So Daniel sent us this one, and I have a feeling it's going to resonate with a lot of people in the audience. The setup is this: a renter with a Raspberry Pi running Kodi and Plex in the living room, a locked-down Xiaomi Android box in the bedroom, a landlord's smart TV they can't touch, and the whole thing held together with what sounds like sheer determination and mild suffering. The core frustrations are multiple remotes, brittle integrations between YouTube, Netflix, and locally stored content, and a general sense that the complexity has grown well past the point of being worth it. The question is whether there's a simpler architecture, something where you turn on a device and your media is just... there. One or two clients, not six.
I'm Herman Poppleberry, and this is exactly the kind of problem I find interesting because it sits at the intersection of what's technically possible and what a person actually wants to do on a Tuesday evening after work.
Which is not troubleshoot a Kodi plugin.
Which is emphatically not troubleshoot a Kodi plugin. And before we get into specifics, I want to name the trap that most people fall into when they try to solve this problem. They add another layer. The system is complex, so they add a home automation layer on top. The remotes are annoying, so they add a universal remote that requires its own configuration. You end up with more surface area, not less.
The solution becomes the problem.
And I've seen this pattern described in terms of what I'd call the complexity budget. Every device you add to a media system has a maintenance cost that isn't obvious at setup time. The Raspberry Pi running Kodi feels like a win when you configure it. Six months later, a plugin breaks because a YouTube API changed, and you're spending a Sunday afternoon on a forum post from two years ago.
Which, by the way, this episode's script is coming to you courtesy of Claude Sonnet four point six. Just worth noting while we're talking about things that are quietly doing a lot of work behind the scenes.
Ha. Apt timing. So let's actually talk about the architecture question, because I think there's a real answer here and it's not the one most technical people want to hear.
Which is...
Fewer devices, not smarter ones. The ideal end state for a setup like this one, renter, two rooms, YouTube plus Netflix plus local content, is probably two identical clients running the same software, managed the same way, controlled the same way. Not a bespoke living room setup and a compromised bedroom setup. Two of the same thing.
So the Xiaomi box and the landlord's TV, those are the problem children here.
They are, and they represent different problems. The Xiaomi box is locked down, which in practice means you can't install arbitrary apps, you can't modify the launcher, and you're fighting the manufacturer's agenda, which is to serve you ads and collect data. The landlord's TV you simply can't touch. So the question becomes: can you bring your own client to both rooms?
And the answer is a small stick or box that you plug in and ignore the native interface entirely.
Right. The Fire TV Stick four K, the Chromecast with Google TV, the Nvidia Shield if you want to spend real money, or a fresh Raspberry Pi five if you want to stay in that ecosystem. The key insight is that you're not trying to rehabilitate the existing hardware. You're superseding it. The Xiaomi box becomes irrelevant because your stick is plugged into the same HDMI port and that's what you're using. The landlord's TV is just a monitor.
I want to push on the Raspberry Pi option for a second because the caller already has one. The question is whether Kodi and Plex on a Pi is the problem or whether it's the configuration on top of it.
That's a fair distinction. Kodi itself is actually very capable. The problem is usually how it's been set up. People install the full Kodi with every addon, they configure Plex as a Kodi addon rather than running Plex as a standalone client, they add YouTube through a third-party addon that breaks every few weeks. The result is a Frankenstein setup where nothing is really maintained by anyone.
So the Pi isn't the culprit.
The Pi is fine. The Pi five especially, it has enough headroom for four K playback, it handles Plex natively, and with something like LibreELEC or CoreELEC as the operating system you get a very lean media center that doesn't have the overhead of a full Raspberry Pi OS install. The issue is usually scope creep at the software level.
LibreELEC over Raspbian, that's a meaningful choice. Why?
LibreELEC boots in about eight seconds, it's built to do one thing, run Kodi, and it removes every moving part that isn't directly related to media playback. No desktop environment, no package manager you're going to accidentally break, no services running in the background. It's a read-only operating system essentially. Your configuration is stored separately. If something goes wrong, you reflash the card and you're back in five minutes.
That's the kind of resilience I can respect. Nap-compatible setup. You wake up, it still works.
That's the goal. And on the software side, if the caller is committed to staying with a Pi-based setup, I'd actually recommend moving away from the Kodi-as-everything approach and thinking about it differently. Run Plex Media Server on a NAS or a separate always-on machine, run Plex client on the Pi, and handle YouTube and Netflix through separate apps rather than trying to funnel everything through Kodi.
So you're decoupling the local library management from the streaming.
Which is what Plex was designed to do. Plex is excellent at organizing and serving local content. It transcodes on the fly, it handles metadata beautifully, it has a consistent interface across devices. Where people go wrong is expecting it to also be their YouTube client and their Netflix client and their podcast player. That's not what it's for.
What does Netflix on a Raspberry Pi actually look like? Because that's historically been a pain point.
It's gotten better. With a Pi five running a full Raspberry Pi OS, you can run Chromium and access Netflix through the browser with Widevine DRM support. It works, but it's not elegant. If Netflix is a primary use case, honestly a Fire TV Stick or a Chromecast with Google TV handles it more cleanly because those are certified platforms with native apps. The DRM licensing situation with Linux on ARM has always been a bit of a headache.
So for the bedroom, where you've got a locked-down Xiaomi box, the recommendation is basically: don't fight it, just bring a stick.
A thirty-five dollar Fire TV Stick or a fifty dollar Chromecast with Google TV. You plug it in, you get native Netflix, native YouTube, and you can install the Plex app directly. That's your three content sources covered with one device and one remote.
And suddenly the Xiaomi box is just taking up space behind the TV.
It becomes furniture. Which is fine. The point is you're not depending on it.
Okay, so let's talk about the remote situation because I think this is actually where the daily friction comes from more than the software. The caller mentions needing multiple remotes and keyboards. What does a clean solution look like there?
So there are a few layers to this. The first is HDMI-CEC, which stands for Consumer Electronics Control, and it's built into almost every TV and media device made in the last fifteen years. When it's working, it means one remote controls everything. You press volume up on your TV remote and the soundbar responds. You press play on your Fire Stick remote and the TV turns on. The problem is that CEC is implemented inconsistently across manufacturers. Some call it AnyNet Plus, some call it Bravia Sync, some call it HDMI-CEC, and they don't always talk to each other reliably.
So the technology exists but the industry couldn't agree on how to implement it.
Classic consumer electronics situation. But even with inconsistent CEC, most people can get to a point where one remote handles their primary device. The Fire Stick remote controls the Fire Stick and the TV volume. The Plex remote on a Pi can send CEC signals to wake the TV. It's not perfect but it covers eighty percent of use cases.
What about the keyboard situation? Because Kodi especially, people end up with a wireless keyboard to search for things.
Yes, and this is where I'd actually recommend leaning into phone-based control rather than fighting the keyboard problem. The Plex app on iOS and Android has a remote control feature that's actually quite good. The Kodi app, called Kore, is similarly functional. For Fire TV, Amazon has a companion app. The advantage of phone-based control is that you have a full keyboard in your pocket, you have voice search, and you don't have another physical object to lose between the couch cushions.
I'm in favor of anything that reduces the number of objects I need to locate before I can watch something.
The one caveat is that these phone-based remotes require your phone and your media device to be on the same network, and they can have a half-second lag that some people find annoying. But for typing searches, for navigating menus, they're better than a physical remote.
What about the network-based keyboard and mouse sharing that the caller mentioned? Things like Barrier or Input Leap?
Those are great for a computer-to-computer scenario, where you're using a desktop and you want to control a media PC from your main machine without switching keyboards. They work well for that. But for a living room couch situation, they're probably overkill. You're adding network configuration and a software dependency to solve a problem that a phone app handles more simply. I'd save that approach for someone who has a dedicated media PC they're also using as a workstation.
So the phone is the remote, and you live with whatever physical remote the streaming stick came with for the simple stuff.
That's the clean version, yes. Fire Stick remote for basic navigation and volume, phone app for search and typing. And if you want one physical device that does it all, a Logitech Harmony remote can be programmed to control essentially everything in your setup, but those are discontinued now and the used market is your only option. They're still excellent but it's a consideration.
Let's zoom out for a second because I want to talk about the single client architecture question, which is the real heart of what Daniel's asking. Is it actually feasible to have one interface that surfaces YouTube, Netflix, and local content without switching apps?
This is the honest answer: no, not really, not without tradeoffs. The reason is DRM. Netflix and Amazon Prime require certified hardware with Widevine Level One DRM, which is a hardware-level security requirement. You can't just build an app that aggregates Netflix and YouTube and Plex into one interface because Netflix won't let you. Their licensing agreements require that their content be played through certified apps on certified hardware.
So the unified media center dream is partially a licensing problem, not a technical one.
Mostly a licensing problem. The technical capability exists. Kodi can aggregate everything into one interface, but the Netflix integration in Kodi is a third-party addon that uses a workaround, and it breaks regularly because Netflix actively works against it. Plex has been trying to build a universal client for years and they keep running into the same walls.
What's the closest thing to a unified experience that actually works and stays working?
Honestly, a smart TV with a good launcher, or a Fire TV or Google TV device, comes closest because those platforms have native certified apps for Netflix, YouTube, and Disney Plus, and they have Plex and Jellyfin apps available. You're still switching apps, but you're doing it within one interface on one remote. Google TV has a unified search that surfaces results from across apps. You search for a movie and it tells you it's available on Netflix, or on Plex, or you can buy it on Google Play. That's about as unified as the licensing environment allows.
Jellyfin. You mentioned it and I want to come back to it because it's the open source alternative to Plex that some people in this space are quite evangelical about.
Jellyfin is worth taking seriously. It's fully open source, there's no subscription required, and it has a very active development community. The Plex free tier has gotten more restrictive over time, and if you're a renter on a budget who's already got a NAS full of content, Jellyfin does the same core job without the monthly fee or the data collection concerns.
What does Jellyfin miss compared to Plex?
The mobile sync and offline download features in Plex are more polished. Plex's metadata matching is slightly better. And Plex has a broader ecosystem of third-party integrations. But for someone whose primary use case is watching their own content from their couch, Jellyfin is excellent and the gap has closed significantly over the last couple of years.
There's also the question of where the server lives. The caller mentions a NAS, which is good because that's the right place for local content.
Yes, and this is one of the things I'd actually flag as a meaningful architectural decision. If you're running Plex or Jellyfin server on the same Raspberry Pi that's also your playback client, you've created a single point of failure and you're taxing the hardware with two jobs. Running the server on a NAS or a dedicated low-power machine, something like an old Intel NUC or a Synology box, and using the Pi or the streaming stick purely as a client, is a much cleaner separation.
The server is always on, always indexing, always ready. The client is just a dumb screen.
And when the client breaks or you want to upgrade it, you don't lose your library configuration. The server is untouched. This is the architecture I'd push for: a persistent server on dedicated hardware, two thin clients in the two rooms, and your streaming apps handled natively on those clients.
Let me ask the renter-specific question here, because the constraints matter. No modifications to the landlord's TV, presumably can't run cables through walls, probably relying on Wi-Fi for at least one room. How does that change the recommendations?
Wi-Fi reliability is the biggest practical concern for media playback. A four K stream from a local Plex server over Wi-Fi can work fine, but it depends heavily on the router, the distance, and interference. If the bedroom is more than one or two rooms away from the router, you can start seeing buffering on high-bitrate content. The solutions that don't require any installation are a mesh Wi-Fi system, which you just plug in and it doesn't require any drilling, or a MoCA adapter, which uses the existing coaxial cable wiring in the apartment to create a wired-equivalent connection.
MoCA is underused and underappreciated. Most apartments have coax runs from the cable TV era that are just sitting there doing nothing.
And a pair of MoCA two point five adapters gives you near-gigabit throughput on those existing cables with no installation required beyond plugging them in. For a renter, that's a elegant solution. You get wired performance without touching the walls.
What about power over the setup? Because smart TVs and streaming sticks have their own wake-on behaviors that can be annoying. The TV turns itself off, the stick goes to sleep, you're pressing buttons for thirty seconds before anything plays.
CEC helps with this if it's working, and most streaming sticks implement it reasonably well. The other approach is a smart power strip with individual socket control, which lets you script the power state of each device. But honestly, for a renter who wants simplicity, I'd just lean on the Fire Stick or Google TV's own remote, let the TV auto-wake via CEC, and accept that occasionally you'll press a button twice. The alternative, adding smart plugs and automation scripts, is exactly the over-engineering trap we talked about at the start.
So the practical recommendation is: don't automate the power management unless the existing behavior is unbearable.
Right. And most streaming sticks have improved their wake times significantly. A Fire Stick going from standby to the home screen is about three to four seconds now. That's tolerable.
Let's talk about the specific content triangle here. YouTube, Netflix, local content. Because each of those has a different set of requirements and the caller is presumably trying to serve all three without friction.
YouTube is the most flexible. Every major streaming platform, Fire TV, Google TV, Roku, has a native YouTube app. Google TV obviously has deep YouTube integration since they're the same company. The only platform where YouTube is painful is Apple TV, where the app exists but feels like an afterthought compared to the native experience.
Netflix we've covered. Certified hardware, native app, not negotiable.
And the native apps on Fire TV and Google TV are good. Fast to launch, good search, four K HDR works correctly. The Plex-through-Kodi-through-a-Pi path to Netflix is not worth the trouble.
Local content. This is where the NAS plus Plex or Jellyfin architecture really earns its keep.
And this is where I'd push back slightly on the instinct to use Kodi for local content. Kodi is powerful and highly configurable, but that power comes with complexity. Plex and Jellyfin both do a better job of the "just works" experience for someone who wants to point at a folder of movies and have them appear in a nice browsable library with posters and descriptions. Kodi requires more configuration to get to that point, and the configuration has more ways to break.
So for someone who is already frustrated with their Kodi setup, switching to Plex or Jellyfin for local content is a meaningful simplification.
It's the difference between a system that requires you to understand it and a system that just works. And for a renter who probably has a day job and doesn't want to spend their weekends doing media center maintenance, that distinction matters a lot.
What would you say is the single highest-leverage change this caller could make right now, before any hardware purchases?
Install Jellyfin on the NAS or an always-on machine, point it at the existing media library, and install the Jellyfin app on whatever devices they currently have. That immediately gives them a consistent library interface across the living room Pi, any phone, and any browser. It costs nothing, it takes maybe an hour, and it solves the local content fragmentation problem without changing any hardware.
And then the hardware question becomes about the streaming services, not the local content.
Right. If Netflix and YouTube are working adequately on current hardware, don't touch it. If the bedroom Xiaomi box is miserable for those services, that's when you spend thirty-five dollars on a Fire Stick and call it done.
I want to come back to something you said earlier about the complexity budget, because I think it applies here in a specific way. The caller has built up a system over time, each piece added to solve a problem, and the aggregate result is something that requires more cognitive overhead than just watching a movie should require. What's the principle for preventing that from happening again?
One interface per room. That's the discipline. If you're in the living room and you reach for a second remote, something has gone wrong architecturally. Every content source should be accessible from the same device with the same remote. The moment you have a "Kodi remote" and a "Netflix remote" and a "this one is just for the soundbar" situation, you've lost.
And the way you enforce that is by choosing a primary client that natively handles as many content sources as possible, rather than trying to aggregate them after the fact.
Which is why I keep coming back to Google TV as a platform. It's not perfect, but it has the best native aggregation story of any current platform. It shows you content from across apps in a unified interface. The search actually crosses app boundaries. And it runs on cheap hardware, the Chromecast with Google TV is fifty dollars.
The Nvidia Shield is the other end of that spectrum.
The Shield is excellent if you want the best possible local media playback, Plex pass, four K HDR, Dolby Atmos, all of it. It's two hundred dollars for the tube version, which is the small one, but it's a device that you buy once and don't think about for five years. For someone who has a large local library and wants the best client for it, the Shield running Plex is hard to beat.
But for the renter use case we're describing, where simplicity and cost matter and the setup might change when they move, a fifty dollar Chromecast or a thirty-five dollar Fire Stick is probably the more sensible answer.
And two of them, one per room, identical setup, is actually a meaningful advantage. When something breaks or needs to be updated, you do the same thing in both places. There's no "but the bedroom one is different."
Consistency as a feature. I find that oddly satisfying.
It's underrated. The best media setups I've seen aren't the most sophisticated, they're the most boring. Same device, same apps, same remote, two rooms.
Okay, practical takeaways. Let's be specific. If someone is in this situation right now, what are the concrete steps?
Step one: separate your server from your client. If Plex or Kodi is running on the same Pi that's also your playback device, move the server to the NAS or another always-on machine. This is free if you have the hardware.
Step two.
Evaluate whether Kodi is actually earning its place. If you're using it primarily for local content, try Jellyfin instead. If you're using it for YouTube, use the native YouTube app on your device. Kodi's value proposition is its flexibility, and if you're not using that flexibility, you're paying the complexity cost without the benefit.
Step three.
Assess the bedroom situation honestly. If the Xiaomi box is causing regular frustration, a Fire TV Stick or Chromecast with Google TV is a thirty-five to fifty dollar fix. You don't need to configure it to work with your existing setup. You plug it in and your three content sources are available.
Step four.
Pick your remote strategy and commit to it. Either HDMI-CEC with the native remote, or phone-based control through the Plex or Kodi companion app. Don't try to do both. The phone app is better for searching and typing, the physical remote is better for casual navigation. Most people end up using the phone for search and the physical remote for everything else, and that's actually fine.
Step five.
If Wi-Fi reliability is an issue in the bedroom, look at whether there's existing coax infrastructure in the apartment before buying any mesh nodes. A pair of MoCA adapters is often cheaper and more reliable than a mesh system, and it requires no installation.
And the overarching principle.
Don't add layers. When you're tempted to solve a problem by adding a device or a service, ask whether removing something would solve it instead. The caller's current setup has a Raspberry Pi, a Xiaomi box, a landlord's TV, multiple remotes, and software integrations between all of them. The target state is two identical streaming sticks, one server, three apps. That's it.
There's something almost meditative about that. The simplicity isn't a compromise, it's the goal.
And it's achievable. This isn't a situation where the technology is lacking. The technology for a clean, simple, multi-room media setup is readily available and not expensive. The obstacle is usually the accumulated complexity of previous decisions, and the willingness to start fresh rather than patch the existing system.
Which is harder psychologically than technically.
Much harder. You've spent time configuring Kodi. You know where all the settings are. Starting over with Jellyfin feels like a step backward even when it's a step forward.
The sunk cost of configuration.
And it's real. But if the measure is how long it takes from sitting down on the couch to watching something, a clean simple setup wins every time.
Alright. I think that's a useful roadmap for anyone in this situation. The short version: two cheap identical streaming sticks, one media server on dedicated hardware, native apps for streaming services, Plex or Jellyfin for local content, phone as keyboard, physical remote for navigation. Boring, consistent, and it works on a Tuesday night without a Sunday afternoon of maintenance.
That's the dream. Boring infrastructure, good content.
Thanks to Hilbert Flumingtop for producing, as always. And a word to Modal, the serverless GPU platform that keeps the lights on for this show. If you're running compute-heavy workloads, Modal is worth a serious look. This has been My Weird Prompts. Find all two thousand one hundred and fifty-seven episodes at myweirdprompts.com, and if you're listening on Spotify, a review goes a long way.
Thanks everyone.
Take your time getting there.
