Mitch Downey :pci:

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

Podverse: https://podverse.fm
Github: https://github.com/mitchdowney

Public Key

npub1p7v0a3mgwgqkhyuuwwrdf7kyt8c0gker3asa4lp0l626p0f3ck5sklqql0

NIP-05 Address

mitch_at_podcastindex.social@momostr.pink

Profile Code

nprofile1qqsqlx87ca58yqttjww88pk5ltz9nu85tv3c7cw6lshla9dqh5cut2gpz3mhxue69uhhyetvv9ujuerpd46hxtnfduqs6amnwvaz7tmwdaejumr0ds9k92m3

Publishing to

relay.damus.io

nos.lol

Author Public Key

npub1p7v0a3mgwgqkhyuuwwrdf7kyt8c0gker3asa4lp0l626p0f3ck5sklqql0

Show more details

Last Notes

2026-10-19 04:39:41 UTC

- reply

nextgen currently is setup to pull from podcastindex_dead_feeds.csv every hour for this purpose.

2026-10-18 23:30:14 UTC

- reply

not seeing duplicates now!
Question: for the ones that were handled as duplicates, is the handling still "dead" and "duplicateOf" to signal what is removed and which feed it is a duplicate of?
No major changes to that process from how it used to be? Just a layer that handles detecting and marking duplicates better?

2026-10-17 20:15:17 UTC

Imagine listening to Joe Rogan Experience and using your own personal in-app llm integration to skip over all discussion of elk meat, feral boars, and brazilian jujitsu 🫠

2026-10-17 19:37:38 UTC

- reply

it is getting amazingly easy for "average person" to make a personalized app, if they have the determination to do so.
Making a sophisticated app that handles the myriad conditions someone without deep architectural experience however is still very difficult. Many thousands of prompts and corrections difficult.
Perhaps tho, the biggest barrier to avg joe making his own apps, are the hurdles app stores put in your way to publis on them. Have to be willing to persevere through that.

2026-10-16 17:24:43 UTC

- reply

I have not released an update in 2 years…it has gotten that bad? There was a time I thought Google was faster than Apple.
I wonder if store review issues are enough to keep React Native in the mix, since it can support over-the-air updates for at least some changes.
Although I’m not sure how they allow over-the-air at all given that it seems like a loophole to their review processes.

2026-10-14 04:02:01 UTC

- reply

is always doing that

2026-10-14 04:00:13 UTC

- reply

thanks to transcripts and BoK (this was ChatGPT), you can now communicate in mere grunts and the full context can be unpacked lol

2026-10-14 03:58:28 UTC

- reply

I asked the BoK what you meant by that:
Adam Curry almost certainly meant SOPML = Social OPML.
> He's suggesting that the problems could be addressed using an existing decentralized social-graph concept rather than inventing a new personGUID mechanism.
> In Podcasting 2.0 episode 175, Everyone's an Actor, Adam and Dave Jones explicitly discuss SOPML as a social graph.
> It uses OPML to distinguish between feeds someone publishes (pub) and feeds they subscribe to (sub).

2026-10-13 04:47:16 UTC

- reply

feeds for people could be nice. An official RSS feed for their podcasts, appearances on other shows, and other info they want (blog posts even?).
Maybe it's just a publisher or playlist feed, that can be referenced in the person tag with a guid?
Biggest challenge might be integrity verification to prevent people from being misled to a fake "official person" feed containing something malicious.
Very experimental but I would open to giving it a try.

2026-10-03 19:16:42 UTC

I am behind on my podcasting 2.0 listening. Does the new MCP server open up the possibility for apps to add an AI option to their search pages?

2026-09-30 16:40:44 UTC

- reply

ah if it means “I don’t want to listen when you tell me” then I would disagree. Listening live and listening later are not exclusive UX. YouTube it seems is increasingly popular for “podcast listening” today, and I think livestreams are part of what has driven that adoption. If a listener wants to occasionally hear the livestream, and it’s live only on YouTube, they tend to stay and listen to the recorded versions there as well.

2026-09-30 12:39:46 UTC

- reply

“People wanted to time shift listening with podcasts so live becomes the opposite of what podcasts provide”
What do you mean by “time shift” here? The ability to rewind in an ongoing livestream? If that’s it, then my understanding is PeerTube supports this functionality called HLS DVR since at least 2024.
I do not know if any 2.0 apps support it yet, but we certainly want to with PV NG.

2026-09-29 16:04:11 UTC

- reply

I would suggest that, if we expand or adopt new namespaces, we don’t strictly need to use <podcast:__> namespaces or have them be officially Podcast Index creations. The more important thing is we have a community willing to experiment with new tags.

2026-09-29 15:57:11 UTC

- reply

I have long thought that podcast:soundbite is under utilized for podcasters creating their own clips. Does True Fans allow them to set a title with their clips? I haven’t checked if it has changed in a long time, but at one point Buzzsprout allowed podcasters to create soundbites but without titles. I think titles could help with discovery.
We’ll be trying to feature soundbites (tentatively called “Official Clips” to differentiate from listener created) more prominently in PV NextGen.

2026-09-27 21:00:40 UTC

- reply

that would be great! Any feeds that could be used for development, and any documentation would be appreciated.

2026-09-27 21:00:14 UTC

- reply

awesome thanks Sam. I may hold off until we actually have Podverse NextGen in beta testers hands, because there won't be much I can do until then, and then I'll forget and ask the same questions over again 😂
NextGen should be much more adaptable than Ogen though so it shouldn't be too difficult for us to experiment with new tags quickly when it is ready.
Trying to get Beta in testers' hands before end of October. Mobile app mostly finished. Working on payment and data migration now.

2026-09-27 20:50:04 UTC

- reply

Podverse would potentially be interested in supporting both value-for-value and RSS-based paywalled content.
I see them as separate and non-overlapping features. It's extremely common these days for livestreamers to offer both exclusive premium content AND "superchats" aka boosts aka paid messages sent by the audience (read on both the public and private content).
V4V is more of a philosophy, but sending paid messages is the tangible feature most people engage in V4V through.

2026-09-27 20:43:52 UTC

- reply

this seems like an odd direction for Apple to be going in, when Apple is a device and OS specific company.
Why would a creator want to invest in the effort to publish subscriber-exclusive content, to their audience that is only on one type of device, when cross-compatible platforms are available?
They're trying to compete with YouTube Premium and Patreon exclusives?
YouTube already works on AppleTV so...not sure how they expect this to work. Feel like I'm missing something.

2026-09-22 20:30:08 UTC

- reply

that’s exciting and good luck with the new versions!
Re: man pushing rock uphill for eternity, I know the feeling too well 😂 Hopefully with our new releases we can find some more “synergy”. From what I hear V4V adoption has slowed over the years. Would love to see that trend reversed.

2026-09-22 18:31:03 UTC

Was wondering why Cursor Grok 4.7 kept refusing to work on my plan file, telling me to "try a less sensitive prompt."
Apparently, it's because my plan investigates "candidates" in my code that could explain a UI slowdown, and it doesn't like the phrase "Read and execute candidate..." 😂

2026-09-22 17:53:56 UTC

- reply

does this mean the response shape has changed? Or the query params those endpoints accept are different?
If response shape has changed, is it provided with a v1/v2 type prefix to not break existing requests?
Sorry, I’m not fully up to speed yet on new schema. PV NG mobile is pretty far along so I should be able to revisit PI API more soon…

2026-09-22 17:46:34 UTC

- reply

I’m doing well! A lot of life changes this summer, moving out of state and getting a new job.
Still working on the full rewrite of Podverse. It’s very far along but feels never ending.
How are you and TrueFans doing?

2026-09-21 02:01:17 UTC

- reply

happy birthday Sam!

2026-09-20 21:08:17 UTC

Sneak peak video of the Podverse NextGen mobile player screen
The player screen should be for the most part the same as the previous gen, with the exception of these tabs at the bottom of the screen, and the ability to scroll down.
This is an experiment and not sure what % of people will like it. The upside of this is that the features that are available for the episode/music track will be instantly visible, instead of hidden behind a carousel in the middle of the player.
https://cdn.masto.host/podcastindexsocial/media_attachments/files/117/305/398/461/823/193/original/8e053f6f74f62aa5.mp4

2026-09-15 20:29:09 UTC

- reply

ok, sounds good! That makes sense. I'll wait for docs before asking more questions.

2026-09-15 20:22:52 UTC

@npub1rcn…dw4h listening to last week's P2.0...the nextgen Podverse schema (and ogen) relies heavily on the Podcast Index IDs as basically the authority on feed uniqueness. It sounds like we may need/want to do a substantial rewrite to account for how Podcast Index API nextgen will handle ids?
I don't fully understand what we'll need to do differently yet. Just mentioning it for awareness.
cc @npub1yda…h4tz

2026-06-21 04:53:48 UTC

- reply

thanks TJ! Appreciate your feedback. If anything is missing or there’s something we can improve please let me know.

2026-06-20 21:45:37 UTC

- reply

awesome! And I just found it through your Podroll. https://alpha.podcastdj.com/podcast/02cR5Iwbor
https://cdn.masto.host/podcastindexsocial/media_attachments/files/116/784/621/724/181/787/original/98a53205a8121f7a.png

2026-06-20 20:52:00 UTC

- reply

cc @npub1fmx…wp74 @npub1rcn…dw4h please check out the PV Nextgen "alt enclosure" demo above when you have a chance. I think we're very close to "Beta Web" release.

2026-06-20 20:50:20 UTC

- reply

I'm thinking of letting existing premium podverse.fm members have access to beta.podverse.fm. The tricky part is we cannot go fully live with PV Nextgen until the new mobile app is also completed, bc the new mobile app and website will not be compatible with the old database. We will have to do a full migration process when we finally go live with Nextgen and replace the old version. We need to minimize data loss following the migration so listeners have as seamless of an experience as possible.

2026-06-20 20:49:33 UTC

- reply

This PodcastDJ website is meant as a temporary and separate home for Podverse nextgen. It helps demo how the open source Podverse code can be deployed by anyone on their own servers, and is also a separate data environment so people know it's not to be confused with Podverse.
I'm considering deploying the new Podverse website to beta.podverse.fm soon, and including clear warnings that the data/history/subscriptions/everything will not overlap with people's main podverse.fm accounts.

2026-06-20 20:48:42 UTC

Podverse NextGen Update 🛠️
Been working on audio + video playback enhancements. Here's a demo of what we call the "alt enclosure" system in podcasting 2.0 RSS feeds.
Listeners can set if they want audio or video to load by default. In this demo the user has video on by default.
The "wave" icon indicates that an episode or music track has both an audio and video version.
@npub1ncy…zkql and the Doerfels have an alt enclosure feed that really showcases what it can do
https://alpha.podcastdj.com/album/L93EkHGxGT
https://cdn.masto.host/podcastindexsocial/media_attachments/files/116/784/394/728/139/622/original/61731329c68ccba6.mp4

2026-06-18 14:02:23 UTC

- reply

here is the error message. It happened even after I added the backoff retry attempts.
https://gist.github.com/mitchdowney/fb576de610e98899241c81f849f8aad6
A worker container calls it from here:
https://github.com/podverse/podverse/blob/develop/apps/workers/src/commands/mq/rss/mqRSSAddRecentlyUpdatedFeedsFromPodcastIndex.ts#L40
The PI + MQ function the worker calls is defined here:
https://github.com/podverse/podverse/blob/develop/packages/mq/src/functions/mq/rss/addRecentlyUpdatedFeedsFromPodcastIndex.ts#L15
The most relevant one though, the PI function is defined here:
https://github.com/podverse/podverse/blob/develop/packages/external-services-podcast-index/src/index.ts#L361

2026-06-18 00:33:02 UTC

- reply

I'll also add a slight delay between the paginated requests in case we're getting blocked for too many requests too quickly.

2026-06-18 00:32:04 UTC

@npub1rcn…dw4h @npub1yda…h4tz I notice in our PV alpha that we are getting 500 errors when we request "get a list of feeds that have updates over the last X amount of time".
We make a series of paginated requests until we get the full list. It seems like the process works fine for most of those requests, but eventually we get a 500 error response from one of the pages. Is this a known issue?
I think I'm just going to add some retry handling for when this happens and see if that resolves it.

2026-06-11 04:54:52 UTC

- reply

exactly what I was looking for. Thank you!
Here's a work-in-progress preview of the podverse nextgen embed player...only the audio version is finished. Will be adding a video version next, with alternate enclosure support.
https://cdn.masto.host/podcastindexsocial/media_attachments/files/116/729/682/944/940/932/original/b098549c5821b1bb.mp4

2026-06-11 01:34:16 UTC

Does anyone know of an RSS feed that has music videos in them?

2026-06-01 03:44:08 UTC

- reply

I don't love using these proprietary models btw. I'd love to rely on my own self-hosted and open source models, but afaik the cost of doing that without it being an absolute slog and fighting with inaccuracies is very high. For some reason GPT is much cheaper than Anthropic. I do think Anthropic models are generally better, but I don't think they're 3-10x the cost better, if budget is a concern.

2026-06-01 03:39:03 UTC

- reply

4) When it is confident, and you've reviewed the plans, and you're ready to go, ask it to implement each plan, one prompt per plan at a time, until it completes all the work.
There's more nuance I could get into, and some custom skill/rule files, but that's the general idea. Basically, I try to get the LLM to plan the work out thoroughly and in bite sized, deferred chunks before asking it to do the work. I haven't found the need to use expensive Anthropic models most of the time.

2026-06-01 03:35:49 UTC

- reply

FWIW the process I have used for large tasks with LLM programming looks like:
1) Use Cursor. It behaves far less quirky than VS Code for programming in my experience.
2) Explain what I want in detail, and ask it to create a high level plan file.
3) Then ask it to break the plan into more detailed subplans, and address any gaps until it is "confident" in the plans, and to save those plan files in the repo, so the work can be deferred to later. Repeat until it says it is confident.

2026-06-01 03:19:54 UTC

- reply

Hey! Sorry for the delay. I'll email you a CSV of the data.

2026-05-01 19:20:19 UTC

- reply

explanation:
> Here “scrubbed” means reviewed and cleaned up for accuracy, not a Git or disk utility.
> In the plan it applies to things like .llm/history or .llm/plans under metaboost.cc that still describe abandoned behavior (e.g. “pin ?ref=develop”, old env filenames, k.podcastdj.com as the metaboost alpha GitOps home).

2026-05-01 19:19:36 UTC

@npub1rcn…dw4h one for the "what does LLM mean by that" list. Do I want LLM to "scrub" this directory?
"If you want .llm scrubbed in the metaboost monorepo too (not only metaboost.cc), say so and we can add that as an explicit todo; right now it’s scoped to metaboost.cc .llm/ plus “don’t contradict” guidance elsewhere."

2026-05-01 19:16:50 UTC

@npub1rcn…dw4h have you setup an "entity expansion limit" for RSS parsing before?
LLM advises me to set an entity expansion limit to avoid "XML bombs." I guess it means a feed is trying to waste your CPU resources.
It tells me 5,000 - 10,000 characters is sensible, but I saw what appears to be a valid (I think?) Arabic language feed from PI that exceeds our current 50,000 limit.
Any thoughts? I'm leaning towards keeping 50,000 as a limit and worrying about this if people complain about a feed missing.

2026-04-22 19:38:11 UTC

- reply

I've been running my LLM accounts hard this whole year, in part because I suspect the prices will go up dramatically at some point.
I mainly use Cursor (the max $200 per month plan). Their "Auto" model is great for most tasks, and the most efficient option apparently.
For more complex / higher stakes tasks I have been using Codex 5.3. I don't know if it is better or worse than Anthropic models at this point, but I have no complaints about how it is working for me.

2026-04-21 18:37:51 UTC

- reply

going to try to add support for this to Podverse Alpha over the next few days.

2026-04-13 22:56:33 UTC

- reply

sorry I missed this. I’ll add it to our priority parsing list so it is less likely to fall through cracks in our legacy platform.

2026-03-19 05:33:00 UTC

- reply

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

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.

2026-03-18 19:50:22 UTC

- reply

cool, thank you. I'm working on adding retry handling and better error message displays right now.

2026-03-18 19:02:13 UTC

- reply

do you know if the Podcasting 2.0 node has availability issues? In some of my tests, the lnaddress lookup works, and in others, it seems to fail.
@npub18xh…9nzr can you confirm if any of these errors are expected?
Unless these errors are entirely on Podverse's side, it seems there is quite a bit of cleanup and rejuvenation needed for the V4V Bitcoin Lightning ecosystem. We are happy to help however we can.

2026-03-18 18:58:29 UTC

Podverse Alpha update 🛠️
EXPERIMENTAL: USE AT OWN RISK
Bitcoin Lightning V4V transactions are live in Podverse alpha!
It supports the traditional "node" send approach with metadata for sending messages, and it also supports lnaddress, but does not yet support sending messages with lnaddress yet (it will just send an amount for now).
I tried sending a 1234 sats boost to @npub18xh…9nzr and Podcasting 2.0 with mixed results...unsure if these failures are due to node issues, or buggy PV code.
https://cdn.masto.host/podcastindexsocial/media_attachments/files/116/251/680/484/742/949/original/dd2d496098ddb4e8.png
https://cdn.masto.host/podcastindexsocial/media_attachments/files/116/251/688/802/129/544/original/e2e38ef869640e75.png
https://cdn.masto.host/podcastindexsocial/media_attachments/files/116/251/693/834/951/408/original/4711529cda8efb48.png
https://cdn.masto.host/podcastindexsocial/media_attachments/files/116/251/697/474/068/516/original/4a49e4b841d78232.png

2026-02-28 05:44:10 UTC

- reply

tbh I enjoyed the moments of zen 🧘‍♂️

2026-02-24 21:18:27 UTC

- reply

interesting ok. So type="lightning" is sufficient to signal to us this is a BTC LN wallet. "type" on the recipient lets us know what approach to use to send to the wallet. Thank you! I have some rewriting to do 😂

2026-02-24 21:13:09 UTC

- reply

is it accurate to say that both "lnaddress" and "node" implementations are both method "keysend"?
I have been talking about "lnaddress" and "keysend" as if they are two different things, but it looks like (according to the P2.0 RSS feed) the difference is between "lnaddress" and "node" rather than keysend?

2026-02-24 21:08:53 UTC

- reply

Looks like the Podcasting 2.0 RSS feed has it. I'll use this as an example.
https://feeds.podcastindex.org/pc20.xml

2026-02-24 21:06:44 UTC

Can anyone point me to an RSS feed that has a proper <podcast:value> lnaddress implementation?

2026-02-21 18:04:01 UTC

- reply

nice feed report page! I hadn't seen this.

2026-02-21 00:32:27 UTC

- reply

as for Podverse, we've been talking it over, and we don't really want ownership responsibility of payment metadata, so we're looking into if there's a way to bridge these worlds. Continue using the lnaddress metadata format other apps are using, but look for an opt-in declaration by the podcasters for where they want their metadata sent.

2026-02-21 00:27:55 UTC

- reply

keysend still works as it did back in the beginning, I believe.
A key difference between keysend and lnaddress, is that keysend lets you send metadata with the payment, but lnaddress does not.
lnaddress does however, allow you send a memo, and so some apps are now hosting their own "boost metadata" endpoints, and including a link to boost metadata in the memo of the payments.
https://github.com/Podcastindex-org/podcast-namespace/pull/734
So keysend adoption appears to be fading, while lnaddress adoption is rising.

2026-02-21 00:21:42 UTC

- reply

> I have no idea what kind of features have been implemented in the spec over the last year
I was posting the same thing a couple weeks ago 😂
In short, for BTC, lnaddress and keysend. Each has trade-offs.
lnaddress is nice bc it lets people specify a human readable, username/domain "email style" reference, that can be looked up to get the wallet address. this lets people provide their lnaddress, and potentially change the wallet address it directs to later.

2026-02-18 21:39:54 UTC

- reply

it certainly can produce garbage. It just seems like it can be steered effectively for dramatic productivity boosts in programming.
I'm throwing a pretty complex plan at Codex 5.2 today for v4v integration + metaboost + boostbox + alby prod/sandbox implementation. will see how it goes...

2026-02-18 20:52:40 UTC

- reply

true. I saw "The Tech Report" and assumed it was mostly a SWE audience.

2026-02-18 08:12:15 UTC

- reply

I keep seeing takes like this…and I’m wondering if I’m really bad at coding and naive, or people are in denial of just how effective LLM is at it.
I barely have to write syntax directly these days. I often have to prompt for improvements, and use some of my own knowledge with debugging, but even after all of that, I feel like my productivity is many times faster since December.
I lean towards the skepticism of this tech being more bullshit than the claims that it’s revolutionary.
https://cdn.masto.host/podcastindexsocial/media_attachments/files/116/090/596/387/153/389/original/4a061e2d5decafdc.png

2026-02-15 02:31:04 UTC

- reply

I didn’t know about the metaboost tag. Seems like exactly what we’d need to support LNURL without losing boost message data.
Boost message data should probably be pulled out of the keysend implementations anyway since it’s so restrictive.
Podverse might be open to hosting a metaboost API as well. I don’t know how we’d manage the spam and abuse problems though…Alecks if you have ideas and would ever be up to talking about them I’d appreciate it.

2026-02-14 22:16:09 UTC

- reply

interesting! I'll keep this in mind. I agree we could offer something better than just Github issues and direct emails.

2026-02-10 23:42:26 UTC

- reply

Speaking of Podverse Nextgen bug reports, we now have ONE place for everyone to report bugs to.
The entire Podverse infrastructure (almost) is within this one Github monorepo. If you notice a bug or have a feature request, all are welcome to create issues here:
https://github.com/podverse/podverse/issues

2026-02-10 23:39:24 UTC

- reply

Podverse legacy also did not have web support for Add by RSS feeds at all. They were a mobile feature only, and an incomplete feature, at that.
With Podverse nextgen, you should be able to add your private RSS feeds on web, and have them available in mobile, and vice versa.
For many reasons, this has been the most complex feature to support in Podverse Nextgen yet...there are many different situations to test for, and we will likely have to iron out bugs as people report them to us.

2026-02-10 23:36:12 UTC

Podverse Alpha Update 🛠️
NEW: Add By RSS feeds on PV web 🎉
"Add by RSS" are feeds that are NOT in the Podcast Index, and are not supposed to be.
For example, private, premium content feeds that you get a link to when supporting a creator on Patreon.
Podverse legacy supported these feeds in the mobile app, but it did not let you add content from those feeds to your queue or playlists.
Podverse nextgen does support Add by RSS queue and playlist functionality. (cont.)
https://alpha.podverse.fm/add-by-rss/add
https://cdn.masto.host/podcastindexsocial/media_attachments/files/116/048/956/453/428/312/original/24e4450f358a892e.png

2026-02-06 18:04:05 UTC

- reply

hah I have not thought that far ahead yet...
But as long as the data can map to the same schema we end up with from parsing RSS, I imagine there's no reason we can't support AP or any other data format.
I do however lean towards RSS being a good enough data standard for podcasts/music/etc. What does AP offer in this context that RSS does not?

2026-02-06 17:50:27 UTC

This is a test thread for <podcast:socialInteract> tag testing.
(I need an activity pub post url for us with test RSS feeds.)

2026-02-06 17:11:06 UTC

- reply

that makes sense. However since the feed URL can change at a later time, I believe from the v5 requirement is not actually required any more.
I'll just add validation that it is any uuid (36 chars, 8-4-4-4-12).

2026-02-06 16:22:41 UTC

- reply

LLM suggests v1, v2, v4, and v5.
v3 apparently is deprecated.
v6, v7 seem reasonable.
v8 is "custom" which maybe we don't want.

2026-02-06 16:18:48 UTC

@npub1rcn…dw4h @npub1j0h…u952 @npub1yda…h4tz I think we discussed this before, but wanted to confirm.
The PI docs said a podcast:guid should be v5, but you also said that shouldn't be a strict requirement.
I'd like to implement at least some validation on that column in our database. Do you think v4 and v5 would be enough? Or just validate that it follows any of the uuid schemas? (apparently there are 8 versions)

2026-01-29 22:38:00 UTC

- reply

that’d be great! Let me know if there’s a good day/time to chat.

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

2026-01-28 00:24:43 UTC

- reply

I have not even heard of it yet. Oh man...I have written so many skills already (well, telling LLM to write them mostly) forming a framework for the new monorepo. My impression is this may nullify or conflict with a lot of them. If it's a game-changer and the clear standards leader though, I wouldn't want to be left behind.
Also, I'm hoping to not stay confined to Claude / non-open source models. Can the plugin work with any model? I imagine it's just skills files?

2026-01-27 21:03:18 UTC

- reply

...then after Sonnet 4.5 seems happy with everything, I switch to Opus 4.5 to review the plans that Sonnet 4.5 created. When I do this, Opus 4.5 often finds gaps or issues that Sonnet 4.5 did not.
Once Opus 4.5 is happy with the plans, and I'm happy with the plans, I switch back to Sonnet 4.5 to run them (in parallel, when possible).
So far, the process seems to work well and save me a lot of money on API credit usage.

2026-01-27 21:00:42 UTC

LLM programming lesson learned...I have been using Opus 4.5 for large, complex tasks where I want the highest quality model available for the work, but Sonnet 4.5 (also by Anthropic) is also very good, and costs ~67% less $ than Opus 4.5 to use.
My latest approach is, when working on a large and complex task, have Sonnet 4.5 plan all the work upfront, and split the work into many plan files, and save them locally (target under 300 lines per plan, to keep them reasonably sized), then...

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)

2026-01-23 04:58:26 UTC

- reply

I still haven't tried Claude Code. Curious to hear how it goes.
My workflow has been Cursor using Auto mode ~95% of the time, and then for big or difficult tasks, I switch it to Opus 4.5, and in the rare case that doesn't work, I switch it to
I think I literally got 2 weeks of tedious refactoring, validation scripts, and documentation done today over 8 hours. Plus a feature on the website that I've been wondering how to do cleanly for like a decade (cont)

2026-01-14 16:45:02 UTC

- reply

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

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 09:17:50 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

awesome. Thank you!

2026-01-10 23:53:22 UTC

- reply

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 06:29:02 UTC

- reply

and yeah cost is a big concern. You can set budget limits with the LLM APIs though to prevent surprise overspending. Planning the prompts and trying to write the most efficient way to get the work done is important when cost is a constraint.
So far GLM 4.7 seems pretty good. Not Opus level but reasonably close, and 10% the cost is a huge selling point. Only been messing with it for a couple hours but I'm happy with it so far.

2026-01-10 06:23:53 UTC

- reply

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, which 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

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 04:48:24 UTC

- reply

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

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

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

@npub1rcn…dw4h @npub1hta…eg8a 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 @npub1yda…h4tz

2026-01-02 06:28:54 UTC

Podverse Alpha Web update 🛠️
Added the Artists and Artist page.
This is our first implementation of Podcasting 2.0 "publisher" feeds.
A "publisher" feed is an RSS feed that contains references to other RSS feeds.
In general, individual Music RSS feeds represent an Album or Track of music.
Publisher feeds allow artists to group all of their Album RSS feeds under their artist name.
https://alpha.podverse.fm/artists
https://alpha.podverse.fm/artist/FTTwdF9Ral
#podverse #pv2
https://cdn.masto.host/podcastindexsocial/media_attachments/files/115/824/074/377/301/317/original/a939c77565c13003.png
https://cdn.masto.host/podcastindexsocial/media_attachments/files/115/824/083/956/723/904/original/8b46dd9f8275b4e6.png

2026-01-01 01:43:54 UTC

- reply

nice! I’ll use it for deving

2025-12-31 22:53:58 UTC

- reply

I figured you’d already have a solution lol.
This would be perfect for our purposes. It provides a way for apps to display *lists* of publisher feeds grouped by medium type.
@npub1gfn…mk8j @npub1rcn…dw4h any thoughts on this? (I don’t know who else is creating publisher feeds to @)

2025-12-31 22:44:26 UTC

- reply

do you have any thoughts on this use case?
We'd like to be able to differentiate between Music Publishers and Podcast Publishers within the PV2 UX. I'm not seeing a way to differentiate without our own duck typing at this point.