Shout out for ODROID, their product revision cycles take too long (lmao why are they still selling a 32-bit chip that was an iffy investment back in 2013), but when they drop new stuff, it tends to be great.
firmly of the belief that guitars are real
Shout out for ODROID, their product revision cycles take too long (lmao why are they still selling a 32-bit chip that was an iffy investment back in 2013), but when they drop new stuff, it tends to be great.
deleted by creator
Bonus: there is a literally endless supply of used x86 SFF hardware from large institutions, so unlike SBC’s, there’s no special, weird supply chain managed by an English educational nonprofit that could just suddenly decide to not sell to the public for years at a time.
Encrypting your disk only provides at-rest protection, meaning there are entire swathes of physical attacks it provides zero protection against. Tons of stuff a malicious actor can do during runtime with physical access that you’d never notice. it quite literally only protects against thugs smashing your door in and physically walking away with the disk.
So if you’ve painted yourself into a corner with a baby’s first config, what you can do to step up your level of data protection (until you can redo your setup properly) is creating an encrypted filesystem or filesystem image (use fallocate to create a large empty file, then connect it to a loopback device, encrypt with LUKS, and use it as a virtual filesystem), rsync your data directory to it, and then unlock/mount it at boot under the directory where Nextcloud is configured to store your data. It’s god-awful, but this should be more or less transparent to Nextcloud if you do it right, and then at least your data directory gets at-rest encryption, and tbqh if someone is smash and grabbing your hard drive they are probably more interested in your data than they are your OS config.
I wouldn’t say this is an acceptable or preferable alternative to FDE, but it sounds like you’re still figuring out the best ways to set these things up, and this will get you more protection than none. But, realistically, you should probably not worry about it too much and should think about the security of your setup as a learning exercise/study in best practices.
or maybe it’s trying to highlight that it’s also secure?
I wish more content creators would upload to PeerTube (or something like it). I get it, there’s no instances with good monetization options, it just sucks we’re all stuck in various walled gardens because of how expensive video delivery is.
Exactly, It’s a commonly cited example from Manufacturing Consent because he broke it down really well there. But everybody who just read Chomsky for the first time then goes out and tries to correlate every single front-page story with the back-page story it’s supposed to be covering up, like they’ve only ever used a single propaganda technique or being predictable wouldn’t undermine the value of the propaganda.
You know, I appreciate Chomsky, but his work is mainly intended to get you reading and thinking more on your own than to give you all the answers. Not everything that happens in the news media is a distraction from something else just because he broke down that one propaganda trick really well. Sometimes, events stay glued to our screens because they really are the main propaganda event of the day, and they really do want you to spend all day and all night thinking about it.
In this case, Israel needs tons and tons of people frothing at the mouth supporting genocide, and Palestinians are needing just as many, if not more, to consider that genocide may be wrong, and they’re playing tug-of-war in the media. That’s all I’m seeing.
It was a really bad look to scramble to get him back once he triggered a mass exodus. Having him at the helm is either so dangerous for AI safety they had to push him out with a bureaucratic coup, or it isn’t. Doing that severely hurts their credibility on multiple levels (did they really not realize how popular he was within the company and that the price was going to be some of their top researchers?) and after pushing him out the way they did, they should hardly be surprised that Microsoft hoovered him up before the weekend was even over. Why would they give him a few days to process the betrayal and maybe come back around?
After this, we shouldn’t be surprised if Microsoft suddenly starts sabotaging OpenAI until it has no choice but to sell itself off to MS, at which point Altman gets all his toys back.
Stuff like this is why I never took their safety mission all that seriously. It was going to bump up against the business imperatives before long, and given the level of interest business has in AI… what else was the outcome going to be other than corporate sabotage and malfeasance?
Hate that Altman guy, he’s Zuckerberg with more important technology, but somewhere in the mix of articles I read one of the board members complained that “this board is not the group of people you want to see spearheading AI safety.” Yeah, I guess not!
The most useful philosophy I’ve come across is “make the OS instance disposable.” That means an almost backups-first approach. Everything of importance to me is thoroughly backed up so once main box goes kaput, I just have to pull the most recent copy of the dataset and provision it on a new OS, maybe new hardware if needed. These days, it’s not that difficult. Docker makes scripting backups easy as pie. You write your docker-compose so all config and program state lives in a single directory. Back up the directory, and all you need to get up and running again with your services is access to Docker Hub to fetch the application code.
Some downsides with this approach (Docker’s security model sorta assumes you can secure/segment your home network better than most people are actually able to), but honestly, for throwing up a small local service quickly it’s kind of fantastic. Also, if you decide to move away from Docker the experience will give you insight into what amounts to program state for the applications you use which will make doing the same thing without Docker that much easier.
Misinformation is a numbers game. For every 10 people that see the misinfo, only maybe 1 or 2 will ever see the followup proving whatever the misinformation was was in fact misinformation. And out of those 2, half will assume the followup is itself misinformation and have their belief in the propaganda reinforced. Out of the 8 who will never see the correction, maybe two will reject it, four won’t really know what to think (itself a useful propaganda outcome), and maybe two will accept it.
Concerted efforts to combat misinformation can help, maybe nowadays the number of people who see the followup is closer to 3 or 4, but it’s the same basic dynamics behind the Gish Gallop, but on an industrial scale. Making up bullshit is easy, analyzing and explaining why it’s bullshit on average takes longer than spewing out some new piece of bullshit.
Watching videos is like an order of magnitude easier than reading for large swathes of the population. Fully 18% of the US adult population is functionally illiterate – they can pass a reading test, but their reading level is so low it hardly matters. These folks can still watch YouTube/Dystopian Vine (sorry, TikTok).
Also, this much is just my own speculation, but A/V media is a lot easier to push people’s emotional buttons with because it’s much, much faster and easier to consume content via video and we’re likely hardwired to pay more attention to audio/visual stimuli than abstract imagery in our heads. A video+audio track of an explosion is always going to hit people harder than a careful description of the same explosion, and if people expect it to be easier and to provide a larger emotional impact, they’re more likely to go for the thing that makes them feel something more easily.
We are all governed by dopamine more than we like to admit.
It’s like… could they go do some reading, figure out a good strategy, implement it, then make a video about how to do it properly? Well no, that would only be one video per topic, wouldn’t it?
What mistakes are you going to make “building” a Synology? Getting ATA drives?
Hey, it can be the current directory!
deleted by creator
Come back after you rm -rf / or remove glibc, you whipper-snapper! shakes cane
They want to throw this OS on smart home/automative/IoT type things. Android works in these situations, but it’s not necessarily ideal. Thing was designed for phones. It’s likely the only phone firmware in history that’s also been put in cars, espresso makers, washer/dryers, microwaves, and TV’s.
I completely get why the first waves of smart devices tended to just use Android – it’s easy to develop on and “lightweight enough” that the tradeoffs involved were generally acceptable. But those qualities only take you so far. Companies moving on to develop their own in-house OS’s for all these devices was the obvious next step.
Every other Zionist makes it a point to argue that the Palestinians were never a “people” because it was never an independent nation-state, but the only reason you’d split hairs about that in such an arbitrary way (point out where in the UN Convention on Genocide it says "it’s not genocide if they never had a modern nation-state of their own? Y’all) is to muddy the waters as to whether or not clearing out Palestine amounts to genocide. Blows me away people can look at this conflict and not conclude that Israel is committing genocide.
While we’re here, Bibi uses Hamas to weaken the Palestinian Authority and make independent statehood impossible. They don’t dislike Hamas. Hamas makes it very easy for Bibi to accomplish his party’s goals in Gaza, which include ethnically cleansing it. Why would they ever want to actually get rid of Hamas until Gaza’s been completely destroyed?
Hi, sorry I just saw this. “SFF” is short for “small form factor.” It’s just industry jargon for “a small PC.” They tend to be designed to use less power which makes them a good fit for home servers. Pretty much any line of PC sold to businesses, like Dell Optiplex or HP EliteDesk, will have small form factor variants.