I gave it a try, but what turns me off is the weird decentralization that’s sort of black box? Like I have a recovery phrase which I associate with blockchain stuff, and there’s a vague button that says “offload data to our backup node”. And then I seem to have an account with them? The settings mentions deleting an account which is weird, because I thought it was local/lan sync only.
Their website says “No server”, but in the settings on the app it says I’ve used xxMB out of 1GB of remote storage, where/what is that if there’s no server involved? Where is my data being uploaded to?
I can’t seem to find where it stores data in a standard format on my local filesystem, so if anytype shuts down how do I migrate? It looks like my local data is even encrypted for some reason??
Basically both on their website and in the app it feels like the concept is all over the place, it can’t decide if it’s local where you own your data, stored on a server somewhere, or some sort of weird blockchain decentralized thing where your data just might vanish one day.
For the app itself I can’t figure out how to get an editing/format tool bar like I have in onenote, to change font, size, headings, insert tables, and that sort of thing.
Navigation is also confusing, I created a new note (page?) and now I can only find it in “All Objects” which is just a giant mess of stuff, whereas I’m looking for something like a tab bar with my sections and pages organized in a tree or something like onenote does it.
Overall my impression is it’s very confusing to use and understand, with a lot going on in the UI but still missing basic editing tools and organization.
For android, Google uses Firebase Cloud Messaging, basically a server that pings the phone when a notification for an app is available, which wakes the app up to receive the notification. There are alternatives but they need to be adopted by app devs for them to work.
For people running a degoogled android, they’ll notice most apps won’t receive any notifications until they open the apps since most apps rely on Google Play Services to receive a ping from FCM.
I don’t have any google play services so most of my apps don’t give me push notifications but I do have WhatsApp installed and that still receives notifications, they’re sometimes delayed by a few minutes which makes me think Meta have their own implementation/alternative to FCM but I’m not sure.
For Signal, their servers tell Googles FCM servers that you have notifications waiting on Signals servers and to wake up your Signal app so it can communicate with Signals servers to receive your messages.
WhatsApp and Signal claim/have end-end encryption on their messages but that shouldn’t matter when specifically looking at Googles FCM servers so, at most it would be meta data that could be obtained from the FCM servers.
jami.net/unifiedpush/ has a pretty basic explanation of push notifications on android and also showcases an alternative to FCM unifiedpush.org which has a nice little diagram about push notifications on android. Unfortunately, Unifiedpush is not widely adopted by many applications.
So there are ways to avoid Googles FCM servers on android using Unifiedpush or always having the application on in the background but for the most part FCM is used.
I’m pretty sure Element stays active in the background, it may have asked you to turn off battery optimisation and have a silent notification always active. This decreases battery life which is why most apps don’t do this but it allows the app to constantly ping the server to check for new messages and is one way around using FCM.
Fair email uses en.m.wikipedia.org/wiki/IMAP_IDLE instead of FCM, I’m no expert and this is just my guess but it seems to also need the app to run in the background for this to work.
Silence is SMS and MMS only and so doesn’t use internet and so has no need for FCM or any alternative anyway
Curious about this too. From what I could find, for those it seems like the push is being used to wake up the app and tell it to connect to the server where it grabs the data and then creates the notification locally. Even if a bare minimum is used there is room for traffic analysis, and I imagine Google can easily tell the app being targeted for the push, but it shouldn’t mean the contents of the displayed notification are necessarily what was sent through the server. It’s hard to find info without digging because consumer-facing stuff just calls every notification a push notification.
The alternative is an app keeping a constant connection open to the server, which understandably mobile OSs don’t like. With push only the one service needs to keep an open connection to provide updates for all the apps.
I guess that would explain the difficulties some apps face without push notifications and releasing APKs. These big companies want you to rely on their systems. Signal was pushing their app through play store. I don’t know if an equivalent exists, but it really needs to. We need this, combined with f-droid, so we don’t have to use spyware like the Play Store.
Signal does in fact distribute an APK that isn’t dependant on Play Services/FCM on their website. Uses a websocket, so not the most elegant way I guess, but oh well.
It’s rather hidden, which I think is disappointing. But it exists. Updates itself, too.
I tried a bunch of these note taking apps and didn’t really get on with any of them. I now use the Vscode/Vscodium extension Foam which essentially gives all the note taking features and graph view in what is already a good text editor.
So long as you are mainly writing notes and don’t need the database features that AnyType has or the DataView plug in Obsidian has (though there might be another vscode extension for this - I’ve never looked), then you’d be fine.
Neither is a teen freely sending a nude selfie of themselves to their teen SO.
Many family photo albums feature naked kids in various stupid situations that were just thought to be cute and funny at the time, which would nowadays melt peoples brains if they'd hear about it, and certainly would fall under the same laws that would punish the teen in the other example.
Not to mention it is all based on the questionable assumption that porn is somehow harmful for teens to watch but the exact same porn is not harmful for adults.
‘cause if there’s one demographic that couldn’t possibly have the aptitude, resourcefulness or motivation needed to defeat a scheme like this it’s horny teenagers.
There is a 0% chance that AI can accurately determine if someone is 18 or not, even with hypothetical futuristic AI technology. Some 20-year-olds look very young. Some 16-year-olds look shockingly old. And nobody changes very significantly between the day of their 18th birthday and the day they were 17 years, 364 days.
Hell, even 13-15 year olds are looking much older. I genuinely saw one in normal clothes, took a guess based on her being around maybe 20. Saw her a day later in school uniform. And only under 16’s wear them.
Really did make me realise I am shit at age guessing.
To quote the Simpsons: “0% is a percentage as well!” And that will be more than enough for politicians who know nothing about the topic and are blinded by the hot tech-buzzword of the minute (especially if it turns out they or some of their friends can make a shitload of money with it. I love capitalism and democracy.)
As an outside observer, UK politicians (even Conservatives) seem to tend to be a bit better at this sort of thing than American or Australian (“the laws of mathematics are very commendable, but the only law that applies in Australia is the law of Australia”) politicians. There’s a much stronger tendency for their back benchers to vote against the party line than we have, too, which is great for deliberative democracy.
To be fair, that at least is hypothetically possible. Working out someone’s age exactly purely based on their face is not even possible, so you can argue against it very easily from a purely technical standpoint.
Facial recognition with a database is quite good today, and will only get better. To argue against that you need to start getting into the privacy and ethical arguments.
Exactly, this is something that even humans would have a hard time doing. Even though AI can do many things better than humans, humans are better at vision at the moment.
Eh, humans are better at certain kinds of vision—particularly on tasks that deal with non-white people where the AI was trained mostly on white people.
But things where the vision is looking at very fine detail, AI is very good at. Like determining if a patient has a disease based on a retinal scan or other medical imagery.
And I think it’s fair to say that, at least superficially, the problem in this thread seems like it might be more similar to those medical cases where an AI could do a really good job. The problem is that actually, no. There’s no known marker that could determine age with the level of accuracy that would be required for this task.
The government says selfies for AI verification are to “protect” children from such content but in reality we know it’s all to create a database with biometric hashes of everyone and create the long-awaited dystopia of 24/7 mass surveillance. The government that completes the task first comes out ahead, and it seems that China is winning the dispute
privacyguides
Oldest
This magazine is from a federated server and may be incomplete. Browse more on the original instance.