I've been meaning to move to Home Assistant for years, but most examples are like yours and I'm not sure if my seemingly simple use case is covered. Would you mind providing some additional insight?

My use case is to get off of my Alexa devices. The main feature I want is voice control of the lights. As far as which smart bulbs are supported, I'm comfortable dealing with that (I know half of mine are unsupported by Home Assistant and will need replaced). So my biggest question is, what's the easiest way to add voice support for basic control of lighting groups (ex. "Turn off the living room lights", "Turn on the bedroom", "Set the kitchen lights to 50%", etc..)? Might be nice if I could convert my existing Alexa devices into remote speaker+mics for Home Assistant, but I'm hoping there are better solutions (and hope they're affordable).

Most of the features I use are very basic Alexa features, but my normal use of them is via voice, so that has to be integrated:
* Time (What time is it? What is the today's date?)
* Timers (Set a timer for 3 minutes. Remind me in 40 minutes to get the laundry.)
* Weather (What's the weather? What's the weather tomorrow? What's the (current) temperature?)
* Music (Play White Stripes on Pandora. Pause. Resume. Set volume to 5. Raise volume. Lower volume.)

All the cool home automation features you listed are things I don't own yet. I could see adding some of those and growing my system once it's going, but I don't have any immediate need for them (apartment life plays a part there). I can't foresee myself building this out just for those sort of automations (ex. I can hear my smoke detectors from anywhere in the apartment, just look at my thermometers, have no thermostat (steam heat), etc..). But I also loathe these Alexa devices cause I can't get them to just do their job without butting in and suggesting bullshit in follow ups and ads, and I don't need all my dumb timer requests sent to the cloud just for speech-to-text.

FWIW, there's a lot more to it than that. Here's some good info on it: https://www.youtube.com/watch?...

Some points from that:
* Sugar (sucrose) in an acid will quickly be split into Fructose and Glucose (about 50/50 ratio).
* Mexican Coke tested with near zero sucrose content (it had split).
* Corn Syrup is nearly the same ratio of Fructose and Glucose (about 55/42 ratio).
* No, you can't taste the difference between those sugars.
* Mexican Coke had nearly twice the sodium... something you can taste.

I thought I had wasted my time watching that video, but it really stuck with me. I love the idea of sugar instead of corn syrup in Coke, and I like the taste of a Mexican Coke a little more, but this broke my assumptions about why I like the taste of it.

Why would you think that? The fuckin slashdot page title even says, "... Black Market Starlink Terminals ...". Starlink isn't even active anywhere in Russia.

Regardless, if they were authorized terminals, then they'd certainly be able to disable them and be well aware of what all terminals are connecting.

I'm flummoxed; Is the Starlink service incapable of authenticating terminals? Is it less advanced than 90's era satellite TV? IE: how do Russia's black market Starlink terminals get any access in the first place? (FWIW, this isn't a direct response to the parent post... just seemed like a natural progression to ask about individual terminals versus geofencing)

If one lives in the US and has a Starlink terminal, can you access the network even if you are not an active subscriber? I'm certain that's a no. So how are the black market terminals getting around that? And if it's that easy and untraceable, then where can we all get black market terminals?

My first thought as well, though I was looking at it from the perspective of, "The entire US population is only 341 million!!!". 1/3rd of the global population is far fetched as fuck.

We're not talking about the same premise. You're including exfiltration of data. I'm taking the comment literally - "They will likely still listen to and record everything."

The first part of what I'm saying refers to all the "telemetry" data that gets sent to MS. The latter part refers to the massive amount of data collection that already takes place with MS Recall, of which I am not claiming any direct sharing with MS.

You expect me to follow a link AND read that article... from a comment no less??? You've been on slashdot long enough to know we rarely RTFA, let alone TFS, let alone some "see also" link at the end of a comment. If they're talking about data exfiltration, then they should have used their words :-P

In case you're unaware, copy/paste, even on Linux, is context aware and does smart things. No LLM integration I'm aware of, but it's not just plain text. You can try it. In a (possibly Linux) web browser, copy a big chunk of some webpage, then paste that into a new gmail message (or somewhere that accepts context-type: text/html). Bullet points, tables, font sizes, etc.. will be retained. Those aren't plain text.

Even more simple of an example: you can copy/paste images.

Given that there's already some hooks in there to do things with the clipboard data based on context, it's not *too* big of a stretch to see them try to hook in other stuff... not that I want it. I frequently find myself pasting to a temporary vim text file just to ensure I'm removing all the unseen markup before pasting a message (email, comments, IMs, etc..). I should probably setup a shortcut to paste as plain text someday.

Agreed. My first thought was that likely meant they'd do local transcription, then use that for other things. That's how most audio stuff gets processed anyway, and that would skip the bandwidth heavy step of transferring the full audio (when needed, they can just transfer the transcription text). If that sounds like I'm gung ho about it, I'm not. But it is still an improvement in user control; Maybe we could probably swap the local transcription engine with a no op?

WTH are you trying to say? OP said, "They will likely still listen to and record everything." That doesn't state anything about ex-filtrating that data. So if the processing now has an option to be done locally, that means it's doing it one way or the other, does it not?

We're already aware of troves of data that does get sent to them, and I suspect you've heard of MS Recall? If they can justify regularly saving full screenshots, then could easily justify saving a transcript of all the audio they record.

SO MANY posts excusing this behavior because of the licensing fee BS! Wow! Anyway...

Let's go with that and say that's a fine reason to disable HEVC. Should they not offer the ability to re-enable it for said fee!?!?!?

IMHO, it should be required to do one or the other:
A) Ship it enabled. Eat the cost or include it in the purchase price if needed, but ship with the features the hardware has.
B) Ship disabled, but include options to re-enable it, possibly for a fee. Ensure users can actually make use of the hardware they purchased.

If "B" is too complicated for them, they do "A"; Don't just shit the bed.

Right? At least link to the site for the game: https://kingdomsofthedump.com/

I don't get this sort of complaining. If one is using Windows, or Xbox, or a Playstation, or Nintendo, they're on proprietary systems already, and nearly all the games are as well. So what if Chromebooks are invited to the party?

The difference on this, compared to a typical Windows gamer, is the streaming of the entire game. However, this isn't anything new. You can already use Xbox Cloud Gaming to game on a Chromebook (and a lot of other devices). Nvidia already has GeForce Now, which also does it. Amazon has Luna, which also does this stuff.

Maybe you're angling at the bandwidth usage? ... and I'd agree this burns a lot more than running a game locally. I don't think the involvement of Chromebooks makes a lick of difference though. FWIW, Xbox Cloud Gaming also supports Quest VR 2 and 3. For an occasional game session, or to try out other games, I think these services can be terrific compared to the alternative (owning a full gaming system and buying full copies of games).

My question is, why only 10 hours a month!?!? I'm sure that's the only reason it's free, but that should also alleviate some of the bandwidth usage concerns.

While the enumeration of accounts is not impossible for Signal[^1], I'm unaware of any reports about profile pics and profile descriptions being leaked during the enumeration, as we find in this WhatsApp example.

IE: the full impact of this issue is not possible on Signal (AFAICT).

[^1] They did it back in 2020: https://www.privateinternetacc...

But for what values of "Trump"? Here, I believe it still holds true if Trump=0, though maybe it loses a exclamation point or two?

Guess I also overlooked the use case - pro use, traveling, concerts... I was thinking just a static home setup.

Dunno if I agree on the micro-sd card issue. There's loads of ways around that (boot from microsd, data elsewhere; or just boot from a USB hard drive), and I wouldn't trust the storage in a cheap android tablet over good microsd cards. Also, you could actually carry spares of the microsd card with everything loaded and ready to go on it, whereas the tablet option may mean carrying extra tablets... or using the exact same sort of external storage options that are available on the RPi.

But whatever... if you're happy with what you've got and it's working for you, great!

Regarding the picture frame... in my ideal world, it would be pulling the pics from my NAS. I'd be a lot more concerned about the storage on an Android tablet going bad than whatever I boot a Raspberry Pi from (the latter I'd have fully backed up and mirrored; The Android will likely have an out of date OS by the time I get it and no means of replacing the storage). Not to mention screen size restraints - very easy to get a much bigger (optionally portable) monitor for cheap than to get a very large android tablet near the same cost.