So the pirate podcast feed at https://jumble.top/, maintained by our comrade @jumble@hexbear.net, is down again. Seems like the server is fully unreachable at the moment. I know there were issues with how Patreon premium feed URLs work, but it seems like that was solved for some time.
For those unaware, the way that site works is that people who are subscribed to some podcast’s premium feed on Patreon donate their premium feed URL to Jumble’s site, and users can then access a mirror of that premium feed for free through the site. I myself have donated a few feeds that I pay for, and I then sub to a few more for free through the site. As a chronic slop addict I genuinely feel the absence whenever the site is down, and I know I’m not alone based on the posts I see here every time it happens.
Given that downtimes have been getting more frequent, I’m wondering how me and other volunteers from the comm could contribute to maintaining and improving the stability of this service. So far it seems to have been mostly a one-person operation by Jumble, who deserves plenty of kudos for it (if there were other contributors I’m not aware of I apologize and extend the praise to them as well). The issue of course is that a one-person hobby project is inherently fragile and opaque. If Jumble - understandably - doesn’t have time to respond to incidents then we don’t get any slop and we don’t even know if we’ll ever get it back.
First off, for @jumble@hexbear.net if you see this: what’s the status of the project at the moment? How much time and effort are you able to give it these days? Would you be willing/able to involve other volunteer contributors? Do you have an idea of what that could look like? If not that, would you consider allowing someone else to fork the codebase and create another instance of it? Would you be willing to share at least some of the working feed URLs you’ve gathered so far? The URLs donated by other users should probably not be shared without their owners consent, but the leaked ones at least could be.
Unfortunately Jumble doesn’t seem to very active here these days, so as an alternative I’m calling out to the rest of you hogs to figure out if anybody else would be willing to help me work on a replacement for the slop feed. What I can offer is:
- a couple hours a week of my free time
- software dev and web hosting experience
- spare compute in my existing self-hosted setup
- a few bucks a month to contribute to running expenses
- a few podcast feeds I’m subscribed to
What’s needed on the other hand would be:
- contributors to help developing and maintaining a new site in a more collaborative/democratic way
- small financial contributions for web hosting costs (potentially not necessary if I just use my existing setup)
- more feed URLs to collectivize (this is the most important one)
To start with it would be good to know how much interest there is in this comm for this kinda service, if it even makes sense to put work into it at all, and if there are enough people still willing to donate feeds for it to be viable.
You must log in or # to comment.
I can contribute feeds and some $$ to help pay for hosting. I can help add feeds to the site as well if the process is simple enough
Awesome, thanks comrade! I’m working on the new site and I’ll post it in this comm as soon as it’s ready. The process for donating feeds will be the same manual one as for Jumble’s old site: PM me the link here and I’ll add it.
I have a couple feeds too, lmk when you need them
Here’s how to do it:
Get your list of feeds
Make it so their content is mirrored on your own storage periodically, this is so the hundred people using the replacement jumble service don’t all hit “Jim’s personalized private feed url” and blow up Jim’s spot. He was nice enough to give you his private feed after all…
Change the rss url to point to your store of content
Post the changed rss url on a website you host.
Tell people about it
Loss!
Yeah without having looked into it properly yet it doesn’t seem like a particularly complex project, unless there’s some trickery I haven’t foreseen. I don’t think you even have to mirror the actual media files, just the RSS feed with the signed download URLs. At least it doesn’t seem like Jumble did that, I can still download the audio for episodes my podcast app has cached. They told me previously that they do some processing to anonymize the feeds by removing identifiable components, so that will take a bit of work, but otherwise this could be done with just a shell script in a cronjob and an NGINX file server.
I’m just wondering if it’s worth putting the effort in the first place. Otherwise I’m just making a fancy mirror for the few podcast feeds I’m already paying for…
iirc the download urls are personalized so that patreon can monitor account sharing. Mirroring the media files is necessary to prevent the feed donors getting got.
I don’t know if that’s the case, Jumble told me some time ago in a PM that they had crosschecked a few different users’ feeds and the download URLs were the same for the same episode. Jumble’s site doesn’t mirror the media files, I can see that the download URLs for their feeds point to Patreon. In don’t know if that’s a risky way to go, but it seems to have worked for their site so far.
Also, the media files themselves are not always hosted by Patreon, sometimes it’s an external service like Fireside or iTunes, in which case it would be hard for Patreon to track account sharing across platforms.
I will defer to Jumble’s expertise. I had noticed unique elements to shared patreon urls in the past but evidently it is possible to strip those.
I wish Jumble came back online to share their knowledge. Looking at the download URLs of episodes my app cached from their feeds it doesn’t seem like they strip anything from the download URLs. They look just like the ones in the original feeds. And I don’t think it’s possible to do that without fully mirroring the files. I also don’t have the capacity/budget to create a full mirror of the media files.
Edit: see this comment by Jumble.
Just confirmed this is still the case with a feed a friend is paying for vs its version from jumble. The only differences are the link to the rss, (‘ns:0 link’ element), ‘pubDate’, and ‘lastBuildDate’. ‘enclosure url’ is the same across both feeds. And actually ‘ns:0 link’ isn’t identifying, it’s the same between my expired patreon sub and a paid sub. So nothing inside the feed looks identifying, only the link to the feed itself.
I don’t have the resources to mirror any rss right now, but have two feeds I’d like to donate to the revolution.
My I’m broke idea for sharing those:
- cron job that downloads the xml file
- xml file is uploaded to a) a public shared folder (on Mega, dropbox, whatever). Easy for the uploader to just add a sync to the cloud drive and share a link, but more complicated for the downloader to get those files into whatever rss app. b) some site that hosts text files for free. Not sure what an easy option for that would be. Maybe there’s a way to do it on a free blogging platform.
Fantastic, comrade, thank you so much for the research! I have a first version of this project almost ready, and it works almost exactly as you described:
- a cronjob downloads, processes and dumps the xml files into a folder
- a Caddy file server exposes those files through my sellf-hosted VPS
I think Jumble’s site was more elaborate than that, but as a quick replacement this should be fine. I’ll post the site here as soon as it’s ready, and you’re welcome to donate your feeds then.
I think so. But I like a project
You raise a good point. Perhaps there should be a layered approach with a distribution feed that is shared only with mirrors who then provide public feeds.
I don’t get it, why would you have that? Seems like that’s just increasing the number of failure points…
If you are trying to avoid downtime from inadvertent ddos then it makes sense to avoid having a single point of failure exposed publicly. There would be more failure points but ideally resulting in a hydra situation rather than a blackout situation. You could ofc also just pay for a higher capacity central server but relying on others to set up mirrors is a way to distribute the costs without needing any transactions.
Fair enough, that would make it more resilient, but it would also require trusting each mirror to not accidentally expose the distribution feed. Also only one person so far has offered even just to chip in to the hosting costs, let alone to host stuff themselves. I don’t think I can expect any help on that for now…
I have a PoC ready for a very simple implementation, but I’m waiting to add some extra safety precautions before I post it here.
I can contribute a couple feeds… unfortunately I don’t have the technical skills to do more
Several of the feeds had messages from the hosts saying that something was changing with the feed URL. I think it was in iTunes
What do you mean by this? Were there banners on the site? Or text posts in the individual feeds with warnings? My podcast app (AntennaPod) doesn’t show text posts in feeds so I might have missed them, but I think all the feeds I sub to come from Patreon.
Either way right now the whole website is down. It might be that a few broken feeds took down the whole site due to bugs, but it’s more likely that the VPS expired or something similar…
That was about a month ago when the feeds went down the first time. They came back on after that. They went down again briefly last week, and now they’ve been down for about a day
I guess it’s a good thing I hadn’t noticed it being down yet but I am also a big slop connoisseur. I have no technical expertise to offer but am quite interested in its continued existence and would happily contribute for any hosting costs
i thought that slop was neoliberal nonsense; am i wrong?
It was used for the podcast Chapo Trap House because it was liberal nonsense. Way too far right wing. They don’t even advocate for a violent revolution smdh.
This site originates from the chapotraphouse subreddit, where “slop” referred to podcast episodes, especially premium. It’s still a derogatory term, but it’s self-deprecating. The “post other peoples’ bad takes” community on this site adopted the name “slop” because people started reading about the racist origins of the original name, “the dunk tank”. Among the broader public, “slop” has come to refer to unwanted AI-generated content (which is to say, AI-generated content in general), and people here will use it that way as well.
