mpweiher

Yeah, Verse also confuses me as an onlooker. Good to know that people that are closer to it share that impression.

As a language nerd, I find some of the concepts interesting, though pervasive use of failure has been tried in Icon and didn't work out well.

I am really struggling to see how

a) the weird bits lead to concrete improvements in actual programming practice

b) that translates to "scripting" usage.

But yeah: it'll be interesting, however it turns out.

Insider Trading as a Service.

Poe's law continues to be a bitch...

We need programming languages that can express the problems at densities similar to the English language prompts.

So: better modularity.

https://blog.metaobject.com/2019/02/why-architecture-oriente...

For "code", substitute "object code".

For "ideas", substitute "code in a higher level language". At least if we actually want to control the ideas. Right now, with those ideas expressed only in English, our control is limited.

What LLM coding does is generate lots of code from a fairly small English (natural language) prompt.

If you look closely, I think you'll see that this is two processes:

1. Translate from fuzzy English to precise machine-executable language

2. Translate from high-level description to lots of low-level code.

Although these two are still pretty great when they are mashed-together, I think they'd be even better, lots better, if we could separate them.

But today we cannot.

Maybe the key difference is that you are so far off the beaten path that there simply are no examples of what you are doing that the models "want" to emulate?

I've also had reasonable success with the models generating fairly idiomatic Objective-Smalltalk, my own language of which there are likely few to no examples in the training data.

I do steer them towards my own sample programs.

A new language now has to clear the ever growing hurdle of not being in the LLM training data.

I found this to be far less of a prblen than I thought it would be.

Do you have practical experience?

TFA:

The type safety we gave up hasn’t been noticeable in any concrete way yet, especially considering our test coverage has never been better.

Algol?

Did you mean APL?

The biggest source of electricity in Europe in the year 2025 was natural gas.

Closely followed by nuclear.

In the first half of 2026, nuclear was the #1 electricity source in Europe, by quite a margin.

Offshore wind is cheaper than coal in China now. Which also makes it much cheaper than nuclear in China.

Citation needed.

China reportedly builds the CAP-1400, a localized and uprated version of the passively safe Westinghouse AP-1000, in 5 years and for around $3.5 billion.

Serial production of a known-good design with a savvy workforce rocks!

If those reported numbers are correct, which I cannot verify, they can profitably sell that electricy at 2 cents/kWh or below.

Because the nuclear issue in Australia is highly politicized and the report is deeply flawed?

https://www.afr.com/policy/energy-and-climate/the-flaws-in-c...

https://theconversation.com/known-unknowns-controversy-over-...

according to Lazard. ... Nuclear new build low end

Here's what Lazard actually says:

“We do not, in this study, try to cost out new nuclear” (2:35)

“We think nuclear will be a big part of the future” (2:47)

“the costs of nuclear should go down “ (12:54)

“next five to 10 years the nuclear bar the one that's most likely to change the most in in terms of cost reduction” (14:06)

https://www.youtube.com/watch?v=16HVh_Fx6LQ

And to claim this is the "low end" is a bald-faced lie.

They used a single nuclear power plant that was the most expensive nuclear power plant ever built in the US and one of the 3 most expensive in the world, ever. Of course, they note that this is the case, and that this is unrepresentative. Alas, all the anti-nuclear activists quoting Lazard are not this honest.

Potentially, but it is much, much safer to dispose of it here.

What's even better is to recycle it, because 95% of the original energy is still in the "waste". And when you do use all of it, the remainder remains radioactive for a much shorter period of time.

Why would you evacuate Long Island?

In Fukushima, there were no radiation deaths, and the long term effects of radiation on the population will be undetectable. The deaths that did occur were due to the unnecessary evacuations.

https://www.sciencedirect.com/science/article/pii/S095758201...

So due to Radiophobia, not radiation.

https://en.wikipedia.org/wiki/Radiophobia

The forced evacuation of 154,000 people ″was not justified by the relatively moderate radiation levels″, but was ordered because ″the government basically panicked″

https://www.nytimes.com/2015/09/22/science/when-radiation-is...

Personal note: the Fukushima accident turned me from a nuclear skeptic to a nuclear supporter. This happened quite a bit. At least for people who actually paid attention.

https://www.theguardian.com/commentisfree/2011/mar/21/pro-nu...

And remember that this was all due to a historically unprecedented earthquake and Tsunami that killed 18000 people and caused half a trillion dollars in damage (in 2025 dollars).

https://en.wikipedia.org/wiki/2011_Tōhoku_earthquake_and_tsu...

During that earthquake, more people died due to breaking dams than of radiation in that natural disaster. Are we dismantling our dams?

There is no 100% safe technology. Nuclear power is the safest form of electricity generation we have, although solar and wind are so close that the differences don't really matter.

According to this NASA study, nuclear power saved 1,8 million lives up to 2011, with many millions more lives saved in the future.

https://www.giss.nasa.gov/pubs/abs/kh05000e.html

On the flip-side, the most consequential negative health effects of Chernobyl and Fukushima came from turning off nuclear power plants and not building more.

https://www.sciencedirect.com/science/article/pii/S030142151...

If the US and the rest of Europe follow Germany's example they could lose the chance to prevent over 200,000 deaths and 14,000 MtCO2 emissions by 2035.

https://www.sciencespo.fr/department-economics/sites/science...

We estimate that the decline in NPP caused by Chernobyl led to the loss of approximately 141 million expected life years in the U.S., 33 in the U.K. and 318 million globally

And we absolutly know how to deal with the waste, and it's not particularly difficult. In fact, we have multiple ways of disposing of the small amounts of waste. NPPs are very secure against terrorism.

Interesting.

I am convinced we have done exactly the opposite: encode structures and particularly connections as sequences of operations over time.

Which are much harder to comprehend.

A lot of software is what I call accidentally algorithmic.

Thanks for the pointer!

You're welcome!

in-process implies exactly the absence of the isolation guarantees that OOP!Kay and microservices share.

OOP objects are in-process and are isolated using language mechanisms rather than machine/process boundaries.

Feasible with in-process REST.

https://news.ycombinator.com/item?id=48731266

Alan Kay's distaste for (static) types

Citation needed.

TFA quotes him saying almost the opposite:

(I'm not against types, but I don't know of any type systems that aren't a complete pain, so I still like dynamic typing.)

Very similar, yes.

Except that Kay did not envision the distinct computers communicating via REST.

https://blog.metaobject.com/2019/11/what-alan-kay-got-wrong-...

Also: microservices currently require at least process and frequently even VM/Container/Machine boundaries.

In-Process REST scales that down:

https://link.springer.com/chapter/10.1007/978-1-4614-9299-3_...

Language support for In-Process REST:

https://dl.acm.org/doi/10.1145/3359591.3359729

https://objective.st/

We did that in my physics high school class.

Then we took away components until we had virtually nothing left, a diode I think(?), and still we had some signal.

Turns out there was a transmitter on the top of the hill the school was also on.

Fun times.

There is valid criticism to the barriers to entry that do exist.

The problems the post describes are not that. They are barriers that the author created himself by selecting a complex corporate setup.

It's not.

Nope.

But the GmbH & Co KG setup the poster wants is not needed for limited liability.

You get that with a plain GmbH (or UG), which is much, much simpler to set up.

The GmbH has the tax structure of a business, as it is a business.

He wanted something more special than that.

Which is possible, but complicated.

I am pointing out that his claims are, in fact, not true.

It is not this complicated to set up a limited liability company in Germany. It is this complicated to set up his choice of a two company setup. Which is, I repeat, his choice and much rarer than a GmbH, at around a 5-10% ratio. Because it is more complicated, not just to set up, but also to run. You have two companies, so two sets of books, two sets of audited returns etc.

Which is why most companies are not GmbH & Co KG. They are plain GmbH. A GmbH & Co KG is a much more complex setup and known to be a much more complex setup. Which he knowingly chose.

And his claim that this is somehow necessary for limited liability, which would be a legitimate "significant tradeoff" is simply not true at all. That's what the GmbH is for.

Now maybe he was badly advised by his lawyer, but then complain about the bad advice. Not about the consequences of choices you made.

One of us is lying.

It ain't me.

As one example of many see previous post on nuclear production records.

Yes, what you are doing is complicated.

I did a GmbH with "Musterprotokolle" so virtually no lawyer fees and quick and easy to set up.

And a GmbH is limited liability. It's in the name. Gesellschaft mit beschränkter Haftung. GmbH.

The Geschäftsführer is liable for certain (avoidable) things in either case.

The difference is that that the Co KG can be taxed like a "Personengesellschaft". So you wanted to optimize taxes, which leads to a more complicated structure.

You are certainly allowed to choose such a more complicated setup if you think it's better for you. But then don't moan about it being complicated. It was your choice.

That is an obviously mis-designed "market" as it only prices in marginal costs. This worked when most if not all producers actually were reliable. It doesn't work when you have a high share of intermittents.

For mis-designed electricity markets, see also:

https://en.wikipedia.org/wiki/2000–2001_California_electrici...

It can be fixed easily by giving priority to reliable producers and only allowing intermittent producers in once the reliable producers are used up.

The whole baseload argument when talking about renewables is a strawman

No. It is actually essential, because intermittent renewables cannot reliably supply electricity, yet we need reliable electricity supply.

Intermittents are not baseload technology, because they cannot reliably supply electricity when needed.

It's even worse: the problems of that reactor type were, in fact, well-known at the time.

Which is one of the reasons that reactor would never have gotten an operating license in any western country. Oh, and not having any containment.