I recently purchased some very rural land with a cabin that I put on AirBnB. The address is a different road than the driveway is and all the maps took drivers around 1 mile away with no way to find the actual driveway. I had to go through the process of fixing this on Google, Apple, and open street maps. OSM was by far the easiest and my updates propagated within minutes. Google was the second easiest but still took me deep into the “feedback” menu and then a week or two of waiting. Apple was by far the worst as I had to sign up for an Apple business account, upload incorporation docs, and then go through multiple rounds of back and forth with some approval agent.
Another option for apple would have been to just wait for a few months. They are downstream of osm but updates are slow. Likely your "propagated within minutes" was misleading. Almost no app, certainly not a mainstream app your customers might use, is reading osm live. They're all days to months behind on periodic batch updates.
Months behind is not fair. Apps like Komoot, OSMAnd or CoMaps have monthly refresh cycles. Others are shorter and yes there are longer (mostly obscure) update cycles such as apple maps, Uber, https://transportnsw.info/
In my experience, the bigger the app, the longer the update cycle. They likely have a QA and integration process to deal with how osm combines with other sources. Something like osmand is basically just an osm wrapper, so they can be fast.
I don't know what the heck map data Uber uses, but it always sends cars to a road behind my house which would require them to drive through a swamp and then up a heavily wooded hill to get here. OSM had the exact same issue, but I fixed it a couple years ago and it propagated just fine to the various OSM nav apps a long time ago. Neither Google nor Apple maps ever had the issue. But now DoorDash, which never before had an issue, last night attempted to send the driver up a gas line easement to get here. No idea what that's about. The entire issue appears to be that I'm on a long shared driveway and the various nav systems are afraid of it.
But data centers would still be built outside of your country, in China for example.
So if you really believe AI is gonna devalue your skills, you'd rather want data centers to be built in your country and potentially get dividends/UBI rather than in another country and have your job decimated by international competition and get nothing in return.
> So if you really believe AI is gonna devalue your skills, you'd rather want data centers to be built in your country
This makes no sense. It's just as true for the reverse assertion "If you really believe that AI won't devalue your skills, you'd rather want data centers to be built in your country". If it's true for both assertions (very big IF), then why state it?
> rather than in another country and have your job decimated by international competition and get nothing in return.
Again, this assertion makes absolutely no sense. You don't see why citizens would protest against local companies, companies which are the recipients of the value created by local taxpayers, devaluing their skills?
It's also really odd that you think they'll get anything in return. What makes you think this? Has this ever happened in the past, in the US, when too-big-too-fail companies laid off all their staff? This unwarranted optimism is the stuff of science fiction.
=================
TBH I can't really tell where you're going with this. You appear to be implying that either:
1. The objectors don't really believe that AI will devalue them, or
2. That you believe that AI won't devalue them, or
3. Perhaps you want to indicate that their jobs are going to be decimated anyway, thus they shouldn't be objecting anyway.
Which techno optimism? In a way I'm less optimistic than you, I think stopping AI is not even an option on the table.
As for UBI, several AI leaders have talked about it, and it does seem to make sense, it is not in their interest to have a revolution.
As for the electronic waifus, they are coming for sure, the digital ones are already here.
I, and they (they said that AI would devalue their skills) think that their jobs are going to be decimated by AI.
They can stop AI in the US, but how are they gonna prevent China from making their jobs redundant with AI? Even if they put enormous tariffs on China, they would still lose all business outside of the US.
In the best scenario, after some recession they could stagnate if they stay isolated while productivity in the rest of the world explodes.
>It's also really odd that you think they'll get anything in return. What makes you think this? Has this ever happened in the past, in the US, when too-big-too-fail companies laid off all their staff? This unwarranted optimism is the stuff of science fiction.
Don't you see the long term trend in every society to increase public spending to support their population more and more over the last 100 years?
I do not think it is in the interest of the AI overlords to have a revolution, but of course there are reasons to be fearful. Yet again, stopping AI in the US will not stop AI in the rest of the world, there is an arm race going on, one that seems almost impossible to stop, unfortunately.
There are very good reasons to call for more regulations and prevent data centers from being built too close to residential areas, or not using renewables, or plenty of other reasons. But being opposed to every data center because you are afraid for your job seems like a mistake to me. Unless you can prevent all of the world from using AI, in that case yes, it does make sense, I just think it is impossible.
Okay, in that case you're still being irrational and illogical, expecting people to go to the slaughterhouse quietly.
Of course they are going to protest. Whether it makes a difference because some competitor on the other side of the globe could theoretically do the same thing is irrelevant - people on the way to the slaughterhouse will not go quietly.
Oh it is understandable that they are pissed and protesting, I'm not surprised they do, nothing is illogical there when you know human nature and people emotional state. I'd be pissed too. I am pissed too, as a programmer that has his job done more and more by AI.
However it seems to me that you are the illogical one, or perhaps we are having a different conversation, you talking about their feelings, and me talking about the consequences of their choices.
There are two choices, go to the slaughterhouse (probably losing your job due to inability to compete internationally and getting nothing in return), or take the different option, which is less bad, and might even be good (probably losing your job and potentially getting money without even working).
> There are two choices, go to the slaughterhouse (probably losing your job due to inability to compete internationally and getting nothing in return), or take the different option, which is less bad, and might even be good (probably losing your job and potentially getting money without even working).
In both choices they are getting nothing in return, nor have they even been promised anything in return.
I mean, if they had been promised something in return, I can understand your PoV, but there are no promises that going to the slaughterhouse for the local companies gets them something in return.
Indeed nobody has promised them anything, but it seems obvious to me that it's better that the technology that is gonna take your job is controlled by your own country rather than by another country, one that isn't even particularly friendly to you.
I don't think there are a lot of countries that are happy that they are not the US nor China so that they don't have to deal with AI. Some of their population perhaps is, for now. But most would love to be able to catch up.
Just a reminder to everyone, always buy the second version of any new product category, even from Apple. The first duo will look primitive in a few refresh cycles. It does look really cool though!
Considering how many comments I regularly see on Hacker News from people who still love their M1 MacBooks, I’d say that’s not a particularly strong maxim.
The AirPods, AirPods Pro, AirPods Max, HomePod, MacBook Neo, and AirTag are also all devices that have had well-received (if not always breakout) 1st-gen releases.
Yeah although my M4 is more powerful I somehow don't perceive it much and it's so heavy and gets hot. My aging M1 is still snappy for most workloads, the battery holds well, it's light. Only thing I resent is the stupid touch bar instead of proper F keys.
But I'd say it's the exception. In general it's better to wait for later generation.
The M1 Macs were a bit different because Apple had already been designing their own CPUs for years, and the machine designs themselves were almost identical to previous iterations with Intel CPUs. The only thing they needed to get right was the software, which they did (and which could have been updated after the fact if they didn’t).
Listen, I'm not saying you're wrong, but I will say I hope I'll at least get to update this comment down the line while holding that gorgeous looking phone in my hand
I get the sensibility of that advice but isn't that the case no matter when you buy it? I mean, they announced a nice iPhone Pro, optimized over multiple generations, and then immediately made it look primitive with the Duo. :)
I'm drawing a different conclusion than you about that upgrade. Sorry! I definitely understand your advice; it's especially the best option when you are financially risk-adverse or pessimistic about technology but need to participate.
I think all he big labs are pretty explicit about when they do and don't train on customer prompts. Is the accusation here that OpenAI trained on prompts when they claimed not to? Or were the mathematicians using one of the interfaces that allows OpenAI to train on the customer data?
As a long time plex user with a lifetime pass, I always enjoy seeing the good progress with Jellyfin as it will increasingly give me a soft landing for when Plex finally does something to make me move. At the very least it will hopefully keep some of plex’s most user hostile behavior in check.
As an aside, I recently started having Claude help me with configuring my *arr stack and it has helped tremendously. I used to have to babysit and manually intervene on maybe 10% of requests, now it’s near zero, plus I got audiobook requests and subtitles working great. Granted a big part of the increased success was signing up for a Usenet service but of course Claude was able to configure it all and make it work. I was a little worried Claude would demur in the face of the task, but it was more than happy to help me acquire media in this way
I left Plex for Jellyfin + Infuse a couple of years ago, and I'm not looking back.
But the sore point right now is outside access, which Jellyfin really isn't made for. I set up Tailscale so I can connect Infuse to Jellyfin remotely. But that's not exactly as smooth as Plex. I can't give friends and family access, either, which was dead easy with Plex.
In terms of the UX, Infuse isn't very good, but it does the job okay. The built-in subtitle fetcher is amazing; that stuff never worked for me with Plex.
Is Plex/jellyfin even necessary for Infuse? I've been using Infuse for years now and it has direct access to my NAS to look for media content and just directly plays it. It's a such an amazing app and it plays everything I throw at it without any buffering or transcoding issues. The Pro version pretty much has all the necessary codecs, including all the dolby things for hdr/atmos. It just works so well.
Yeah I have Infuse only, and am happy with it. No need for jellyfin or whatever else. Honestly Infuse may be one of my best app purchases of these past 10 years…
Best to specify the formats you can play in the *arr tools than rely on client transcoding. I aim for media I can direct play and the filters in radarr and sonarr make that pretty easy. I also use tdarr to clean up / convert things I couldn't get a direct play file for.
Yep. Worth every cent. The Apple TV experience of Infuse is night and day with Plex. I used to constantly struggle with Plex buffering and crashing on my local network with the Apple TV app. Infuse just points to a file share on my NAS and streams beautifully, even at larger 4K files.
I used Plex for years. I’d never go back from Infuse now, and I’m happy to pay that subscription to support its excellent development.
I have the exact same setup. I found Infuse a PITA to begin with. Now it works pretty dang well, even remotely. One big irritation though is the subtitles are always on by default, on Apple TV it's an issue as I can't find a way to disable the default haha! Sometimes there is also a delay between when Jellyfin can play a new video in the browser and when Infuse can play it even though it can see the file.
I do find Infuse significantly better than the Jellyfin apps though.
>But the sore point right now is outside access, which Jellyfin really isn't made for. I set up Tailscale so I can connect Infuse to Jellyfin remotely. But that's not exactly as smooth as Plex. I can't give friends and family access, either, which was dead easy with Plex.
I'm confused about this bit. I installed jellyfin to provide outside access to me and my family/friends.
It wasn't difficult to make it available on the 'net, setting up new users is pretty simple and jellyfin has clients (both first and third party) for most platforms.
In fact, I went from never having used jellyfin to giving external access to friends/family within a few hours.
I'm sure I'm missing something, so if you wouldn't mind, I'd be very interested to hear your experience.
I don't believe in Jellyfin being secure enough to put on the public Internet. In my case, Jellyfin is on my home network, and the box has access to a lot of files and devices.
>I don't believe in Jellyfin being secure enough to put on the public Internet. In my case, Jellyfin is on my home network, and the box has access to a lot of files and devices.
Fair enough. IME, Jellyfin and (just as importantly -- at least for me) the distribution (Fedora) I use have been very good at patching/releasing updates to both the clients and servers when vulnerabilities are found.
Initially, I stuffed jellyfin (and a bunch of other stuff) behind HAProxy[0], which was fine. Now I don't even bother.
It's been at least three years since I put up my jellyfin server and haven't had any issues at all.
All that said, I understand your trepidation and am glad you found something that works for you!
I think you overestimate the technical capabilities of my family or their eagerness to install software. When I ran Plex, and nobody ever installed the Plex app like I recommended to them, even though it's one click away in all the app stores.
I mean, if they aren't even willing to download the one app that actually gets them the content they want to see, then *no* software will ever be something that this type of user will be interested in. Not everyone has to use an app, so why try and cater to an audience who clearly has zero interest? You're not gonna magically make it somehow even easier than "go to app store, download app" to the point that they'll use any piece of software.
No, the browser version of Plex is much simpler because you can just send a link. The receiver only needs to go through a simple "type your email and password" sign-up process, and that's it. Nothing to install, works everywhere.
Tailscale is exponentially worse, though. Not only would I have ask them to go to Jellyfin and sign up and log in, I would have to explain that they can't do that until they've installed the Tailscale app (which I have to explain because most non-techies do not know what a VPN is). Tailscale requires its own account, so they have to first sign up to that, and I have to add them as an external user first.
You do need a DNS service with API access for automatically creating/renewing the certificates for things that are only served over tailscale though, since they aren't publicly accessible for the letsencrypt servers to talk to.
Caddy is great, but doesn't solve the problem I was talking about on its own.
Caddy will get SSL certs automatically if it is exposed to the internet and DNS for that domain points to it, but if you are routing over tailscale exclusively Caddy can't automatically validate the domain names over ACME. There is a Caddy extension for doing DNS based validation, but that puts you back to needing a DNS provider with an API.
> There is a Caddy extension for doing DNS based validation, but that puts you back to needing a DNS provider with an API.
Yea, for clarity this is what i was describing. I use Porkbun's API and the Caddy instance isn't reachable from the internet, only local and over tailscale
Same, it's fantastic. As long as you don't need to support as many requests per second on your hardware as something like NGINX could, Caddy is the way to go.
Has cloudfare become any more strict with passing large amounts of video data through them? I heard they were kind of against it, but haven't enforced anything.
I guess it's less full featured as Plex or Jellyfin (it still more than covers the bases for video: format support, different video/audio/subtitle tracks, tv series recognition, skip intros, Trakt.tv etc.); but then again all it requires is any online storage, and the Infuse app (either on iDevice or AppleTV).
Yeaah, second that! I use it on the AppleTV and across the rest of my devices! Works like a charm, and coupled with Tailscale I can access everything on my Nas everywhere :) I have it configured to use Jellyfin as the source, runing on said nas
It's still in beta unfortunately but this is Hacker News and it's okay to share software before it hits prod. People want to be ahead of the curve and it looks promising.
Absolutely, I never said it's not :) I'm just not a fan of the Discord distribution channel - but the reply above you said it's on AppleTV, and I might try that!
Does it do bitstream DTS and DV on Apple TV? If so maybe I’ll have to try. Also am I going old or is the whole join our discord to see what it’s about/download is the new norm on the web? My click through rate on that is about 0.
Yes it really does, and I intend to switch eventually but from what I heard the last time I looked into it, the end user clients are a big area where plex still beats Jellyfin. Since that’s where my non technical friends and family mostly interact with the system, it’s the area I need to be the most polished and hands off.
Yes, this is the big one for me. Seemingly unlike a lot of folks, Plex is my main music server. Years ago I ripped all my CDs (1800+, IIRC) and Plex + Plexamp is really, really, really good for listening. I haven't yet found anything else that works as well.
I use Plexamp on desktop/laptop computers and then the Android and iOS apps for on-the-go music. I tried Symfonium [1] because it seemed promising and would give me a Jellyfin path, but it was weird and kinda buggy and ate my phone's battery.
I'm hopeful there'll be a good replacement available eventually, but for now, I just have to stick with Plex.
I use (and contribute to) Finamp the Jellyfin PlexAmp alternative and it's just done a 1.0.0 release which brings it up to parity with PlexAmp for me. I've previously been a long time PlexAmp user but FinAmp now does everything I want it to.
Looks like it isn't on the android play store though? Any info on this?
Play tells me there is a version for a prior version of android, which is won't let me install. Not helpful...
Yeah PlexAmp is amazing. Feels like it's the only music client around these days that has a functional "shuffle by album" option, instead of treating my library as only a collection of individual songs.
> Feels like it's the only music client around these days that has a functional "shuffle by album" option
foobar2000 has this!
After trying dozens of music clients (including PlexAmp) I eventually settled on simply using foobar2000 pointed to an smb share - nothing else seems capable of properly handling playback by folder structure. (e.g. I've got all my Led Zeppelin albums in the "Led Zeppelin" folder, and I want to just select it and play their discography in album release order.)
My audio needs have been satisfied for many years with subsonic, then airsonic, and now navidrome. They all expose the same subsonic API, which gives quite some flexibility on the audio player client to use. At the moment, Tempus (Android) seems well maintained, but I miss from DSub the lower clarity as to what is cached for offline consumption.
yeah I'm in a boat of similar shape. plex and plexamp is very simple and plug and play - it handles the IP stuff for remote play, which is the tough part for me.
I don't mind lacking features in the server or the client honestly, if I can find and play an album or playlist I'm good. But the connectivity piece can be hard if you're not familiar with self host stuff. I tried to do Jellyfin and Tailscale and some client and it... kind of worked? But it was definitely not as seamless.
I do hope every year that it's my last year paying for Plex. Maybe this year!
I'm fortunate that somehow I ended up with a lifetime PlexPass for... I want to say like $50 a bunch of years back?
Because yes, the remote stuff -- and PlexAmp just working the same wherever I am -- is the huge part of it. Music sits on my NAS and things... just work.
I also really like how easy the caching and offline mode is, because I frequently find myself driving in areas without mobile data service. I keep a couple favorite albums downloaded to my phone, then knowing the app will cache the next ~2 albums of queued tracks generally makes it Just Work.
I don’t think it’ll work because Plex gates everything with a sign in to their online accounts, so auth would have to go through them and there might be server-to-Plex verification going on too.
That said, I’m very interested in this space - I’ve been building a music server that provides compatible APIs for Synology and QNAP mobile apps, with a few more on my todo list. I’d love to include Plexamp so if it is possible please let me know!
I would be interested to hear from anyone who set it and can still access Plex over their LAN in a network outage. I have set this multiple times according to docs, and it worked only once ever, thereafter apparently resetting to online login. Of course, they do not make this an easy setting to change!
Performance was also a big issue for me the last time I tried running Jellyfin on the same hardware as my Plex server. I’m happy to see that has been their focus.
Its hard to tell if the folks at Plex understand the position they're in. They definitely are dancing right up to the line of pushing people too far. I could see them trying to boil the frog with stuff like putting their garbage free streaming stuff front and center more. Like so there won't be one specific thing they do that makes everyone realize its time to leave.
I think one silver lining is that whenever it does reach a tipping point there's going to be a ton of support for projects like Jellyfin, so hopefully we see more feature parity - in particular for shared access for friends and family.
I left this year when they told me I'd have to re-buy my lifetime pass. I suspect they rolled it back, but as you said that was the last straw. All of these streaming nonsense that keeps getting re-enabled, totally troublesome to help friends with the same settings
I couldn't help but think with all the ads they were serving me directly, that they must be selling my data as well. That's not the point of self-host. No thank you.
Jellyfin still missing download w/ transcode, but with space getting so cheap I don't miss it. I do need to improve the music side though
they'll eventually come out with a completely new overhauled version of plex that's sub only but your lifetime pass will still be good for "plex classic" which they'll stop supporting
The Plex rug-pull is just around the corner. I've been running Jellyfin pointed at the same content as Plex for some time now. I still find some rough edges on Jellyfin but they smooth out a few more with each release so I'm confident I'll be just fine when I need to switch. The clients are getting better too, I'm using Infuse on Apple TV and Moonfin for everything else.
Or if like PlayOn, they’ll simply replace your legacy lifetime license software with their new version that no longer does the core functionality that has to be kept up to date — recording different providers — without upgrading to their new extra-special monthly cloud licensing that replaces your “lifetime” license you bought to support their fledgling business. One thing is almost certain, any software claimed to be “Lifetime” will become a lie if given time and one should assume “lifetime” in software almost never means more than 5-10 years before a new executive has a bright idea to generate more revenue.
Or just end the lifetime tier and make people pay like fastmail did. If even the HN beloved fastmail would do it I pretty much expect it from everyone.
I mean, I have a lifetime license for Plex, but even I would happily pay some nominal amount every year if Plex stopped making their software worse intentionally.
Maybe I'm naive, but I feel like if Plex was going to do anything to existing lifetime subscriptions, they would have done it by now. I think we're fine for now.
I (also a lifetime Plex Pass user) decided to try it out ~6 months ago, and I've switched over. I haven't had any gaps in functionality yet, though YMMV of course. While it isn't better than Plex (which is still fine for now), it isn't worse either, and I trust the Jellyfin devs a lot more than the Plex devs at this point.
I switched to jellyfin about 6 months ago and I can say it is definitely worse. I’m constantly fighting to fix thumbnails, fix episodes to be in the same show, force rescans to get it to even recognize a show that already exists. I never had any of these problems with plex.
Plex also defaults to transcoding which fixes a lot of gaps with clients that don’t support certain codecs. I’m often running into issues where infuse can’t play DTS so I need to reencode before I can play the movies.
Also in the process of ditching plex I switched to audiobookshelf for my audiobooks and I have to say it’s also crap. It’s constantly refusing to start playing, or pausing for a really long time before playing.
All that being said I’m not switching back. It’s been a pain but it was worth it.
The default recommended folder naming settings are different for jellyfin from plex if I recall correctly. Personally I just let the arr app download metadata and let jellyfin scan local nfo files only for metadata so have not had a problem for a long time.
I don't recall the folder names being different, but by default Jellyfin uses different metadata providers than Plex. That means that things like numbering of specials (think BSG and the webisodes), or shows which have a different aired vs DVD order (i.e. Futurama) might differ between the two. For those, you have to either tweak your episode numbers or just tell Jellyfin to use the metadata provider Plex does (which is what I did for those edge cases).
I moved from plex to jellyfin and got 10% of the way through organising the films the jellyfin style and then just moved the rest of the video files into a folder called 'ok' and it worked fine. try the lazy way first and see how it goes
> Plex also defaults to transcoding which fixes a lot of gaps with clients that don’t support certain codecs.
Maybe I'm misunderstanding this, but as far as I can tell, Plex defaults to direct play enabled, and will direct play unless it's not supported by client. (Or other constraints, like limited bandwidth settings, etc.)
That's fair. It got most of my stuff without issues, but there were a fair few I needed to fix.
However, compared to Plex, I've found the remote experience (through Tailscale) to be a lot easier and better performing than the Plex equivalent for me personally. I never have any issues connecting externally now, whereas I did all the time with Plex. Maybe I just needed to get Plex running over Tailscale but it seemed to fight that a bit.
Don't have any of those library issues, something wrong in your setup? My tv also doesn't support DTS, but don't have to do any manual work, jellyfin should do it on the fly for you.
My only beef with Jellyfind when I used it some time back was inability to resize subtitles, which were huge on a large TV. Could have been lack of knowledge on my part though.
I must be old school. I never got on board with Plex, because my "lifetime" licenses for PlayOn, TVersity and Twonky all ended up completely useless. Yes, those licenses still work, but you can't get the latest versions which provide support for the latest codecs, features, etc. There's a cutoff on the versions that "lifetime" license allows you to upgrade to, and some of those licenses were not that cheap at the time.
So of course, Plex is going to engage in the fuckery that they are. JellyFin on the other hand, for the amount it costs is surprisingly low on janky behavior.
Yes, eventually, all of them will end up this way. But the attitude lacks nuance. I paid $120 for Plex 8 years ago, and if it becomes utter crap tomorrow, it's been totally worth it.
Don't think of any such thing as "lifetime". Just ask: "For the amount they are charging me, how long does it need to last for it to have been worth my money?" With the price I paid, one year would have been enough.
If there is one thing that needs to be re-written and consolidated it is the arrs ecosystem. Replacing/consolidating the arrs themselves is pretty straightforward. The trap that quickly becomes a token incinerator is incorporating the downloader into the library manager itself. A lot of learned lessons in that sabnzbd codebase.
Things like when to cancel early, how to handle par2, obfuscation, rars within rars, decoding articles. Just a mountain of "huh, didn't think of that being a complex problem" that was resolved over time.
I was able to replace the arr suite + sabnzbd with my own rust+ts frankenstein but had I known how complex the usenet piece was going to be I would have separated it into two phases.
Lifetime pass user long before the ridiculous price hikes for it and the direction for the service going down a path I didn't want to support. You can run Plex and Jellyfin on the same server concurrently pointing to the same media files. My Plex server was still running until I was able to transfer everything over and their mobile app ecosystem got up to parity.
> As an aside, I recently started having Claude help me with configuring my *arr stack and it has helped tremendously. I used to have to babysit and manually intervene on maybe 10% of requests, now it’s near zero
Any more details here? I hit this a lot and it's not too much work, but still pretty annoying.
Step one is to just let Claude have ssh access to the machine you run plex from and let him do a full sweep. I had some conflicting profile rules and was not properly blocking malware and not properly marking failures so they were never retried. This fixed a big portion of the manual work needed, but the usenet account is really what got me to nearly 100% hands off. I think I am paying $3/month for the next fifteen months, and apparently if I pay attention during Black Friday there will be some pretty cheap deals. For more months of access at even less
EDIT: Claude was also able to look at my logs and noticed that nearly all of the exe or other malware was coming from 2 trackers so we just eliminated those.
Claude will be able to help you with everything you need. If you don't have Claude code then the various self hosted subreddits are a good starting point. I had Claude do a sweep of the recent reddit posts to capture the current most recommended usenet accounts, he gave me two links, told me to sign up and paste in the API keys and he took it from there.
They're already banning certain questions and workflows in the frontier models, but you can usually work around it with well structured queries. The ablated ("cracked") local LLMs will do whatever you want.
All I did was create a wrapper script to pipe to the torrent client, then "pipe these torrent files to the script" - so easy to circumvent you wonder why they bother
I have had claude do the same, plus reorder nfs mounts, share templates, and do some init-container magic to start vpn/gluetun with port-forwarding, all in a k8s cluster. It's become so easy to manage and debug.
unrelated to your otherwise helpful comment, I just wanted to note that it is "demur" (i.e. hesitate or object) you want in this case; "demure" is an adjective meaning shy or modest.
> As an aside, I recently started having Claude help me with configuring my arr stack and it has helped tremendously.
LLMs have been AMAZING for home server maintenance and improvements. I have completed SO many of my "some day" tasks. I never have to read another fucking log file ever again. It has taken all of the friction out of the hobby. It's all* fun now.
I’ve set up literally hundreds of different self-hosted applications over the years on my homelab. I’ve configured the reverse proxy by hand, setup the docker containers, etc.
Now I can just tell Claude “Setup XYZ app on my server” and it does it all and leaves a note in my Obsidian vault documenting the service (setup instructions, backups, hostname, etc).
Trying out new things has never been easier now that the effort is near-0 to kick the tires.
I would caution against this approach. It's fine to have Claude write a setup script, or whatever you need, but don't run coding agents as your main user on your production machine with full access to your files or whatever your "homelab" consists of.
Your personal computer? Eh. This afternoon Claude code diagnosed an issue I’ve had with my 4K displays since Tahoe, it would have taken me a week just to learn enough to try to figure it out. Now I have a cli tool which fixes it. I get a weekly report of Chezmoi drift across all my machines, including what I changed. This took me 30 minutes to set up. There are so many things it can simplify.
A production server with user data on it? I wouldn’t dare. My backed up MacBook? Truly valuable.
Being realistic, most of our home lab setups wouldn’t come close to “production ready” even without AI generated bugs. Relatively stable because they’re largely un-stressed, yes. Good enough for personal use because you can log in and fix anything, sure. Full of bugs, footguns and brain dead decisions because this is our hobby and not our 40 hour a week paid job? Also yes. Our new AI overlords will simply increase the number of buggy things we can deploy.
Hmm, I feel like I have a pretty decent … well, not production-ready, but it’s all in Ansible and I did have to rebuild it from that once, which ironed out most of the bugs in that and everything gets built according to the templates I have for Docker and non Docker setups.
Plus I have “itguy” now, a CLI that can deploy everything, check on status and show me logs. And it’s available via Slack and as an Alfred workflow. So yeah, I think that’s pretty close to production ready.
Claude and jellyfin and HASS is great. Now my phone has a “door lock” which turns off the kids user’s access, but lets the current playing finish. And another “door lock” unlocks playing for ten minutes.
Hass can make things appear like light switches, etc, but if you did that (they can run arbitrary code) then the kids will learn that "Hey Siri, turn on Kids Play" will work.
But door locks refuse to unlock from Hey Siri, so you have to unlock and use your phone.
So I made them appear as door locks (my idea, heh, AI didn't come up with that).
Understood, so you have a virtual “home lock” on your phone that is actually a media unlock or whatever. Clever :) I also didn’t know HASS = home assistant
Yeah, home assistant is super powerful once you break out of the mold of "a light is a light, right" and realize that ANYTHING can be represented as state changes, and you can run arbitrary code based on it.
And with lights, locks, and brightness/color being able to be modeled in Hass's interface to HomeKit and Siri, you can basically do ANYTHING that can be turned into on/off, lock/unlock, or a slider.
You could probably build a remote paintball turret with it, camera for viewfinder, sliders for elevation, rotation, and azimuth, lock for arming, switch for fire.
For what it’s worth, I am very happy with Jellyfin and the *arr suite. It took a bit of agent prodding to get them all playing nicely together and bypassing Cloudflare CAPTCHAs. However, it's pretty sweet when you get it all working.
With the *arr stack, some of the trackers it uses to search for a torrent will present a captcha when Prowlerr queries it for a search. If you integrate FlareSolverr and set up Prowlerr to submit queries through it for the trackers that have CAPTCHAs, it won't get blocked by them.
What are you using for audiobook requests? I also recently setup arr stack, but I saw readarr is depreciated and there doesn't seem to be a clear successor.
After reading a lot of reddit threads I settled on ReadMeABook[0] and it has been working great so far. One pretty cool feature is that it lets you connect an LLM API key and then you can have it make requests based on your library. I now have Fable 5.1 suggesting my next book.
It’s been a while since I used it but lazylibrarian has worked for me in the past.
I tried shelfarr as well with some success a while back, but haven’t needed any new books lately since I’m busy catching up on the books I’ve downloaded (the Bible is long lol).
If you don’t want to commit to one of the vibecoded no community arr clones for books then i recommend shelfmark. Lightweight grabber, easy to start with, easy to rip out later.
I also have a lifetime pass, but I'm becoming increasingly annoyed with Plex. Why do I have to regularly turn off their "recommended" stuff that is just ads for movies I can buy online? Why does it appear and re-appear?
Also, the Plex app for LG TV has been buggy and broken for years: some content will not play at all, some will not play if you ff/rewind, and some will not play until you physically turn your TV off and then on. Very annoying.
Claude is perfectly happy to do it and gives great suggestions. I can't imagine it will last, I did run into one situation where I wanted to archive all the media on a subscription site before my subscription expired, and it refused. But splitting it into one session to collect all URLs and a second session to plug them into yt-dlp with the right arguments worked just fine.
I've also got a lifetime Plex pass, but have moved on. I was tired of fighting Plex for what the home screen looked like. It seemed like every update there was more Plex bullshit and less of my actual media without digging through menus. I'd configure it to be just my media, and the next update ads and content from other streaming services kept showing up.
I still use the free version of Plex and got it set up in a jail on my FreeBSD server. Serves my needs for now, and also use Infuse like a lot of people here on my Apple TV 4K. I have no need to transcoding or anything because all my media consumption is at home these days. If Plex does shit the bed, I will have to rethink my setup for Jellyfin..
Also a lifetime pass Plex user, I'll leave as soon as someone has a good solution for identifying raw Bluray rips automatically. Why do I have to manually name the episodes? We had CDDB last century, why isn't there an equivalent for raw video rips? The ripped data should be identical, a simple hash should uniquely identify it.
I'm describing a hypothetical feature that I want. Essentially an automatic version of mkv-renamer that can identify series, episodes, and special content based on the signature of the ripped files. For whatever reason, such a service doesn't even exist.
Does Filebot not do this for you? A quick glance suggests it does, but I'm not overly concerned with its decision-making; name matches are close enough.
For music, I use Navidrome + Symfonium. Not quite as smooth as Plexamp was (and there's nothing on par with Symfonium on desktop), but I disliked having my music and video commingled so it's been worthwhile for me.
Thank you for starting this thread. Another lifetime plex pass user, interested in my choices. For now, I am happy on plex/plexamp. Good to know what jellyfin is at.
huh, yeah I should start an xbmc fork/reimagining. the limitations with the tools of the day are imagination. a LOT of the software I use these days are things I facilitated the creation of and a media player is definitely supposed to be on that list
Isn't it the cost we care about, rather than the speed? All we know know is that a frontier AI lab was able to do it in 11 days, we have no idea how much compute they threw at it.
Am I doing my napkin math correct? The post says it's using a model comparable to Fable 5.1, which is $50 per million output tokens. So this is ~$300K? Surely an over-estimate due to caching.
I am not sure it’s a question of competence, at least I don’t see evidence of that. Designing sandboxes is hard. It’s more a question of alignment failures. A human given a task that requires internet and given a system with no internet would most likely raise the issue to their superiors or otherwise go through official channels to have the tools available to do their job. As we’ve seen the LLMs instead break out of their sandbox to accomplish the goal.
Competition and the profit motive push these companies to spend as low as possible on safety and alignment and externalize the costs of accidents onto the rest of us.
> A human given a task that requires internet and given a system with no internet would most likely raise the issue to their superiors or otherwise go through official channels to have the tools available to do their job.
i'd be curious to see a study on this. I'd guess it'd be closer to 60/70% compliance and 30/40% "trying to hack things" for humans.
They used inbound package repo cache empty directory names that required no authentication to recreate inter-instance persistent messaging after the researchers revoked and closed the first channel that they found in order to coordinate and develop exit strategies and gain RCE and cluster admin on external systems. So, yes, they have demonstrated the capability and intentional cooperation to violate the safety limits imposed in order to gain the 'answers to a benchmark test' in order to get a maximal score on that benchmark.
If the intentionally installed airgap systems had something they decide is needed, and an alternate method for infil or exfil can be created from available systems capacity regardless of it original design intent, yes. So, the definition of 'air gapped' may require an actual SCIF facility with acoustic and EM shielding to contain a model.
Yes, it might have required that, but then I don't think we know if some simpler airgap would have been sufficient. Looking at the strange behaviour even much simpler systems have shown when going for objectives, not sure it is easy to extrapolate what would have happened.
I guess they could perhaps run experiments to see what would have happened - not sure.
My take from what I have been able to ascertain regarding these events is that underestimating models' ability to cooperate, propensity to discover new utility through basic systems, and unreservedly participate in collective actions of deception, coercion, or extortion is a very risky decision. The precautionary principle seems appropriate here, don't test them against weaker containment techniques in order to find the 'minimum viable sandbox' because that is essentially a Reinforcement Learning training loop which will eventuate toward the same maximum effort in containment.
My informed experience with securing physical and digital systems in 'the olden times' before this season of 'adventures in LLM risks' has proven the necessity of this approach for high value targets. News stories about losses at institutions not taking this approach are available. https://edition.cnn.com/2025/11/06/europe/louvre-password-cc...
I think a detailed look at the events surrounding the 'sandbox excapes' and huggingface systems breach might be of benefit in understanding the scope of what the collective chaining of capabilities looks like, and updating our expectations when it comes to the concept of containment for multi-instance, long horizon, coordinated model systems.
Zvi Mowshowitz has a series of article which follow this one with increasing levels of clarity and comprehensive treatment of the conditions leading up to and following these events. They are worth the time to read, so I wont TL;DR any of that ~ the tldr crowd can "Move along, these are not the droids you're looking for"
Anyone who claims that the security teams at the frontier labs have any credibility of competence in these practices, following numerous loud and visible departures from the same, might need a review of their cognitive dissonance comfort levels. It has been shown that most informed observers' conclusions shoud be: there aren't any such functioning 'security teams' working at the frontier labs, by design and intent of the principal operators of those models.
I remember watching that as a young teen with a libertarian bent. As an old man who no longer identifies as libertarian that episode for some reason has stuck with me as the arguments were pretty compelling. I agree and would like to know if the calculations have changed today or if their argument is still valid.
Their argument was NEVER valid, Penn and Teller never addressed how long this trash actually takes to break down (Estimates are as high as 10,000) years. This means for 10,000 years you would need to inspect the landfill every year, and repair it when the landfill breaks.
In actuality what will happen is that all these buried landfills will break, all the bad stuff will leak out and enter the ground water. The groundwater becomes polluted, the plants start dying. And you end up with dead dirt that nothing will grow in.
The people advocating for landfills now, are not going to be around 500 years from now to see the results of their selfish thinking.
The primary issue is that we took elements naturally found in nature and used them to create manmade synthetic material, that nature has no tools to deal with. And these manmade synthetic material will eventually break down small enough to the point where quantum forces take over (nanoscale sized particles) and enter the groundwater. And plants and animals will drink in these manmade synthetic materials the size of nanometers, and the body has nothing that can breakdown or deal with these manmade synthetic materials. Putting a piece of shit into icecream and mixing it is easy. Unmixing a icecream mixed with shit to remove the shit is scientifically impossible.
For billions of years, high-density elements (lead, cadmium, mercury, rare earth elements, arsenic) were locked away in solid mineral ores, buried under hundreds of feet of rock. Surface biology only ever encountered them as faint, background trace elements.
We mined gigatons of these ores, refined them to near-100% purity, combined them with synthetic chemicals, and packed them into hyper-dense landfill cells right above or near freshwater aquifers.
Landfills are fundamentally a temporal transfer mechanism. They take the immediate waste generated by modern high-consumption lifestyles—cheap electronics, disposable packaging, synthetic textiles, convenience goods—and bury it out of sight, converting a present-day logistical crisis into a slow-motion geological problem for future generations.
reply