phba

Awesome work. It's great to have a FOSS alternative while (Embarcadero) Delphi gets worse with every new release.

"Turing's Cathedral" by George Dyson is also great. It's a kind of broad biography of all the great minds that started the computer age.

Maybe I'm imagining things, but every HN thread about a new AI model seems to follow the same pattern, has the same arguments and talking points. The only difference is the version numbers of the AI models mentioned.

Ted Nelson apparently predicted it as early as 1965. In his autobiography Possiplex there is a chapter named "My Box of Crayons, 1965", where he shows some of his early sketches, one of which looks just like any ad-infested blogspam website from today. He used it in an ACM presentation, but I can't find it on the web.

> For instance, I am wondering what exactly you are responding to with this comment

It's an ad for their website.

There is a game on Steam called "Turing Complete". It teaches you how to build a fully functional computer from scratch. I recommend you play this game while you work through the theory.

Not just him, but all the people in his cultural sphere. I've been to a Banksy exhibition, and it also had videos of "critics" commenting on his work. The overtone was how inspiring and brave it is to protest things like war and injustice nowadays in a western country. It's repulsive how ignorant these people are towards their own privilege, while taking the moral high ground and lecturing others.

And of course there was a fucking gift shop at the end.

Yeah, I don't understand why it has to be agile XOR waterfall. Agile development simply doesn't work in projects that have so many externally imposed constraints that there is barely any flexibility left.

Why would someone go to a place in an actual conflict zone that is under attack by actual drones right now to flee from hypothetical drones in a hypothetical conflict?

The primary goal of these efforts is to control communication and the flow of ideas. Information is a control mechanism, since we act on what we believe.

In history we had four media revolutions (printing press, radio, television, Internet), each greatly disrupting and reshaping society. This is the fifth (social media and maybe AI).

All these revolutions had the same theme: increased reach of information, increased speed of transmission, increased density (information amount per unit of time), and centralization of information sources. Now we seem to reach the limits of change. No more reach, since our information networks span the entire globe. No more speed, since transmission times are close to how fast we can perceive things. The only things left to change are even more centralization and tighter feedback loops (changing the information based on how the recipient reacts).

Given all that, this media revolution might be the last one, so there is a gold rush among the elites to come out on top.

Maybe this is the first step towards the Big Beautiful Bailout when the AI bubble inevitably pops.

1. Ask questions, and write down the answers in a way that you will find them again. Anki and spaced repetition is useful to learn the terminology or any info that isn't intuitive.

2. https://github.com/ohenley/awesome-ada has links to pretty much every Ada topic and resource; if you want to try Ada using open source tools, the best starting point is https://alire.ada.dev/docs/

3. Compared to C/C++ I can't really think of any pitfalls. It requires more discipline and formal reasoning, but you will get used to it (and appreciate the lack of footguns, at least I did).

Congrats and good luck.

Yup, same for DuckDuckGo.

> In literate programming you do not write the code then present it to a human reader. You describe your goal, assess various ideas and justify the chosen plan (and oftentimes change your mind in the process), and only after, once the plan is clear, you start to write any code.

This is not literate programming. The main idea behind literate programming is to explain to a human what you want a computer to do. Code and literate explanations are developed side by side. You certainly don't change your mind in the process (lol).

> Working with LLMs is quicker though

Yes, because you neither invest time into understanding the problem nor conveying your understanding to other humans, which is the whole point of literate programming.

But don't take my word, just read the original.[1]

[1] https://www.cs.tufts.edu/~nr/cs257/archive/literate-programm...

Pretty impressive!

Yeah. This is the curse on any legacy software that doesn't enforce strict separation of logic and UI. Any larger change to the UI requires an awful lot of manpower that open source projects usually don't have.

I wonder if it would be possible to extract the spreadsheet data model and logic into a library completely separate from the UI. This would enable a diversity of UIs, and also interoperability between different tools.

> a hobby that already raises eyebrows

Sounds very interesting, but may I ask how this actually works as a hobby? Is it purely theoretical like analyzing and modeling, or do you build real rockets?

There is no incentive to implement such a feature. Bots and paid social media workers drive engagement. Also social media sites are designed to avoid any triggers that make users click away (like showing origin flags that would allow a user to easily dismiss a thread as fake). This is the same reason why Youtube removed dislike counts.

A good place to start is OSINT (open source intelligence) for your city/municipality because it requires little commitment, is scoped with regards to complexity and amount of information, and usually risk-free. Gather publicly available information about the companies in your area, who owns/runs them, your city council, any ongoing projects, the processes of funding stuff with public money and so on. Don't bother finding the best collection method or way to structure all the data, just start, you will figure things out on the way. Also be aware of your personal bias, which might make you dismiss important information or affect your judgement.

The next steps highly depend on where you live. Your HN profile says Australia, so at least safety-wise you are in a better spot. Connect to people in your area (preferrably offline), for example by organizing a local meetup, maybe there is one already. Activities can range from exchanging ideas to spreading awareness in your community to actively going against corrupt affairs. Make sure you know what and who you are up against, or you will have a very bad time.

Anticorruption is a group effort because it requires a lot of work and often special knowledge (info tech, law, finance, opsec, public relations and propaganda, ...) and, more importantly, a group provides safety from corrupt actors. On your own you will not be able to deal with lawsuits, misinformation, character assassination and worse.

Thanks for the link. This is a great analysis of Microsoft's open source strategy.

> AI enables precision influence at unprecedented scale and speed.

IMO this is the most important idea from the paper, not polarization.

Information is control, and every new medium has been revolutionary with regards to its effects on society. Up until now the goal was to transmit bigger and better messages further and faster (size, quality, scale, speed). Through digital media we seem to have reached the limits of size, speed and scale. So the next changes will affect quality, e.g. tailoring the message to its recipient to make it more effective.

This is why in recent years billionaires rushed to acquire media and information companies and why governments are so eager to get a grip on the flow of information.

Recommended reading: Understanding Media by Marshall McLuhan. While it predates digital media, the ideas from this book remain as true as ever.

Apparently X11 has a security extension [1]. There was a discussion some months ago [2].

Xenocara (X on OpenBSD) improves security by dropping privileges and using features like pledge [3], but I don't know how this affects the feasability of keyloggers.

[1] https://www.x.org/releases/X11R7.6/doc/xextproto/security.ht...

[2] https://news.ycombinator.com/item?id=44768745

[3] https://man.openbsd.org/pledge.2

This "joke" is neither funny nor original as it comes up on social media everytime someone mentions Russia as a threat to Europe.

Oh, look at that. A fresh account just to make this comment. What a coincidence.

Adding to that from https://en.wikipedia.org/wiki/Thorn_(organization):

> Thorn works with a group of technology partners who serve the organization as members of the Technology Task Force. The goal of the program includes developing technological barriers and initiatives to ensure the safety of children online and deter sexual predators on the Internet.

> Various corporate members of the task force include Facebook, Google, Irdeto, Microsoft, Mozilla, Palantir, Salesforce Foundation, Symantec, and Twitter.

Apparently Thorn scratched that list from their current website, but the Wiki page has an archive link.

Thank you for the recommendation! I was looking for compiler texts that take a more practical approach.

Well, it's impossible to know if Eichmann really thought about the ethics of his job. With regards to his trial, it didn't matter anyway. Every captured Nazi official claimed they were just a cog in the machine and had to follow orders. The judges rightfully dismissed this, because otherwise Hitler would have been the only one responsible (how convenient!). They were all put on trial for their specific actions and decisions.

My take from Hannah Arendt's work is that normal people will do evil things if they think they can get away with it.

The legislation simply says if you collect more data about your users than necessary, you must inform them and they must consent. This has nothing to do with cookies or any other tech.

The question a user should ask is why is this website collecting my data. Marketing and adtech companies are trying to shift this question to why is the EU making websites worse.

> there is no need for government regulation here

You don't need to care about this if you respect users' privacy in the same way you don't need to care about waste water regulation when you don't pump waste into rivers.

Using uBlock you can filter HN topics by keyword:

  news.ycombinator.com##tr.athing span.titleline > a:has-text(/(Israel|Gaza|Palestin)/):upward(tr)
  news.ycombinator.com##tr.athing span.titleline > a:has-text(/(Israel|Gaza|Palestin)/):upward(tr) + *

Credit: https://news.ycombinator.com/item?id=33546513

Much easier than arguing whether this topic belongs here or not.

I don't agree with two of the four given reasons for the rewrite.

> Modern codebase: Easier to maintain and evolve compared to 30-year-old C code.

It's not a modern codebase, but a NEW codebase, so ofc it's easier to maintain and evolve. How many 30-year old Rust projects are there to actually validate this claim?

> Younger contributor base: Young developers are opting for modern language like Rust instead of C. Rust's safety features also make it easier for new developers to contribute more confidently.

I like that a lot of young programmers work on open source projects, but the old-is-bad bias hurts software projects because so many hard earned lessons get lost and relearned over and over again.

Besides that, why call this a "takeover"? I like Rust, but this whole endeavour screems immature. Sudo is a critical component in millions of systems, please treat it appropriately.

From the second link:

> The privacy considerations for digital credentials are not static. They will evolve over time as the ecosystem matures, and may be informed by the behavior of other actors in the ecosystem, improvements in other layers of the stack, new threats to user privacy, as well as changing societal norms and regulations.

Boil the frog slowly and carefully, and look out for opportune moments that could help to speed up the process.