Obsidian has a plug in for this… here is an announcement from the plugin author: https://www.reddit.com/r/ObsidianMD/comments/1bsa6dy/alpha_release_of_my_handwriting_plugin_ink/ (sorry for a reddit link)

Switched from Onenote to obsidian. There was a small learning curve and I had to install some plugins, but I love it. It looks amazing and runs so much faster than OneNote ever did.

It seems like the most significant new feature is called anti-gravity which is kind of an allow list?

We’ve had white lists for a long time so I’m not exactly sure how this will impact the system. That being said I read through the release notes and there are a lot of changes and improvements throughout the system, so congrats to them team on the 6 release.

I’ve had pihole running on my home network for years and I love it.

I think it’s funny because it’s true. Long form written communication used to convey a lot more subtlety than just its content. It’s a tradition that we will lose a bit like other formalities because it no longer tells you useful information about the sender.

I’m not completely sure if the post you replied to was sincere based on its tone but appreciate your well reasoned comprehensive analysis.

I think sites that feel they have valuable content can deploy this and hope to trap and perhaps detect those bots based on how they interact with the tarpit

Honest question, what are the incentives for instance operators to play nice, so to speak? And not just recreate new oligarch safe havens?

It seems like each instance is a miniature zone of centralization and it’s still incumbent on individuals to create their own circles of influence. For better or worse that’s how we get hivemind echo chambers and I’m not sure it’s even in human nature to seek anything else.

Alternatively we have to rescue our friends and families when they start to fall for BS and educate them aggressively on improving the sourcing of their information.

Had one for a while. It’s ok but the places you can safely ride are more limited than you think. E-scooters are a far superior invention for mobility.

My TikTok feed is full of content that I find interesting and educational, from creators who work hard to make something valuable.

For them, banning TikTok means the work they put in to curating an audience will be partially lost, they’ll retain only the followers who find them on another app. If they are monetizing, they’ll potentially have to start over. That may discourage some who are just getting started from developing their craft.

If china, bytedance, meta, or any other platform is collecting user data in such a way as to be a national threat they definitely need to cut it out and this should be regulated. For example, it should be impossible to identify the location of military generals based on where their wives access TikTok from, or who’s having an affair with who based on proximity to each other, or to develop a vast dataset of individually identifiable profiles of every user that could be used to selectively damage their character.

Aside from these problems, which are potentially solvable, I think the individual creator/maker economy is an awesome way to give more power to the people.

The TOTP changes every time. For modern totp hashing I’m not sure how many sequential codes a keylogger would need but I’m guessing more than I will ever enter.

Edit, asked ai for an answer to that because I was curious (maybe it’s right):

Start AI

That being said, if an attacker were able to collect a large number of TOTP codes, they might be able to launch a brute-force attack to try to guess the private key. However, this would require an enormous amount of computational power and time.

To give you an idea of the scale, let’s consider the following:

Assume an attacker collects 1000 TOTP codes, each 6 digits long (a common length for TOTP codes).
Assume the private key is 128 bits long (a common length for cryptographic keys).
Assume the attacker uses a powerful computer that can perform 1 billion computations per second.

Using a brute-force attack, the attacker would need to try approximately 2^128 (3.4 x 10^38) possible private keys to guess the correct one. Even with a powerful computer, this would take an enormous amount of time - on the order of billions of years.

My bank uses a TOTP and they not only block paste, they also block all typing. Instead they popup a modal with a 0-9 digit keypand and the location of each number changes every time.

Effing obnoxious.

Bots don’t paste. If it a selenium related bot it would inject the value or type out each keypress.

It only causes real users pain

Bluesky is being run by a funded professional startup team and is aimed at the masses. Mastodon is run by activists and software devs and brings in other like minded folks.

The smart ones all know how to use VPNs as well. They know what’s up.

Crazy thing is they only need to control the masses who are mostly uneducated or don’t care enough to figure out what’s going on. Turns out that even the USA has a massive group of the latter type.

Accelerate, brake, accelerate, brake, accelerate, brake on repeat when on a wide open road.

Some drivers do this seemingly without being aware and once you notice it’s impossible to ignore how irritating the feeling is.

I guess this device needed to connect to some remotely hosted server that enabled its functionality. And the company was losing money and hoping that sales would eventually pick up enough to make them profitable. But their latest investor decided not to put any more money in, and the company ran out of cash and can’t pay its bills anymore.

The entrepreneur thought he could get more investor cash and ran the business in such a way that it would fall off a cliff if he didn’t. And… He failed to secure more financing.

I have mixed feelings about products like this… If the device somehow needed to host an entire internet’s worth of data to function, it certainly wouldn’t have cost only $800. But when you buy a product that depends on the ongoing viability of the seller, you’re in a position of caveat emptor - You better vet them out yourself, especially if they’re new.

Hopefully the founders feel some emotional attachment to their product and the trust bestowed upon them by their unknowing customers, and release whatever on the back end makes the thing work so that motivated customers could reactivate their devices somehow.

Hah, I installed Postiz just yesterday, interesting to see this thread. It’s like buffer or one of the other paid tools to schedule your social media posts and track engagement. Of course, of particular interest to our community, Postiz is self hosted.

It doesn’t have as many features yet as the major SaaS businesses, but the software is looking good and quite usable right now. I’m sure the more people who use it and support the developer, the more this tool can grow.

For example you can plug in your OpenAI API key and get an LLM chat interface inside the software while writing social posts. But I don’t think it learns your style or creates posts using any kind of system prompt yet unless you type it in each time.

Another thing I couldn’t figure out so far is how to limit which social media channels individual users can see. For example my business has several different units and there’s a different marketing team on each unit, so they shouldn’t be able to post into other channels.

If you’re in the business of needing to post regularly on a lot of channels I think postiz is worth checking out.

Training takes more resources.

Querying takes less resources.

Get a supported version of Linux on there, give it a static IP on your LAN, install Pihole by following the guide at

https://pi-hole.net/

Update the DNS ip that your router gives out to the IP of your netbook, and voilà, most ads can’t load any more.

Even mobile phone apps that have ads won’t show the ads when you’re connected through your home network.

Pihole