For clarification, Organic Maps was the project that has been accused of mismanagement. A significant portion of that community departed to create this app, CoMaps. The goal from the outset was to create a more transparent and open community, hence the name, community maps or CoMaps.

CoMaps is a fairly recent fork of what was already an excellent app, but likely with poor management ethics behind it. I’ve been extremely impressed by the rapid pace of development on a pretty sizable project. There are already hundreds of small (and a few not so small) improvements over the project it was forked from.

It’s a very different approach. Personally, I could never fit OSMand into my daily routine, but this one has been great.

It’s not a reskin, it’s built from the ground up. Although it is technically a fork of a fork (not OSMand).

For me, much more user-friendly and intuitive and even quicker. They both use open street maps data, but I think they are worlds apart. I haven’t done any testing with OSMand for a couple years, so I couldn’t tell you which specific features are different.

I find it very easy to read, day or night. It’s quick to add a destination for navigation. It’s very easy to create updates directly from the app that will upload to OSM.

Sometimes I forget how brutal the early 2000s were.

Having a number of different editors allows manipulating the discussion and concensus protections built into Wikipedia.

Depending on the topic, it may not be necessary. A complimentary article about a new technology product or company founder just takes a few press releases that get picked up. Manipulating world events and leaders requires more coordination.

Although manipulating the sources cited is a great way to manipulate Wikipedia. You have to recruit 10-40 people to act as a group of editors to manufacture concensus across topics. Or you can just create a website or series of press releases.

“Hey, this small-town museum has an article about a historical event. It must be true. Link it at the bottom.” Or “well, this local newspaper article says it is happened, so into the article it goes.”

Even more effective, especially for political groups, is just publish dozens of supportive articles, while miring competing articles in edit wars and the bureaucracy that comes with it. For sources, just cite expert books that are favorable. It’s not easy, but hiring or recruiting 10-40 editors is trivial for political entities.

We honestly need to end the myth that Wikipedia is some impenetrable white tower. It can and has been infiltrated by corporate and political groups, and even creative vandals.

It’s the most valuable digital property in the world. You think people break into the Louvre but can’t touch Wikipedia?

I had the same reaction! I had to log into the screwy web portal and test it to realize it was something else entirely.

If they built out a Mastodon network with government support, then it would.

Uh, Mastodon exists?

So the people killing women for partially uncovering their hair are the good guys?

The people murdering thousands of protestors are the good guys?

So the religious fundamentalists imposing doctrine at gunpoint are the good guys?

I think the people protesting for their lives and freedoms are the good guys, but that’s just me.

My top customizable FOSS launchers:

Kvaesisto
Search-based with native-themed widgets and app drawer. Lots of integrations, reliable, and my personal daily driver.

Lawnchair
Drop-in replacement for Pixel launcher with some friendly features. Very active development.

Ion Launcher
Automatic app folders by category and decent customization options.

Einstein Launcher
Highly customizable and built from scratch. It’s under 3MB and already pretty capable, although it’s a young project and still in alpha

People don’t read them but I think that’s not usually the point. The people I know who have written them usually end up with boxes in their garage that they eventually give at to friends and family.

It’s still a nice accomplishment and a good personal growth thing.

I actually have no recollection of why some records had the big holes in the first place. Were there players with a chonky spindle in the middle?

Keep protesting for the people of Iran. Keep pressure on Western governments to act and support the Persian people. The Iranian people never wanted to live under a fundamentalist terror state.

Economic isolation against every individual official and commander killing their own people is the most important tool we have.

Keep the information channels open at all costs.

London:

Berlin:

Paris:

Just to be clear, companies know that LLMs are categorically bad at giving life advice/ emotional guidance. They also know that personal decision making is the most common use of the software. They could easily have guardrails in place to prevent it from doing that.

They will never do that.

This is by design. They want people to develop pseudo-emotional bonds with the software, and to trust the judgment in matters of life guidance. In the next year or so, some LLM projects will become profitable for the first time as advertisers flock to the platforms. Injecting ads into conversations with a trusted confidant is the goal. Incluencing human behaviour is the goal.

By 2028, we will be reading about “ChatGPT told teen to drink Pepsi until she went into a sugar coma.”

The headline also does not say the same thing that the post claims!

Headline: “15% of content” --> every 6th or 7th post or comment is a corporate troll

Article: “15% of subreddits contain” --> the vast majority of subreddits contain no troll content

Actual study: [file not found]

I also cannot find any Pew research study resembling the one described. The link is a 2017 report that doesn’t mention reddit.

It’s called a jazz bar and they’re not that great.

Unpopular opinion: I have a second phone logged into my kid’s YT account. I train the algorithm while he’s sleeping.

It takes a significant time, and YouTube doesn’t have good options for blocking content, but it helps keep out the worst of the brainrot.

I can’t attest to any as I don’t use PDFs this way, but here are a few links:

Stirling PDF

PdfDing

BentoPDF

All of these are self-hostable and FOSS. I’m not sure about NextCloud integration.