo...podverse-llm or something) (cont)

2026-01-22 18:43:29 UTC

- reply

@nprofile…cyex awesome. makes sense to me. thanks!

2026-01-21 07:12:53 UTC

- reply

@nprofile…cyex API is for the main http data router backend.
We also have workers that make requests on the backend from outside the API router app. Should we distinguish those from the router API?
Do you think specifying environments is not worthwhile?
Could you explain a bit more what you mean by distinct?
Haven’t started the mobile app/s yet but I’ll add at least as much info as the Spotify example.
Sorry, barrage of questions 😅 you’re definitely an expert on this.

2026-01-20 19:28:43 UTC

I have been thinking of using a user agent pattern like this for PV nextgen:
Podverse/Local/2/API
Podverse/Local/2/iOS
Podverse/Local/2/Android
Podverse/Alpha/2/API
Podverse/Alpha/2/iOS
Podverse/Alpha/2/Android
Podverse/Prod/2/API
Podverse/Prod/2/iOS
Podverse/Prod/2/Android
1st param = owner
2nd param = environment
3rd param = major version
4th param = device
Does that order look good, or does anyone recommend a better hierarchy?

2026-01-18 05:11:40 UTC

- reply

@nprofile…tsmm yeah that sounds right. The exploit in that case is limited to only that specific browser though, since the server has to grant the httponly status to that device. Without httponly, the JWT value could be extracted and set in a different client/device and used freely from there.

2026-01-17 16:14:59 UTC

- reply

@nprofile…tsmm no, a web browser can't set an httponly cookie. What the LLM did was modify the existing login endpoint in our router API so that it would return the JWT in a response body, instead of setting it as a cookie.
The browser then received the response body, and set it as a non httponly cookie in the browser.
And yes, httponly is needed on cookies to prevent XSS attack vectors. When httponly is true, browsers prevent the JavaScript runtime environment from being able to read it.

2026-01-17 13:32:11 UTC

- reply

This is an egregious and easily avoidable security vulnerability. JWTs should not be set client-side in a browser or they can be scraped by 3rd parties.
The actual solution was to simply add ‘withCredentials: true’ in the request helper’s options. If we were flying in blind without knowing that, it would have appeared like the LLM gave us a valid solution that worked, but in reality it was a critical mistake.
I think the models will probably get smart enough to avoid this particular error tho

2026-01-17 13:27:23 UTC

I was showing my brother how to do some agentic development, and it went great, except for one notable problem.
We were adding a Login page to a website. The auth endpoints were already written. It generated the form, and made the request to the endpoint, and logged-in…
But the HttpOnly cookie wasn’t set in the browser. So we asked LLM for help debugging.
The solution it wrote was to return the JWT as a token in the browser, then set it in a cookie on the client-side ☠️

2026-01-17 12:11:25 UTC

Guys…I have no idea what people mean when they talk about multi-agent LLM development workflows.
I’ve been using Cursor with “Auto” mode, basically prompting one feature at a time, and I’ve been thrilled with the results.
But I keep seeing ppl talk about multi-agent and I can’t picture what they’re working on. The LLM takes like 10-30 seconds to complete my prompts. Where is there even time to type more prompts before it completes? I can’t tell how much is larp and how much is my own ignorance

2026-01-14 16:45:02 UTC

- reply

@nprofile…05ue it is! Web only though. I would say it’s about 80% finished.
I’m hoping that leaning into LLMs will help me finish the mobile app faster than I expected…we can’t fully release nextgen PV until mobile is finished since we need a major data migration. Trying to make the website and infrastructure a standalone open source project before then tho.
https://alpha.podverse.fm

2026-01-14 14:43:53 UTC

- reply

@nprofile…05ue the new Podverse we are working on has alternate enclosure support built-in from the ground up. We will try to provide the best alternate enclosure functionality we can for your feeds.

2026-01-14 09:24:27 UTC

- reply

I still don’t understand all the common terminology with LLM development I’m seeing thrown around though. Problem could be my own lack of knowledge.

2026-01-14 09:17:50 UTC

I’m finding trying to keep up with LLM advances for coding pretty disorienting…on one hand there are some real and magnificent breakthroughs, and on the other there is a lot of useless slop masquerading as breakthroughs.
Improved LLM models are already accelerating my productivity by several times, but I’m still seeing enough errors and anti-patterns, that I have a hard time believing the “you can do it all with prompts” and “you don’t need to review every line” camp isn’t being naive.

2026-01-12 05:45:53 UTC

- reply

@nprofile…k8p7 awesome. Thank you!

2026-01-10 23:53:22 UTC

- reply

@nprofile…peg3 do you use a full CLI approach? I haven’t made that leap yet, and with how I’m using LLMs at least, it doesn’t seem like they’ve eliminated the need for manual edits in all cases.

2026-01-10 06:23:53 UTC

- reply

@nprofile…nh7l yeah, and it's important to be cautious. I still reviewed all the code, and there were some issues with it, and some choices it made that were not my preference...but even after the manual changes I needed to make, it saved me so much time, it's crazy.

2026-01-10 04:41:10 UTC

- reply

The cost of using Opus 4.5 is substantial. A long 12 hour day of work cost me almost $50 in Anthropic API credits.
Since we haven't turned a profit yet with Podverse, that comes out of my pocket and is not easy to sustain, but the time and energy savings make it worth it.
I'm currently trying to replace Opus 4.5 in my setup with GLM 4.7, with is the highest rated open source / MIT LLM for coding available today. API credits are 10% the cost of Anthropic's, so hopefully it will be good enough.

2026-01-10 04:36:51 UTC

I don't know if there is anyone left who is skeptical...but the LLM updates over the past couple months for programming are a giant leap forward, and it seems like we've crossed an inflection point for the industry.
Using Opus 4.5, I was able to get what would have been 5 days of work for me finished in 1 day.
One stage of my work included writing 8 paragraphs of instructions for Opus, and it wrote hundreds of lines of code across 7 repos, and was ~99% accurate and finished after a few minutes

2026-01-10 04:28:49 UTC

- reply

@nprofile…k8p7 I don't see an artist RSS feed for Hartlight in Podcast Index:
https://podcastindex.org/search?q=hartlight&type=all
Our search directly queries PI API, so if a feed is not discoverable there, it won't be discoverable in PV.
I see that the album feed has a publisher tag that is pointing to the artist feed, which is great, but we have as a requirement that only feed that are in PI can be added to our database.
Does Fountain typically add artist feeds to PI? Or does PI not seek to include them?

2026-01-09 09:00:35 UTC

- reply

@nprofile…td70 @nprofile…7td6 @nprofile…0fv8 @nprofile…peg3 we don’t have an advanced log implementation at all right now. It’s just STDOUT to files from docker containers onto the host machine. I run all the log statements through Winston helpers tho with the intention of doing something more useful with them later. I don’t know what a good logging system looks like yet tbh

2026-01-09 04:48:24 UTC

- reply

@nprofile…td70 @nprofile…7td6 @nprofile…0fv8 yikes 😬
I guess there must have been an error that caused it to abort the parses instantly and bombard you with requests.
Maybe I'll add I short timeout at the beginning of parse operations to make sure you don't get spammed that badly if something goes wrong.

2026-01-09 00:31:29 UTC

- reply

@nprofile…td70 @nprofile…7td6 @nprofile…0fv8 the user agent should be:
Podverse/Alpha/2/Worker
Maybe there was a high volume of requests. I had "recursively parse remote items" enabled in our parser, which not only parses one feed, but also parses the remote items from that feed.
I could add some form of rate limiting to our parser to not barrage PI too quickly.

2026-01-08 21:51:09 UTC

- reply

@nprofile…td70 @nprofile…7td6 @nprofile…0fv8 after an hour, it appears that the Cloudflare restriction was lifted.
However, we could really use some way to ensure we don't get blocked by Cloudflare. I would think a valid PI API key should be enough, but I guess the Cloudflare protection must preclude API key validation. Any ideas would be appreciated.

2026-01-08 21:23:57 UTC

@nprofile…td70 @nprofile…7td6 how can we avoid Cloudflare breaking our connection to PI API?
I was able to make 3 successful requests to PI API from Alpha server, and I didn't change or redeploy anything...and since then, my 4 attempts to make a request to PI API have failed. The error message is 429, and it is an HTML respnse that includes:
Access denied | api.podcastindex.org used Cloudflare to restrict access
cc @nprofile…0fv8

2026-01-07 03:35:07 UTC

- reply

@nprofile…peg3 actually...I am just realizing that Unified Push is only compatible with Android. I thought maybe it would work with web also...
Not sure if there would be value in letting people change UP settings through the website. Seems like something that should be configured within the mobile app on each Android device.

2026-01-07 03:11:32 UTC

- reply

@nprofile…peg3 thanks! I'll see how far I can get, and may have some questions for you.

2026-01-07 02:21:05 UTC

- reply

Next, I will try to add Unified Push notification support, so listeners can use their own open-source, self-hosted push notification systems.

2026-01-07 02:20:13 UTC

Podverse Alpha Update 🛠️
Push notifications have been added to Podverse Alpha web.
Listeners can configure 3 different types of notifications:
- New item (episode or music track)
- Livestream scheduled
- Livestream started
They can also set different settings for different feeds.
This was a lot of work to add. The code includes not only support for web, but an orchestrator to send notifications to also iOS and Android (when the mobile app is ready).
https://alpha.podverse.fm
#podverse #pv2
https://cdn.masto.host/podcastindexsocial/media_attachments/files/115/851/379/637/421/016/original/a908a9d29f10b179.png
https://cdn.masto.host/podcastindexsocial/media_attachments/files/115/851/381/592/141/112/original/78f15838e0804554.png
https://cdn.masto.host/podcastindexsocial/media_attachments/files/115/851/383/779/817/232/original/b4031c4fda251068.png

2026-01-03 20:03:51 UTC

- reply

@nprofile…7td6 @nprofile…8zna thanks! Sounds like it should be a simple fix. We just need to prioritize the artwork field over the image field from PI API?

2026-01-03 03:15:27 UTC

- reply

@nprofile…km5y @nprofile…8zna it looks like the code is already setup to prefer the itunes:image over the image. If you can provide an example feed with the issue in Podverse Alpha that could help with debugging.
https://github.com/podverse/podverse-parser/blob/v5-develop/src/lib/compat/partytime/channel.ts#L97-L107

2026-01-02 22:48:44 UTC

- reply

@nprofile…km5y @nprofile…8zna can you link me to podcast you identified the issue with?

2026-01-02 21:10:47 UTC

- reply

@nprofile…km5y @nprofile…8zna thanks! I’ll look into it.
We have a helper function that should use podcast:image, itunes:image, image in that priority order. I must need to fix that.

2025-12-30 21:40:12 UTC

- reply

@nprofile…5yxp looks good 👍 should work fine

2025-12-30 21:37:48 UTC

- reply

@nprofile…8nsk @nprofile…5yxp @nprofile…y5nd ok, as long as it has to be a UUID, then that won't require a lot of changes on our side.
right now our channel.podcast_guid column in postgres is a UUID data type. i have some extra validation before updates are saved to confirm the UUID is UUIDv5, but i can replace that with a check for any valid UUID type.

2025-12-30 21:04:18 UTC

- reply

@nprofile…8nsk @nprofile…5yxp @nprofile…y5nd hmm I designed our data schema to match the PI namespace docs, and it seems to specifically say UUIDv5 for <podcast:guid>. If other UUID types are valid for <podcast:guid> I need to make substantial changes.
https://github.com/Podcastindex-org/podcast-namespace/blob/main/docs/tags/guid.md

2025-12-29 12:04:17 UTC

- reply

@nprofile…5yxp but I understand that people who are early adopters of a tag like remoteItem may be manually creating these feeds by hand, which makes it possible for something like copying 35 characters of a podcast guid when a podcast guid needs 36 characters in UUIDv5 format to adhere to the P2.0 spec. This is super easy for an RSS generator to programmatically validate and make sure never happens…but not easy when people may be typing or copy/pasting fields like this manually by hand.

2025-12-29 12:00:15 UTC

- reply

@nprofile…5yxp I will try parsing your other list feeds later this morning.
I also may be able to add filtering before we do a bulk save Postgres transaction, to remove remote items with invalid guids so the error will be avoided.
Also, new Podverse can handle if a remoteItem or value tag points to a feed that does not exist, but trying to save “invalid data” like this edge case (an invalid podcast guid structure) errors the parser. imo it should be the RSS generator’s job to not publish invalid guids

2025-12-28 19:44:58 UTC

- reply

@nprofile…0lgv I think I’ve made progress on this. The he.decode line in the parser config was stumping me for a while. Things seem to be working when I remove it. Now I gotta figure out what problems removing that could cause…

2025-12-28 08:40:55 UTC

- reply

@nprofile…0lgv I tried to make progress with an upgrade from v3 to v4, but I have not been able to figure out how the new validated and parsed output format is supposed to work.
I imagine many helper methods that you wrote would need to change. I'm open to doing this work, but I don't have a clear direction on how to proceed yet. If there's any insight you can offer I would appreciate it, but if you're too busy no problem.

2025-12-28 08:37:30 UTC

Hey @nprofile…0lgv I started working on a fork of partytime. The codebase and dev setup you created is great! I was able to add the "medium list" types, as a first feature addition.
I am wondering if you have any insight on how upgrading v3 fast-xml-parser to v4 fast-xml-parser would work.
There is a moderate security vulnerability with v3, and I tried upgrading, but apparently the output of the parser is different with v4, and your validation methods are incompatible.
https://github.com/podverse/podverse-partytime/pull/1/files

2025-12-27 00:17:34 UTC

- reply

@nprofile…dzgg @nprofile…y5nd we should be able to integrate with the Rust parser outputs. I think we may want a layer of Typescript types to go with the output. That’s just a data mapping exercise tho, and not strictly necessary.

2025-12-26 18:38:16 UTC

- reply

@nprofile…dzgg but also...partytime just seems really stable for our purposes, so I'm not sure if it is worth it for us to break from our full-stack JS pattern to incorporate feedparser/Rust. Also partytime uses Typescript types which is good for our builds and workflow.
Anyway, if no one else is eager to maintain partytime, I think I will fork it tomorrow, make the changes needed, and maybe longer term start a migration to feedparser. I see you made some PRs that are not merged and I can take a look.

2025-12-26 18:35:06 UTC

- reply

@nprofile…dzgg I have seen it. I'm open to using it at some point, but podcast-partytime works well with our full-stack JavaScript infrastructure.
I'm thinking it would take me a day of work to add the missing pieces we need to podcast-partytime, but switching to feedparser which is written in Rust would be a bigger initiative.
partytime logic is only used in podverse-parser/src/lib/compat and podverse-parser/src/lib/rss, so swapping it out doesn't seem crazy difficult
https://github.com/podverse/podverse-parser/blob/v5-develop/src/lib/rss/channel/channel.ts

2025-12-20 02:21:41 UTC

- reply

@nprofile…efpk @nprofile…hs8h looks interesting! Will the RSS feeds support any Podcasting 2.0 namespaces?
A simple one could be the `<podcast:medium>music</podcast:medium>` within the `<channel>` for every feed. This signals to apps that a feed should be handled as a "music experience" instead of a "podcast experience".
https://github.com/Podcastindex-org/podcast-namespace/blob/main/docs/tags/medium.md
The new Podverse (open source, Podcasting 2.0 compatible app) has a heavy RSS music focus. A live alpha demo can be found here https://alpha.podverse.fm/album/YX0N5AHegy

2025-12-18 20:31:50 UTC

- reply

@nprofile…5yxp sorry, too early. Anything on Alpha could be deleted later. There will be a complex migration process from PV1 to PV2 when it finally goes live, and the alpha data won’t be part of that (hence the disclaimer and fact we’re not charging for alpha memberships).

2025-12-18 20:04:42 UTC

- reply

@nprofile…5yxp well I’m happy to refine it to get it to where it needs to be.
The issue of not having everything possible in our database (until people manually add it) may not have an easy answer…our new infra is designed like this to prevent spam and database bloat.
However, if there aren’t too many music feeds today (<10000?), we could potentially just add every music feed that is in the index to our directory.
A V4V only filter could be nice to highlight those feeds.

2025-12-18 16:43:06 UTC

- reply

@nprofile…5yxp that way. Share is common enough maybe we add it but I am concerned about “choice paralysis” if the dropdown menus have too many options.
Thinking of adding a Like button to quickly add items to a Favorite Podcasts or Favorite Music playlist, similar to the heart button in PV1. Infrastructure already there, would just need to add buttons for it on the page.

2025-12-18 16:40:15 UTC

- reply

@nprofile…5yxp yeah auto may need to allow customization but presumption is listener wants to hear the album OR the playlist. Playlist autoqueue may be incomplete, not sure.
What tracks are missing from global? It’s “global” in the sense they’ve been added to our directory by someone searching for it and pressing add feed.
“Top” filter not implemented yet until stats tracking is enabled.
Shuffle and repeat coming soon.
Share button for track can be added although you can Go To Track and copy url

2025-12-18 16:34:21 UTC

- reply

@nprofile…5yxp @nprofile…8nsk reorganizing your queue with thousands of items a huge operation bc every rows position in your queue would need to be updated when 1 row changes 🤦‍♂️

2025-12-18 16:33:10 UTC

- reply

@nprofile…5yxp @nprofile…8nsk it was probably due to an “auto add new episode to queue” feature being turned on, which was possibly buggy.
The new history / now playing / queue system is hopefully improved over old version. In PV1 those were 3 tables. In PV2 it’s one table with a list_position column:
history < 0
now playing = 0
queue > 0
It uses large decimal position numeric data types to easily insert rows where we want instead of an integer type. Old version used an integer type which would make

2025-12-18 16:25:32 UTC

- reply

@nprofile…8nsk @nprofile…5yxp currently PV Alpha queue is setup as follows:
Anything manually placed in the queue gets played next.
If nothing is in the manual queue, play next from album.
If track was played from a playlist page, then play next from the playlist (not sure I finished this tbh).
If people want auto queuing off I guess I could put that on the listener settings page.
Down the road, could have a “radio” option that algorithmically guesses what track to play next based on stats tracking.

2025-12-17 18:31:39 UTC

🛠️ Podverse Alpha Update 🛠️
IT'S A BIG ONE
Started to add the Albums, Album, Tracks, and Track pages for Music playback.
The UX also splits the Queue, History, and Playlist into Podcasts and Music.
Any feedback or suggestions for improvement would be appreciated!
https://alpha.podverse.fm
#podverse #pv2
https://cdn.masto.host/podcastindexsocial/media_attachments/files/115/736/331/378/281/651/original/2e21b660fcb53be1.mp4

2025-12-17 05:35:14 UTC

- reply

@nprofile…8nsk @nprofile…wsn5 @nprofile…5yxp if tags is a good fit, i'll use them.

2025-12-17 01:13:20 UTC

- reply

@nprofile…wsn5 @nprofile…5yxp
> Wavlake adopted the category tag, implemented wrongly, but whatever, it works.
What did they handle incorrectly? I see this example from The Underclass:
<podcast:category text="Rock">
<podcast:category text="British Invasion"/>
</podcast:category>
is that incorrect?
https://wavlake.com/feed/music/fa16edd9-d3d2-4c6f-a6e1-c9d43307a0bf

2025-12-17 01:04:47 UTC

@nprofile…5yxp @nprofile…wsn5 i've asked this before but forget 🤦‍♂️
is there something comparable to "categories" or "genres" in use for P2.0 music feeds?

2025-12-16 21:18:48 UTC

- reply

@nprofile…wsn5 @nprofile…5x2e @nprofile…5yxp @nprofile…vwpk this is good to know. thanks!

2025-12-15 21:01:51 UTC

- reply

Thank you to @nprofile…5x2e and the Doerfels for this complex feed test case 😂
Some items in the feed have a season and episode number, and some do not.
I'd recommend continuing to add them going forward for apps to handle auto-play ordering like you expect, but if not, apps can use fallbacks similar to what we have.
https://www.doerfelverse.com/feeds/music-from-the-doerfelverse.xml

2025-12-15 20:58:11 UTC

- reply

That isn't very complex, but things get complicated when a music feed contains tracks that have an "episode number", and tracks that do not have an "episode number". Where do the non-numbered tracks go in the auto play order?
To handle this, I (::cough:: llm) wrote the query to handle all tracks in a music feed that do not have an episode number as if it had a season of -1, and to sort those by pub_date, while the sort for properly numbered tracks get handled as normal.
https://github.com/podverse/podverse-orm/commit/3470108ab6c81de9dac9a1f0a7c9fd6bbc810ffa#diff-c3343c792900d109ee2695d00313b8b2e68844b2d088a3271b7527b4c88d459eR384

2025-12-15 20:54:51 UTC

Podverse Alpha update
Started to add Music player support.
One of the key differences between Podcast and Music in a Podcasting 2.0 app is how the "auto play next from queue" is handled.
For a podcast, you would typically autoplay the next episode, sorted by its pubDate.
For a music album, you need to play sorted by a track number. The convention used is the season and "episode" tags. Instead of sorting by pubDate, tracks are sorted by the "episode number"...

2025-12-15 05:08:51 UTC

- reply

I *think* I'm clear to start adding the Music pages tomorrow.

2025-12-15 05:08:17 UTC

Podverse Alpha update
Added a new database table named on_demand_parser_event, which will keep track of when a listener adds a feed to our directory, and help prevent overuse / spamming of our "add feed to directory" feature.
Not an exciting update from a listener's standpoint, but an important step, since we are giving listeners the ability to trigger RSS parsing on our servers in PV2, and that can be a data and resource heavy operation.
#podverse #pv2

2025-12-13 04:03:18 UTC

- reply

Podverse Alpha Web progress:
- Redesigned the Subscribe button layout.
- Added a "Podcast Settings" page and a "Check Feed for Updates" button to it. That button lets users manually trigger a feed re-parse in our directory.
- Some misc style improvements for mobile devices.
I did not start on the music pages yet, but that's next on the list.
Changes are not deployed yet. I think I will wait until the music pages are finished before deploying next, possibly Sunday night.
#podverse #pv2
https://cdn.masto.host/podcastindexsocial/media_attachments/files/115/710/233/783/352/045/original/c94a1e867c9b6374.png

2025-12-12 20:18:06 UTC

Today's goals for #PV2 are:
- Add the ability for listeners to manually trigger a RSS feed refresh on our servers.
- Improve the Subscribe button layout. It's awkward having the button so far on the right side.
- Start work on the Album, Track, and Music Livestream pages

2025-12-11 02:46:54 UTC

- reply

@nprofile…8nsk it seems like a natural fit to me, but the fallback I mentioned doesn't seem bad in the meantime. I'm curious if other people would want it.

2025-12-10 23:33:33 UTC

Does Podcasting 2.0 have a way to link an "ended" livestream, to the recording of that broadcast within an RSS feed?
I imagine it could take the form of a guid in the <podcast:liveitem>, and then when the broadcast is finished, that same guid is included in a tag for the <item> within the same feed.
If we had something like this, then an "ended" livestreams page like this could make clicking on an ended stream redirect to the item (if it exists).
Or is this a behavior we'd rather avoid?
https://cdn.masto.host/podcastindexsocial/media_attachments/files/115/697/865/986/817/684/original/e956d32fcd84358a.png

2025-12-10 06:21:28 UTC

- reply

The pagination widget on the bottom showing too many pages is a known issue. Will be fixing that soon.

2025-12-10 06:20:17 UTC

- reply

On a better note, here is a work-in-progress gif preview of the Livestreams page I worked on today for Alpha.
Next I need to add a Livestream page, then we can deploy these changes.
Albums and Tracks pages should be finished shortly after. They will basically be copies of the existing Podcasts and Episodes pages, so I'm thinking it'll only take a day.
#podverse #pv2
https://cdn.masto.host/podcastindexsocial/media_attachments/files/115/693/820/518/179/316/original/2ad33a1bc32bf7d6.mp4

2025-12-10 06:16:12 UTC

⚠️ Podverse Alpha Update ⚠️
I'm shutting off the Alpha website for the night, as there is a critical security vulnerability with the version of Next.js it is using (CVE-2025-55182).
Should have it redeployed and patched tomorrow.
Good video explaining the CVE: https://www.youtube.com/watch?v=s81dVUM-cQM
Thanks to @nprofile…ezja or notifying me.
#podverse #pv2

2025-12-09 04:19:56 UTC

- reply

@nprofile…t46n thanks for the heads up! I change the rate limit rules for function.

2025-12-06 19:13:10 UTC

Decided to merge "audio" and "video" into one grouping. Rewriting the API and modules to be able to handle it.
I think the primary groupings listeners will see will be Podcasts and Music.
#podverse

2025-12-05 14:33:37 UTC

- reply

@nprofile…5yxp it helps! I'm leaning towards merging the experiences in the UI...

2025-12-05 13:22:34 UTC

Working on the Livestreams page for Podverse Alpha Web today.
I'm also wondering if the UI should even separate Audio "podcasts" from Video "channels".
Music listening is a distinct experience than a podcast, imo (like the way the media player handles album track order), but I personally listen to a bunch of "video podcasts" without even looking at the screen.
It's also kind of awkward grouping podcasts under "audio" when podcasts can be video. Anyone have thoughts?
#podverse
https://cdn.masto.host/podcastindexsocial/media_attachments/files/115/667/165/907/296/981/original/882807d7e0ac6b27.png

2025-12-04 19:16:12 UTC

- reply

@nprofile…4t4s I’ll be working on the pagination funkiness. Relatively light fixes to make.

2025-12-04 19:15:15 UTC

- reply

@nprofile…4t4s maybe we should make subscribe a bit smaller, and fit it along side the other buttons on the bottom left of the header?

2025-12-04 19:11:22 UTC

- reply

@nprofile…4t4s if you search for your podcast again, it should direct you to the podcast page this time.

2025-12-04 19:10:11 UTC

- reply

@nprofile…4t4s I need to make that UX smoother.
"Add Feed" should navigate you to the Podcast page after it starts parsing. From there, you can subscribe to it.
Maybe we should auto-subscribe whoever adds the feed, since that is an expected behavior?
Also, that blank screen should have a message to explain why it is blank (you're not subscribed to anything yet). Thanks for the report

2025-12-04 19:07:23 UTC

- reply

@nprofile…4t4s our new search will search the Podcast Index directly, and then if you want to add it to Podverse, there is an "Add Feed" button.
We're doing this to drastically reduce the size of our database. Old Podverse has millions of podcasts and hundreds of millions of episodes, and it's very heavy and unnecessary.
This is one of the major optimizations for the new version.

2025-12-04 18:38:37 UTC

The Podverse Alpha website is now live!
This is a work-in-progress, public demo of the upcoming Podverse "2.0".
The new Podverse is designed, from the ground up, to support as many Podcasting 2.0 features as possible.
Functionality is limited right now, but the core is there.
The test website can be found at https://alpha.podverse.fm
For more info, please read https://alpha.podverse.fm/updates We will be continuously posting updates there.
We would greatly appreciate help with testing and feedback.
https://cdn.masto.host/podcastindexsocial/media_attachments/files/115/662/741/977/015/912/original/026483f52fc2c99c.png

2025-12-01 20:08:01 UTC

Update on Podverse 2.0...first version of the Alpha website should be publicly accessible this week.
Technically could be shared now...but ironing out some things and finishing some pages before that.

2025-11-29 05:16:18 UTC

- reply

@nprofile…nhmd thanks! I’ve bookmarked this and will look into it when we’re ready to write the mobile app.

2025-11-26 07:07:19 UTC

Podverse 2.0 Preview:
Added the Home and Episodes page.
Next I'll be adding the Clips, Livestreams, Videos, and Music pages. It should be fairly simple with the new endpoints and UI components in place.
The PV2 Alpha website should be publicly accessible for testing before the end of this month.
cc @nprofile…ezja who handles the ops and more for Podverse, and I couldn't do any of this without him.
https://cdn.masto.host/podcastindexsocial/media_attachments/files/115/614/706/982/373/391/original/36cc38baac4947b7.png

2025-11-19 23:05:54 UTC

- reply

@nprofile…nhmd can you remind me what NWC is? 😅

2025-11-16 19:32:22 UTC

- reply

@nprofile…5yxp yessir! Music will be supported as a primary feature, instead of how it is in PV1, where it's kind of an afterthought shoe-horned in.

2025-11-16 19:12:34 UTC

It's LONG overdo, but I think the Podverse 2.0 website will finally be available in beta before the end of this month.
The goal of Podverse 2.0 is to support as many Podcasting 2.0 features as possible, from the ground up. Emphasis is placed on multiple medium experiences (podcast, video, music), alternate enclosures, and livestreams.
Other goals include a focus on optimizing app performance, simplifying code, and closing feature gaps between web and mobile (web will support custom RSS feeds).
https://cdn.masto.host/podcastindexsocial/media_attachments/files/115/560/788/224/496/906/original/10f65aacb202e768.png
https://cdn.masto.host/podcastindexsocial/media_attachments/files/115/560/804/720/898/793/original/9133aff6cb7106f2.png
https://cdn.masto.host/podcastindexsocial/media_attachments/files/115/560/806/157/928/740/original/10cb17c3173718e0.png

2025-11-16 18:25:54 UTC

@nprofile…y5nd @nprofile…tmcv am I correct that the Search endpoints for Podcast Index do not have a pagination query param?
We don't strictly need it in Podverse (I'll just request a large 50 rows limit), but I was surprised to not see it, so I'm wondering if there's documentation for it I'm missing.
https://podcastindex-org.github.io/docs-api/#get-/search/byterm