h4kor

In fact I don't recall ever seeing an ad for tiktok and yet it is humongous.

For the first years of its existence I only new tiktok because they were advertising everywhere.

This is 100% LLM generated; website, documentation and tutorials. There is no link to downloads or a repository. No way to use anything.

Why should anyone care about this?

These exist. There is "Consent-O-Matic" for example

At my former company we paid 900€/month for 1Gbit/s, which we required. That's definitely a problem when the same performance is available for ~50€/month a few kilometers over the border (in the Netherlands).

In rural regions workshops output is limited by their internet speed as the can only download that many CAD files from customer per day.

Reasons I can think of (as a German) from the top of my head:

- Crumbling infrastructure

- Decades of missing investments in education and the public sector

- no digitization

- Unwillingness to move away from ICE vehicles

- Slow internet access and slow build out of fiber network

- Killing future industries (solar, battery ...) by cutting funding/subsidies early

- low wages in European comparison

The main reason I use postgres instead of SQLite is that I have multiple processes accessing the database, often 1 web service for API/Website and a worker running in the background doing heavy tasks (e.g. image processing). Both need access to the database and SQLite will run into locking issues.

How do you overcome this with SQLite and Django?

IIRC there has been a case in Germany where it was ruled that opening devtools is hacking and therefore illegal intend. A person found a vulnerability by looking at the website source (using devtools) and informed the company. They then sued him using the "Hackerparagraph" (§ 202c StGB) for use of hacker tools.

This isn't a trade "just on paper". You need real hardware integrated into the grid.

I'm working on my own mini time series database.

It only stores (timestamped) floating point values with a series id and uses a B+Tree as the backing data structure. Querying is done with a lisp-like query language.

I've added ActivityPub to my blog system (https://h4kor.github.io/owl-blogs/) as just another way of subscribing to the blog, but found that it is a good model to support interactions.

Are companies just sending the output of their liquid cooling into the drain?

basically yes. Closed loop systems require more energy and are more expensive investments. The main concern for data centers is Total Energy / Energy Compute.

There was a good talk about this at 37C3 https://media.ccc.de/v/37c3-11796-energy_consumption_of_data...

I used and am now using plain git over ssh. I only hosted gitea to have my code accessible for anyone interested.

My "main" projects I want to keep public are moved/cloned to GitHub, the rest I just don't bother with any longer.

I had to take offline my personal gitea instance as it was regularly spammed, crashing Caddy in the process.

Quiet impressive!

I've build a very similar game for a 3 hour game jam once :D

https://h4kor.itch.io/herding-simulator

Same in Germany. Farmers (most conservative population group) love renewable energy. Every farmer I know tries to get wind turbines build on their land and have their roofs packed with PV. Yet the conservative parties are the biggest blockers for the energy transition.

IMO it's only a matter of time before LLMs will include ads. Current pricing isn't covering the costs of running LLMs and brands will pay a high price for being favored in responses.

I'm using SQLite as the single storage system for my blog, including media files. I did a simple performance comparison for this use case: https://blog.libove.org/posts/serving-binary-files-from-sqli...

TL;DR: No real difference in low throughput environments, 2.3x slower on high throughput

I'm working on print outs only, as I don't use digital maps.

I'm working on my dungeon planning tool for tabletop games.

Currently I'm implementing the export of a player map that can be partially revealed.

https://github.com/H4kor/dungeon-planner

I watched the last starship launch on a scam YouTube stream. It was 30 minutes delayed, which a thought was because Musk wanted to promote Twitter. They said multiple times that Elon will announce something big after the launch. Directly after lift off it cut to Elon holding a speech and I only noticed this been a scam channel when he talked about the QR code and crypto.

Find hobbies, things to do besides programming. You will find problems to solve for yourself eventually.

I play tabletop games and lacked a good tool to plan dungeons as a DM. So I wrote a dungeon planning tool. ( https://h4kor.github.io/dungeon-planner/ )

I like blogging, so I wrote my own blog software ( https://github.com/H4kor/owl-blogs ) as I disliked WordPress and static sites were too limiting for me.

Policy failure. South Korea enforced a "Sender Pays" rule for networks, eliminating peering between ISPs. This resulted in companies moving there server to neighboring countries to avoid paying for traffic, which was free before.

More details: https://www.internetsociety.org/resources/internet-fragmenta...

Get a binary format supported by all standard browsers with the same simplicity as JSON and you will see a rapid adoption.

How does this compare to list based solutions? An overblocking/underblocking comparison would be great

Yes!

No. There is an interview/documentary about South Park where they talk about celebrity appearances in the show. They said that they never request celebrities to play themselves in the show. If the person declines the offer, they can get into legal trouble for imitating them. If they don't ask it is fine.

The problem here is that they wanted Scarlett Johansson and imitated her when she said no.

There is also a big market for "celebrity sound alikes" in voice acting.

I had a similar, frustrating experience last weekend trying to copy data from a phone to a windows machine.

Windows is trying to hide all "technical" stuff so hard that it becomes impossible to do anything.

Timescale hypertables automatically have an index on the timestamp. To make this more comparable you could create the same index on the normal table and test the ingestion rate.

I'm a total trading noob. Can you explain why a low latency is worth so much? And how these traders "exploit" that advantage to make a profit?

This is an ad and it reads as if written by ChatGPT. Some surface level bullet points about Access Control, link to their product and the most basic SQL statements.