cyberwolfie @ cyberwolfie @lemmy.ml

Posts

35

Comments

422

Joined

2 yr. ago

Hm, the Music Assistant at least does not quite accomplish what I hope and it seems to rely on services such as Spotify or YT Music to be integrated. I couldn't quite evaluate the LinkPlay-solution, but his comment on SD card corruption with RPis made me a bit worried for the balenaSound approach. I guess there's a lot of write operations in such a setup, that can easily corrupt the SD cards. I wonder how often they kept failing for him - maybe it'll end up being some sort of a "subscription fee" 😅

My wife says no more toys at the moment, but if I were to implement this, I’d probably pick up one of those Up2Streams for each room and try out the LinkPlay integration.

Then you have something to put on your list for Christmas, if that is something you celebrate :)

Oh, I think this is very difficult. First of all because it is not a single reason why a future where privacy has eroded is a very bad thing, but rather many different reasons. This makes it difficult to know where to start, as it will depend on the person you are talking to what they are more receptive to. Concepts such as security, privacy, secrecy and anonymity are often confused. You have different actors you would want protection from, including corporate and governmental entities.

I don't think most ordinary people you meet will be bad faith actors though, but I do think many tend to take offense if you are outspoken against something that is proclaimed to be about protecting children. Why wouldn't you want to save children?

Some of the reasons below, but not an exhaustive list. As I said, difficult to know where to start.

1. You do have something to hide, even though you might not be doing anything illegal (to your knowledge). Most people dislike people staring into their living room from the street, and will install curtains or other ways to prevent it. Most people closes and locks the door when they go to the toilet. Most people do not say every single thought they have out loud. I think the disconnect comes from people not actually knowing what data is collected, and even if they do, they do not understand how this data can be used / misused to learn things about you or manipulate you, and the privacy threat of having this data stored anywhere even if it is not being used (i.e. risk of data leaks). In terms of manipulation, I think that the story on how Facebook nudged people to vote in the Scottish referendum highlights the creepy influence such a company can have on society, and this was already in 2014. Who's to say the owners of such platforms will not use it to sway elections their preferred way by using such nudging tactics on the population they want to vote, and not on the ones they'd rather stay home. We shouldn't have to trust that they don't abuse such a power.
2. What today might be perfectly legal, might not be legal tomorrow. Case in point are the draconian abortion laws implemented in various states of the US. Facebook had to comply with government requests to hand over chat logs..
3. What today is illegal, should perhaps not be illegal. We do not want 100% law enforcement, as that would mean that we consider today's laws final, however we are constantly evolving our laws. A recent example is legalization of weed in the US. How many have been incarcerated and had their lives ruined on charges related to weed? Yes now the same activities are in many states considered legal. Or homosexuality? Sodomy laws are not a very distant past in many countries (and still exist in other places of the world). If you had Apple or Google scanning your phones and flagging you to law enforcement for illegal activities. Effective mass governmental surveillance (and corporate surveillance that can be passed on to law enforcement) could potentially send countless people to jail on charges that could be legalized in just a few years.
4. Building an infrastructure for mass surveillance is not future-proof. You might trust your government not to misuse it today, but what about after next election? There are countless examples of less-than-democratic forces gaining power in Western democracies in recent years. We need strong protections against potential oppression/suppression, and not just soft protections that are easily swept aside.
5. We are dependent on journalists and whistleblowers exposing wrongdoing in our society. Lack of tools that ensure privacy and anonymity prevents this.
6. Even if our societies are not oppressive regimes today, many around the world are. Political opponents and resistance groups in such regimes need ways to protect themselves. Otherwise authoritarianism will have an too easy time to crack down on dissidents, making organized opposition impossible.

As long as you are connected, I think network adb will stay active but if you leave your network f.e., you have to re-enable it in the developer options. But don’t take my word for it. Feels like google changed this behaviour every major release with android.

Oh, if that is the case, this will not work. I'll test it out later this week to see.

On their blog they say: This project is made possible by the awesome work of various open source projects, including Shairport Sync for Airplay, Raspotify for Spotify Connect and Snapcast for multi-room audio sync. So they “just” glue existing stuff together which leaves you with roughly the same limitations as if you would do it yourself. It might allow spotify to be used with snapcast for multiroom but as I’m a yt music user I didn’t digg any deeper.

Hmm, maybe I misunderstood it. Here is a blog post that shows how it can also use Bluetooth. To me, it sounds like that makes an app agnostic solution as long as you are fine with using Bluetooth. My understanding is that you then just connect to one of your speakers with Bluetooth when you want to cast, and you can then control which speakers the audio should play from. I will research this more. I should order a 3.5mm jack extension to the Pi Zero W, which is the only part I miss to be able to set up a proof-of-concept at home.

Nice. I don't have much experience with VMs yet. Is USB pass-through easily configured?

The Acrylic device also seems pretty nice - I'll dig a little deeper on that one as well. The rabbit hole becomes ever deeper.... :)

It would be nice to have other people being able to use it, but it is not a top priority for me. Also battery drain is not that much of an issue as it will only be used while at home. The last point is a bit more concerning though, so I will see if I can test this out and see how well it works. And yes, the connection seems to be the biggest issue here. But it seems that once configured, it only requires running 'scrcpy' on the recieving end. And KDE Connect can be setup to run commands remotely, for example I just set it up to open VS Codium in a specific folder from my phone. I can't seem to add that command as a quick access tile (which would be my preferred option), but I could add it as a widget on my home screen for quick access so that conncetion is a button press away. How long the connection can stay for I don't know, but I will see if I can't test that out this week.

Did you check out balenaSound by the way? If so, what difficulties did you run into that made you discard it?

fcast looks nice, but if I understand it correctly it would require implementation in every specific application. I think if I were to jump onto the Grayjay-wagon, that could be nice, but I would love for my solution to be app agnostic.

EU is doing a lot of good work to protect the privacy of citizens against corporate surveillance, but continues to propose regulation that would increase government surveillance. News such as this is good, as it seems to show that there are protection measures within the EU to stop such legislation from being effectuated. Another example is the Data Retention Directive, which was first passed back in 2006, but then later declared invalid by the European Court of Justice in 2014. However, while the intent when it comes to corporate surveillance seems aligned with the public interest, the intent when it comes to government surveillance is not. Such privacy violating proposals will continue to be proposed.

I certainly do not have a good overview over all of this. We are completely beholden to the great work of pro-privacy organizations and corporations to keep exerting pressure and making these pieces of legislation known and understandable to the public. But unfortunately, most people can't even begin to consider the implications of such overreach, which is why the "protect the children"-rhetoric is so effective - "I am not doing anything illegal and thus have nothing to hide, so if we can protect the children from abuse by removing encryption which is only something criminals use anyway, I'm fine with that". I am clueless to how I can best contribute here, but I am luckily seeing a shift among friends and family in the awareness on these topics.

Ok, thanks for clarifying that - as I feared. I elaborated on my progress into finding a solution to this below, in case you have any interest in it or any thoughts that could help me along.

Hehe, and for that I apologize! In case you are interested, here are the options I am so far considering. Still very much in the research phase, trying to figure out my specs before I buy any gear I don't already have. I am not very experienced with this, so feel free to point out any baloney in the below text. Anything that could spare me time researching dead ends is gratefully received.

Option 1: Snapcast

Since I will be running a Raspberry Pi 4 with Home Assistant anyway (not yet set up properly), I would like to make use of the Snapcast integration in HA to run the Snapcast server and then set up a Raspberry Pi Zero W (with some 3.5mm extension) with all speakers I want to connect. Ideally these Zeros could be powered by the speaker themselves (through USB for example) to avoid two plugs, but I don't know how realistic it is to achieve this, and I have not done much research into this yet.

The issue that I so far have is that I don't know how I can stream audio from my Android device to the Snapcast server. From my understanding, and what I hoped to clarify with this thread, is that it requires a specific audio source that the Snapcast can recieve audio from. Here is a list. This seems to my limited experience much more doable from my laptop running Linux than from my Android device. But I don't know...

Currently I will be investigating whether I can use audio streaming in scrcpy to stream audio to the Pi, and then route that via PulseAudio to Snapcast. I don't know yet if this is a really cool idea or an incredibly stupid idea. I want to setup scrcpy for another purpose anyway, so why not try? :) It might introduce additional latency from my device to the speakers, but as this happens before the "distribution" from the server to the clients, I don't think this would affect the synchronization. Also, I will never use this to speak on the phone with anyone, so that there is some latency doesn't really matter to me. The biggest issue would be toggling this off and on - maybe via a remote command with KDE Connect or something like that. If I could set that up as a custom tile in the quick access menu in Android, that might work.

All in all it seems a bit too convoluted though, so I don't have too much faith in this.

Option 2: balenaSound

So this is the solution I first learned about as an alternative to Sonos, but I was turned off by the need to connect my devices via the internet to the balenaCloud hosted by the developers. However, either I missed this in the first round of research, or they have released it since, but OpenBalena exists which has much of the functionality of balenaCloud (but not all) and can be self-hosted. If I could get the server to run on my Raspberry Pi 4, and then flash balenaOS onto each Pi Zero W, this could provide what I want as I understand that it allows to stream audio directly to the Zero Ws via Bluetooth (with subsequent sync to the OpenBalena server via WiFi). It would be a much simpler solution than the one above, especially in terms of toggling it on and off on my Android device.

A downside to this solution though, is that I believe I would not be able to install it on a Pi running Home Assistant OS (correct me if I am wrong), and that running HA through Docker makes installing new integrations a bit more cumbersome? Maybe that will pose no problem, as I don't plan on using too many integrations anyway (Zigbee, Netatmo and a MQTT broker). I could perhaps also run a VM that runs Home Assistant OS?

I'm on a Fairphone 4 with CalyxOS

Hm, I don't see any IPv6 addresses anywhere, even on https://ipv6.ipleak.net/

Yeah, that is true - I didn't formulate that quite well. This also happens to me when the kill switch is enabled, as it is disabled when I disconnect. However, it would not disable itself if I simply lost connection, which is in the case where I most want the leak protection. In the other cases, I could make sure to always quit qBittorrent before disconnecting from the VPN. Not ideal, as it would be easy to make a mistake.

tun0 goes away when I am disconnected, so I don't think anything is keeping it open. I noticed that connections in Proton VPN were set to UDP, so I changed this to TCP and this seems to have worked. Will still want to do more testing to be sure.

I don't seem to have v3 installed, by the way. The old GUI tool seems to have been uninstalled when upgrading, and the CLI tool I unisntalled myself.

How could I check this?

That is unfortunately not available in the Linux client.

I'm using the Torrent Address Detection tool, which I assume is checking the IP broadcast by my torrent client.

That was the case until the recent update to the Linux GUI client. I had to change the network interface bind in qBittorrent because the proton0-one stopped appearing. I assumed they had just made some changes, but could this mean that something is faulty with my install?

It is the "Torrent Address detection" magnet link I am using, and it is this that reveals my real IP when the VPN is disabled. The traffic in qBittorrent stops though.

EDIT: And as I mentioned in an earlier post, it works as intended if I open the client when not connected to the VPN. It is just if the client is running while I disconnect that this problem occurs, as far as I can tell.

This is so nice!

If I start the client without being connected through the VPN, my IP does not show up. It is only when the VPN connection is disconnected and I transition from one network interface to another. This is also if I have the kill switch enabled, so I imagine that if the connection is lost, I am safe as all internet access would be blocked then. And that it is only if I manually disconnect myself that this happens.

Could this be a Proton VPN issue? I am pretty sure I checked this previously, and didn't see it before the recent Linux client GUI update, but I am not entirely sure.

I've tried staring with all my might at the different options in the hopes I might identify something. I did turn off the features suggested above in the comment field, but have still been unable to solve it.