Recent comments

Rambler wrote (edited )

Threads got a jump start on it's userbase by making it incredibly simple for current Instagram (and maybe Facebook?) users from signing up to the platform. They already have your data, so you don't need to 'create an account'. It's more of an opt-in, from what I understand. So with that in mind, it's incredibly hassle-free to join which likely is the reason for such high numbers, if they are to be believed to begin with.

It's unlikely to be a Twitter killer in my opinion, but time will tell.

What I find the most interesting when reading comments on the web on places like reddit is the amount of people who dislike Musk so much that they'll support Zuckerberg blindly. Say what you want about Musk but he has at least been part of advancing the automobile, space and telecom industries. Zuck probably still thinks people are "dumb fucks" for trusting him with their data.

(Posted from my Starlink connection. Thanks Musk!)

4

Rambler wrote

Nice, will have to check it out. I'll admit, there are some times where I don't feel like digging through ancient StackOverflow entries or random KB articles and documentation. ChatGPT has come in real handy and has likely saved me a lot of time.

As much as I'm uncertain about AI's role in the future, for now, it's helped me out of a bind or two.

1

Rambler wrote

Care to recommend a 60% keyboard? Bonus if it comes with a travel case!

It'll only be for mobile use with the laptop, as my workstation and daily driver needs a num-pad for me to keep my sanity since I crunch some numbers from time to time. But I'd still like a 60% option for increased portability and ability to pack one in a bag with my laptop.

2

not_bob OP wrote

I'm a big fan of 60% keyboards. Who needs arrow keys? Not when you have tcsh compiled with the build time option for vi key movement! Though, the ones use do offer the option of arrow keys with an FN key. Sometimes you need them.

3

Rambler OP wrote

Absurd.

And like, I get it to a degree. Sure, there are some shitty people online. And sure, some hide behind anonymity.

And there are also people who just don't want giant corporations harvesting data about them or who wish to keep their government out of their daily life, even when doing nothing at all wrong. And for them, there are people that all fall in between or wish to have privacy for various other reasons.

It's bizarre to me that wanting privacy can even ever be considered controversial. No one thinks that because you put up a privacy fence on your property that you're now running a automobile chop-shop or operating a meth lab in the back yard, now that the neighbors and passersby can't see in. Not sure why people assume the worse when people want digital privacy, as well. No one would demand all houses remove privacy fences so everyone can see their yards, which I think can be sort of an apt comparison to digital privacy.

2

Rambler wrote

I just use these cheap Amazon mechanical keyboards, and am happy with them. They're a great upgrade to what I had before, and without breaking the bank.

I'm sort of looking for a 60% mechanical keyboard now for part of a 'mobile office' setup, as I absolutely hate typing on a laptop. Any recommendations?

4

iop23up wrote (edited )

they want to deny scraping of their data. It is not their data. It's the posters data. What about copyrights by the user. This ai thing would fall down very quickly if you have to create/generate your data by yourself by hiring some volunteers on salary. But what they are now doing in twitter is that these data geerators have to pay money for distributing their data. Quite weird. What is needed is a form of copyright which extend its power to segmented/parameterised usage or maybe allow/disallow training usage. Should be easy to verifiy if you can reprodue the original to a limit which indicates that you had this training data. What do thing you will find in some stable diffusion models if you prompt the right parameters. No surprise to find these things there. The consequence of this would be that these datascraper//trainers need to have look to secure that they haven't been trained on copyright material. This discussin is not new. How much parts does make an original and are they consists of parts that are so small that they are "trivial" information and is the amorphed amount of these parts copyright free and become copyrightwith an instruction how to put these parts together in a/the right way? There has been some apps which tries to circumvent the copyright by fractionize the data to "trivial" parts, if iirc correctly the judges don't saw it that way. The difference here is that the instructions will biuld/reconstruct only one output and not variotions of the data like the ai models.

The i2p model does not help to much against this stealing of data i thing. It just make it slower, but it will make the scraper to run nodes,which is good for i2p :). The anonymity of i2p gains some id mixing as you can't really build one id like with an registered user.

My hope is that the people a tired to jump from one to another, but i don't think they in a state that let them recognize the importants of this cooperation fights for their data and access and a huge necessity to have a non big money/data alternative. Which they have to pay/contribute the development/maintainance for.

I don't know if these tech provide the same features as these big cooperation solutions, but i don't think that should be the goal either. The niche is that people need something they scan rely to if they are abandon some service. The question is always how to connect again. That's the niche i think of. Something like: Twitter has deleted my post, but it is still on i2p-x (or bitmassage,i2pemail fed, etc.), here is the link, should be ok as long as 2+ nodes are running... or: Here the link to my always up linklist on i2p-x (i know youc can do this right now by own http server/railroad), here you can find my other locations.

What about something that allows to have a universial post. Using images? If i post i would generate a jpg. It puts metadata from the post like postnumber, reference, keys, hashs into the metadata of the image. The image itsself would be a screenshot of the post(maybe watermarked). The raw post data will be put into the image as stego encrypted or not. This could be useful for reposting etc. and is not format specific. The user will generate it by posting/sharing and stores/send it anywhere for linking/reference. This method will produce bigger post size for sure, but could be acceptable if to find it at all is more important. And you can read the post with every image viewer and do not need to touch any forum/messageservice/app to read it.

1

Rambler wrote

Luckily, faraday bags are (still) legal and easy to make, and cheap to buy if you want one. I've got some name-brand faraday bag that was $25 or so, which works well in all tests I've done. In a pinch can make one with similar effectiveness that may be a bit gaudy to look at, but hey, function over fashion! https://www.instructables.com/EASIEST-FARADAY-PHONE-POUCH/

I'm able to disable the camera and mic on my Pixel 7 w/ GrapheneOS via software, but unsure how well it actually 'works'. I know that when I open an app that requires either, I get warnings that the mic is disabled (As if, I open the phone app) or whatever, which I can toggle back on, if I need to.

France really has gone above and beyond as of late to prove that they hate your privacy.

2