Sir Alecks Gates

@nprofile…td70 with plexor I attach a trace context to every event for this very reason
https://gitlab.com/plexo/plexor/-/blob/alpha-release/plexor-core/src/payload.rs
https://gitlab.com/plexo/plexor/-/blob/alpha-release/plexor-core/src/logging.rs

2026-03-27 03:30:25 UTC

- reply

@nprofile…td70 Strange that a peer discovery process would hang a publishing component.

2026-03-27 02:25:31 UTC

- reply

@nprofile…td70 I might have to finally spend some time with it....

2026-03-25 15:32:23 UTC

- reply

@nprofile…td70 I was using "glow" for a few things. Might try this.
https://github.com/charmbracelet/glow

2026-03-24 16:33:26 UTC

- reply

@nprofile…f3z2
I'd just use dd but clonezilla should get the job done.
Just make sure you set the boot disk in the bios to the external drive after cloning to make sure it boots and works, and validate it's booting from the correct disk.
Then wipe the existing internal disk and make sure it has a new UUID before installing onto it.

2026-02-20 21:52:08 UTC

- reply

@nprofile…td70
#careergoals

2026-02-20 21:29:00 UTC

- reply

@nprofile…td70 It's all good, I'm not waiting on you :)

2026-02-20 16:28:44 UTC

- reply

@nprofile…ax8n @nprofile…zv63 @nprofile…td70
We think it's just the list of endpoints. It hasn't been updated in 2 years. So I'm going to update the list and make it configurable.
Maybe have dave provide an endpoint of known good endpoints the user can opt-in to using.

2026-02-20 16:21:47 UTC

- reply

@nprofile…ax8n @nprofile…zv63
I'm going to try to get the hive writer changes done tonight to help mitigate this for podping.cloud.
Ultimately we are going to try to get a non-Hive solution in place. I have a decent plan but I want to hear what @nprofile…td70 thinks about it first.
Unfortunately, the next few months are going to be very busy for me between work and moving states. And I'm just a volunteer with all this.

2026-02-20 16:14:10 UTC

- reply

@nprofile…ax8n @nprofile…zv63
Primarily JSON-RPC, which is over HTTP.

2026-02-20 16:11:23 UTC

- reply

@nprofile…ax8n @nprofile…zv63
I mean, you can use spurlock's websocket if you want. I try to promote more decentralization.

2026-02-20 16:08:41 UTC

- reply

@nprofile…ax8n @nprofile…zv63
It's ultimately some really basic JSON RPC calls.
You can follow the logic here. Don't need any special hive library.
https://github.com/Gates-Solutions/podpingd/blob/main/src/hive/jsonrpc/block_api.rs
https://github.com/Gates-Solutions/podpingd/blob/main/src/hive/jsonrpc/responses.rs
https://github.com/Gates-Solutions/podpingd/blob/main/src/hive/scanner.rs

2026-02-20 15:49:38 UTC

- reply

@nprofile…zv63 How are you submitting to Hive?

2026-02-20 01:05:17 UTC

- reply

@nprofile…w0mx @nprofile…ax8n
> @agates can’t even be bothered to contribute
What? Sorry, James, that's just disingenuous. And I absolutely have no ill-will against you in this matter.
I've merely stopped caring.
If I need to resurface the issue for something in the future, I will.
However, I imagine it will resurface on its own. It always does.

2026-02-09 14:54:34 UTC

- reply

@nprofile…td70 i'm all about traces these days, probably because i'm using so much async/multithreaded code.

2026-01-30 17:37:01 UTC

- reply

@nprofile…td70 This "standard" leaves out any other apps that want to view them because the metadata is not stored somewhere defined in the feed.
It's not particularly impressive.

2026-01-25 20:19:13 UTC

- reply

@nprofile…vfcp I mentioned workspaces already https://podcastindex.social/@alecksgates/115947745147136523

2026-01-25 20:16:31 UTC

- reply

@nprofile…vfcp Have fun with that in 10 years :)

2026-01-24 02:34:33 UTC

- reply

@nprofile…vfcp I would avoid them if you can, but look into MCP or whatever the latest thing is.
Something I've been thinking about trying is compromising by having a directory that has all of my projects in it, and adding context as to which project does what.
You could probably have a repo of repos, with a bunch of submodules pointing to all of your other repositories. That might work rather well, as long as you instruct the agent how to handle them.

2026-01-18 18:57:57 UTC

I stopped using Tailscale on my phone and the increase in battery life has been dramatic.

2026-01-18 18:54:04 UTC

- reply

@nprofile…f3z2 I honestly haven't used it seriously since it was still XBMC

2026-01-18 18:52:26 UTC

- reply

@nprofile…vfcp it could be automated by building your own agent interface. Or an agent orchestration system. But I feel like it could wrong fast.

2026-01-13 20:38:48 UTC

- reply

@nprofile…kdmg castopod and peertube feeds will for sure.

2026-01-10 20:04:46 UTC

- reply

@nprofile…c2kf just brain dump idea that came to mind based of the last pc20 episode

2026-01-10 19:40:32 UTC

Slop Detector leaderboard

2026-01-10 16:59:40 UTC

- reply

@nprofile…vfcp I'm using Gemini CLI. It's been great. Gemini 3 especially.

2026-01-09 07:30:35 UTC

- reply

@nprofile…vfcp @nprofile…td70 @nprofile…7td6 @nprofile…0fv8
What's your log tracing look like? I recall that was a major pain point in the first podverse server.
At the very least, opentelemetry + some OTEL receiver would get you a long way, you don't need a ton of log retention.
SigNoz looks cool. I sort of want to try it.

2026-01-07 03:36:19 UTC

- reply

@nprofile…vfcp Yeah, I think the way we designed it, the server side UnifiedPush code was 99% the same as WebPush. The only addition was the UP devices.

2026-01-07 03:04:29 UTC

- reply

@nprofile…vfcp Let me know if you want help with that again (not that it's super complicated)

2025-12-30 21:46:59 UTC

- reply

@nprofile…25h2 @nprofile…5yxp @nprofile…y5nd Yeah, we should probably update the specifications. One of those things where the reality didn't match the designed perfection.

2025-12-30 21:32:19 UTC

- reply

@nprofile…25h2 @nprofile…5yxp @nprofile…y5nd
I think this was just one of those things we decided wasn't necessary to strictly enforce because a unique UUID was good enough

2025-12-30 20:50:58 UTC

- reply

@nprofile…9ap7 I've been using split view in Edge for work for a while. Very useful.

2025-12-18 16:27:49 UTC

- reply

@nprofile…25h2 @nprofile…5yxp
Actually I recall I had like thousands upon thousands of episodes queued up in podverse previously. I can't remember if it was a bug or just my negligence

2025-12-18 16:13:55 UTC

- reply

@nprofile…5yxp @nprofile…25h2
Personally, I think it's more intuitive to queue up everything after I click play on a track.
But the UX difference indicates some people don't like that. So I wonder if it's better to be user configurable so people can turn off auto queuing.

2025-12-18 16:13:10 UTC

- reply

@nprofile…5yxp @nprofile…25h2
"If I have an album up and I click on a track, should it play the next one automatically?"
I actually noticed a quirk in behavior in some music players I've been testing.
Some apps automatically queue up every song after the current track in a playlist or album, while other apps intentionally don't queue anything unless you click play on the playlist or album itself.

2025-12-17 03:31:09 UTC

- reply

@nprofile…wsn5 @nprofile…25h2 @nprofile…5yxp
We should just start using this :)

2025-12-13 17:34:36 UTC

- reply

@nprofile…y5nd also it's funny how I'm also writing my own rust message queue system 😂

2025-12-13 02:18:39 UTC

- reply

@nprofile…y5nd hmm shouldn't there be a parse queue in front of the parser?
Also how much of a bottleneck is it to generate SQL statements? In my experience, if we ingest JSON strings as serde structs, it's pretty trivial to use those structs directly in an sqlx insert.

2025-12-11 02:35:12 UTC

- reply

@nprofile…sxxr
The irony is LLMs are great with rust because of the type system.
@nprofile…y5nd @nprofile…wsp3

2025-12-11 02:34:03 UTC

- reply

@nprofile…25h2 This was the use case that caused me to come up with remoteItem, though it doesn't support it. I would repurpose that if we think it's valuable.

2025-12-10 16:08:38 UTC

- reply

@nprofile…5yxp @nprofile…amf2 @nprofile…25h2 @nprofile…ezja @nprofile…wsn5
Week old news, y'all got hacked already 😉

2025-11-07 16:20:29 UTC

- reply

@nprofile…amf2 @nprofile…myyh @nprofile…vwpk @nprofile…l26x
How dare you suggest an open standard with no middleman!
Let's go further and add KYC to prove the sender's identity.

2025-10-19 01:18:05 UTC

- reply

@nprofile…y5nd I think a growth mindset is important to have, but it's a lot more specialized than we're led to believe.
It's not healthy to outright reject such specialization, either. Some things *do* need to scale, even if one can't comprehend why. We all have our own problem domains.
Desire for profit is another matter. People will do anything to retain their social status and power, and I don't see that stopping any time soon.
Maybe the cycle of empires is inevitable?

2025-10-09 03:52:25 UTC

- reply

@nprofile…h82d I like them for basic things on Kagi.
I guess there's a threshold where I have my own confidence level of the response vs the criticality of the information.
I only really use Kagi though.

2025-10-07 18:21:33 UTC

- reply

@nprofile…3ram would be nice to see playlists like this show up in other apps :)

2025-10-06 11:39:27 UTC

- reply

@nprofile…3ram updated with what?

2025-10-05 00:00:35 UTC

- reply

@nprofile…3ram The playlist medium should be podcastL, no?
Title could be "Golden Globes Nominees 2025" or something, as an indicator.
https://cdn.masto.host/podcastindexsocial/media_attachments/files/115/318/607/536/836/681/original/259980f9c14f5f6e.png

2025-10-03 16:15:46 UTC

- reply

@nprofile…pvn3 @nprofile…vu32 @nprofile…h82d @nprofile…zn42
It's still online but maintenance issues like this is exactly why I'm going to turn it off lol.

2025-10-03 00:42:53 UTC

- reply

@nprofile…h82d everybody panic!!

2025-09-30 15:17:24 UTC

- reply

@nprofile…6746 done

2025-09-10 02:45:34 UTC

- reply

@nprofile…6746 I don't know, I haven't seen a single critique of my proposal. Yeah, it needs to be finalized and I would like to help see that through if someone wants to take it seriously.
But the concept of providing the metadata at an HTTP API is incredible simple and about as unopinionated as you can get.
I'm a great mentor, for what it's worth. I would be more than happy to play that role as I have in the past.

2025-09-10 02:21:31 UTC

- reply

@nprofile…6746
I think it's pretty obvious, as a Podping maintainer, that I'm confident in Hive as a solution for the moment.
However, I don't want to bet on it for everything. The same goes for any protocol outside of RSS, including ActivityPub, XMPP, and the N-word.
I'm formulating long-term plans on how to move Podping away from Hive as a central point of failure while keeping the same features. So I think I would argue another dependency on it would not be great for podcasting.

2025-09-06 14:58:19 UTC

- reply

@nprofile…6746 I'm kind of enjoying the "I told you so" moment.

2025-09-05 23:16:53 UTC

- reply

@nprofile…6746 @nprofile…h82d Done

2025-08-19 20:31:33 UTC

- reply

@nprofile…wexf Cool! Wooo Podping

2025-08-17 17:53:51 UTC

- reply

@nprofile…ywdy @nprofile…3j0y @nprofile…kz02 @nprofile…h82d @nprofile…jsxc
I had disabled podcast RSS feeds for remote servers and only had them at the channel level (not account, server, or category). It was a separate implementation. There were already two different RSS "formats" so I added a third.
Also, the concept was very new at the time and I was trying to avoid controversy. You know how FOSS gets.

2025-08-17 01:53:23 UTC

- reply

@nprofile…ywdy Every PeerTube instance has a different nginx config in addition to the application itself, so it would have to be called out as a configuration task in their installation/upgrade/release notes.
This is the default config https://github.com/Chocobozzz/PeerTube/blob/develop/support/nginx/peertube
@nprofile…4vd9 would know more.

2025-08-15 23:11:49 UTC

- reply

@nprofile…6746 @nprofile…pvn3 @nprofile…wuyh
So this is where a process shift comes in. With metaBoost, this app would look up the metaBoost API from the feed and look for boosts the API directly. Anyone can do it.
(If we want private boosts, we can add support for that)

2025-08-15 23:05:10 UTC

- reply

@nprofile…pvn3 @nprofile…6746 @nprofile…wuyh yes Helipad would likely be a candidate to be the public server. We would need to experiment with tailscale funnels or cloudflare's equivalent

2025-08-15 23:02:28 UTC

- reply

@nprofile…pvn3 @nprofile…6746 @nprofile…wuyh
Hmm well the idea is the boosts will optionally be publicly available so you wouldn't need to send them to multiple places. 3rd parties and apps would pull the boosts to list directly.
But, yeah we can probably have a default URL and specify directly per recipient.

2025-08-15 22:55:46 UTC

- reply

@nprofile…6746 @nprofile…pvn3 @nprofile…wuyh
If you're actually using this we can narrow down the spec to whatever makes the most sense. That initial spec was just a draft to get the conversation going.
I think I want to propose using jwt key sets for signing since payment proofs on lightning are so difficult (and providers won't implement what would be required).
Also I think we might want to separate the metadata from the payment recipients. That concept only existed because of keysend

2025-08-08 14:18:36 UTC

- reply

@nprofile…ywdy @nprofile…j90y
https://github.com/Chocobozzz/PeerTube/pull/5487

2025-08-07 13:06:06 UTC

- reply

@nprofile…6746 most of them could be easily modified to use RSS, to be honest.

2025-08-07 12:20:53 UTC

- reply

@nprofile…6746
PRESIDENT, nice taste :)
A lot of the open source players do this on Android. There are a lot of cool ideas. Almost all of them just use YouTube as a backend.

2025-08-06 18:58:37 UTC

- reply

@nprofile…6746 I think technically publisherL as a publisher playlist is technically available.
But otherwise we would need a unifying channel or something which never got off the ground.
Or, maybe a publisher of publishers?

2025-08-04 16:09:23 UTC

- reply

@nprofile…6746 @nprofile…ayy4 Yes I think it's a no-brainer for apps to offer users to share their music playlists publicly like this. It's something I've been thinking a lot about, just have tried to not get distracted so I can finish some Podping stuff.

2025-08-04 15:32:27 UTC

- reply

@nprofile…6746 @nprofile…ayy4 I'm interested, if I can find the time soon I might help you test

2025-08-01 15:12:08 UTC

- reply

@nprofile…6746 @nprofile…h82d it probably is.

2025-07-31 16:09:19 UTC

- reply

@nprofile…ywdy thinking these aren't protected by PCI DSS compliance.

2025-07-21 17:49:25 UTC

- reply

@nprofile…ywdy Could be useful with the Podping JSON schema.
https://valibot.dev/guides/json-schema/
Kinda reminds me of pydantic in Python.

2025-07-20 20:13:58 UTC

- reply

@nprofile…3ram that's Steven Crader's service. I maintain the python one. But also have a rust version I'm working on (podpingd).
The python/JavaScript ones shouldn't be directly affected, but anything not expecting the new mediums or reason codes will need to be updated.

2025-07-20 19:57:26 UTC

- reply

@nprofile…3ram how do you watch and push Podpings?

2025-07-20 15:28:09 UTC

- reply

If I don't get any feedback on this, I'm probably just going to do it :).
Need to get radio mediums added to Podping as well for @nprofile…3ram

2025-07-17 13:30:48 UTC

- reply

@nprofile…7xwn @nprofile…6746 I tend to agree with this... A lot of bros showed up trying to make easy money, making it worse for everyone else. It will pass.

2025-07-14 21:30:04 UTC

- reply

@nprofile…cguj Cool, I've had a lot of neat ideas since I came up with that medium. Newsletters by email kinda suck.
If you have a publisher feed it would be perfect to add there. Maybe podroll too? But not sure that would have the intended affect.