Hacker News new | past | comments | ask | show | jobs | submit | domenkozar's comments login

It would be bring less complexity using https://devenv.sh/ to provide the tooling :-)

That was a mistake, for some reason it left a thumbs down instead of thumbs up.

No biggie, I was mostly worried about having proposed something undesired without receiving any feedback about it. I'm glad this wasn't the case.

Do you have a particular in mind or does the general number bother you?

It's a very ambitious project, so a lot of issues are related to either Nix or general support for each language/service.

In devenv.sh:

  difftastic.enable = true;

Hey folks,

We've shipped a pretty big update this time, incorporating a lot of feedback and issues from the last year.

Here's a 5min demo: https://twitter.com/domenkozar/status/1770444437505929637


Better composability and abstraction, see for example how you can configure a process in https://devenv.sh/services/ that has first-class support for development environments using Nix.

No need to use containers too, which bring cognitive/workflow overhead.

And Nix doesn't? At least with Docker you can get something useful within 2 minutes

> And Nix doesn't?

If you're asking if installing/running software with Nix requires containers like Docker does: no, it does not.

> At least with Docker you can get something useful within 2 minutes

  % nix shell 'nixpkgs#ruby' -c irb
  irb(main):001:0> puts "Hello, world!"
  Hello, world!
  => nil
That took me about 5 seconds to type out, and now I'm in an IRB REPL slinging Ruby code, despite having never installed Ruby prior to running that command. Add another 5 seconds for nix to fetch the necessary files (which are then cached locally, so a second `nix shell` invocation will be immediate) and we're sitting at roughly 10 seconds end-to-end.

If it takes 2 whole minutes to use Docker, Docker must be pretty bad. I guess vintage things have their appeal, but apart from that I don't see why people would be so attached to such archaic, inferior technology.

But is it pure?

That's a mouthful for 14 lines of native development that can generate a dev container using https://devenv.sh

$ cat devenv.nix

{ pkgs, ... }: {

  languages.c.enable = true;

  packages = [ pkgs.ceedling pkgs.cmake ];

  enterShell = ''
    cmake --version

  pre-commit.hooks = { 
    clang-format.enable = true;
    clang-tidy.enable = true;

$ devenv shell

(Note that pkgs.ceedling has been recently added and will only hit the cache in 6-12h).

Note that warm waters is a typical strawman argument against nuclear power.

These are rare anomalies and there's dry-cooling of nuclear power plants (although it's less efficient and costs more).

These are as rare as the week long events of low wind. Meaning that we must plan for nuclear curtailment just as we must plan for unusual wind weather: with backup generation and with storage.

Trying to brush aside legitimate engineering challenges as "not real" seems far too common among nuclear advocates. Which is my guess that their construction projects fail so often; the engineering and logistics and construction are significant challenges that are not taken seriously enough.

If the nuclear industry took engineering and problem solving as seriously as those in solar and wind, we would probably have a lot more nuclear around, a lot more successful construction projects, and nuclear that was cheap enough to build.

> Note that warm waters is a typical strawman argument against nuclear power.

A strawman is a false opposition argument set up to argue against, so, no, its not. I am not even sure what you are trying to say, but “strawman” isn’t it.

Also, most proposed new reactors a aren’t dry-cooled and the arguments, which include cost, for nuclear don’t assume that higher cost option.

Plus, it's a design parameter in a performance/cost trade-off.

It's _designed_ to not work at full power at this heat, because that was thought to be the ideal trade-off.

Maybe it still is, maybe they underestimated the occurrence of high water temperature incidents, but in any case it's a consciously designed safe state.

Sure, but such mitigation strategies mean even higher costs for nuclear power which is by far its largest problem.

It’s clearly possible to make a great deal of nuclear power safely, just not as cheaply as similarly environmentally friendly alternatives. Electric utilities prefer to spend less on battery backed Solar etc because of all the little details that aren’t obvious until you really study what’s involved.

Can future designs have the operating temp shifted in order to accommodate warmer waters?

You mean the design of the ecology around? No, not in a reasonable timespan ;)

Plus, most reactors were built in the 70s/80s when waterflow of rivers was more plentiful and less warm on average. Ironically, nuclear contributed to none of that.

"Ironically, nuclear contributed to none of that."

Only when you think building them, maintaining them, mining Uran, shipping Uran and shipping and storing the radioactive waste has no CO2 footprint.

Now sure, we still might have been better off, if we would have replaced all the coal plants with nuclear by now. But we did not and now we have to work with what we have.

Rare anomalies currently. Will they continue to be rare anomalies going forward? Also, high heat moments are the times when you likely need more power than ever...

Is it not a chicken and egg? If we don’t build waters get warmer faster from the existing primary sources of power, albeit indirectly

There are scenarios where it could be a chicken and egg. But there's also likely many scenarios where you ought to acknowledge that it's a dumbass place to build a nuclear plant because the water supply it depends on are not reliable and are expected to get worse with no regard to the plant itself. I'm not saying that's the case here. But... I would guess it's likely.

Very curious, where do we have the reliable water supply? Just move existing fleet, or build all new now with what we fail already? And then cool them with the unproblematic salty sea water, boiling the oceans that also start getting temperature problems already? Very confused..

Creating heat to turn into electricity is an outdated 19th century idea at this point.

Now that we’ve mastered the technology to turn ambient energy directly into electricity, traditional nuclear reactors are an overly complex technological dead end.

Though it seems inefficient why don't we see alternative methods take over?

We are seeing alternative methods take over....

Check out the interconnection queue for new generation where there's price competition, and you will see it completely dominated by non-thermal tech.

And the thermal tech that is there, natural gas, is partially combustion turbine driven, and without that combustion turbine component it would likely not be competitive at all. It's likely that within a decade a lot of those new natural gas CCGT assets will be completely stranded and uneconomical.

And as with any high-capital established industry, there are a lot of dinosaurs that will not move until they die off. They will be victims of creative destruction, rather than survive and pivot sooner.

I'm seeing them? (:

These are rare anomalies

Note that partial shutdowns due to excessive heat happen regularly in France, e.g. in 2018, 2019 and 2022. The problem's been around for a while, see e.g. this article [1] from 2009 that also mentions the heatwave of 2003, where regulators had to grant special exemptions to allow discharging 30°C water into waterways, well past the 24°C limit.

[1] https://web.archive.org/web/20110612153407/http://business.t...

> Note that partial shutdowns due to excessive heat happen regularly in France, e.g. in 2018, 2019 and 2022.

And they affect a very small number of plants and energy output. The largest disruption so far has been when French government finally got its head out of its butt and stopped a few plants for long overdue maintenance

Even before the most recent maintenance period, there were problems: In 2019, French regulators had to ask industry to cut consumption by 1.5GW to keep the grid stable as utility frequency was dropping significantly [1]. That issue comes up basically every other winter, but so far, we've avoided catastrophic results.

[1] https://www.cre.fr/actualites/RTE-fait-appel-aux-industriels...

They asked to cut consumption because of nuclear power plants or because insufficient supply of electricity?

Insufficient supply in winter times due to electric heating, mostly. Nuclear power plants are relevant insofar that they (in particular the older generation of power plants) are bad at providing energy on-demand. So while everyone worries about Germany and the issues associated with variable availability of solar and wind power, there are also documented cases where Germany had to fire up its coal plants to meet nuclear posterchild France's electricity demands.

I used to read German online magazine Telepolis regularly. They've got a writer who advocates for renewable energy, hence I used to come across related articles every now and then.

> Nuclear power plants are relevant insofar that they (in particular the older generation of power plants) are bad at providing energy on-demand.

This is, of course, an easily verifiable lie.

During winter nuclear power plants already work at near 100% capacity. They can't give you more than 100%, other sources cannot meet demand, and somehow you blame nuclear.

> This is, of course, an easily verifiable lie.

It's not a lie, you just failed to get the point:

Assume, for the sake of the argument, that you have a power plant that always works at 100% capacity and cannot be shut off. To provide energy security, you would have to budget capacity to account for highest possible demand. But if you do so, you will over-produce electricity most of the time, and there are economic incentives against doing so.

> somehow you blame nuclear

I don't blame nuclear energy production, I blame an over-reliance on nuclear energy production.

Okay, you keep approaching this with a abias, false assumptions and false premises.

This is the last I'm going to say on the matter:

1. "Nuclear power plants ... are bad at providing energy on-demand."

This is a lie. All nuclear plants, at least in Europe, are required to increase and decrease their power output on demand.

"according to the current version of the European Utilities Requirements (EUR) the NPP must at least be capable of daily load cycling operation between 50% and 100 % of its rated power Pr, with a rate of change of electric output of 3-5% of Pr per minute." [1]

Daily. Between 50% and 100% of its rated power.

The lie that nuclear power cannot provide energy on demand is a lie that people keep perpetuating.

2. "for the sake of the argument... a power plant that always works at 100% capacity... I blame an over-reliance on nuclear energy production."

If your power plants always work at 100% capacity, it's a failure of planning, not of the plants.

It's amazing that you brought up Germany in one of the comments. It just goes to show how biases make a person completely blind/oblivious.

Germany has shut down its nuclear power plants. Now it needs to burn ungodly amounts of fossil fuels and import ungodly amounts of electricity every time there's a windless night [2]. Because they decided that renewables are enough, are over-relying on them, and made no plans for when they are not enough.

This is colloquially known as "fuck around and find out". Aka "you can't attribute the failure in general planning to the performance of a single actor".

Yes, France relies on nuclear power. And yet, they have done fuck all: they have neglected maintenance of their nuclear plants for decades, they didn't plan for increased electricity demand etc. Much like Germany with shutting down their plants: they never calculated the actual electricity needs etc.

But people like you keep saying things like "nuclear is bad at providing energy on demand" etc.

[1] https://www.oecd-nea.org/upload/docs/application/pdf/2021-12...

[2] On the week when they celebrated shutting down their last reactor, there was a night when they had 0% solar production, and 0.2% wind production, both of which could be covered by just a single reactor from theos they had shut down.

The largest US nuclear plant is Palo Verde just west of Phoenix were temperature can get crazy hot and nowhere near a body of water. It’s running with waste water of the city. So I’m pretty sure in a country way more humid like France you can have a backup plan, like building larger condenser towers.

Sure. It's not so much a physical limitation, but a failure of policy.

To what extent is it a strawman when it comes to nuclear power generation?

These rare anomalies could happen more often because of climate change and the existence of dry-cooling power plants doesn't help if yout already existing isn't

Never heard of dry-cooling nuclear reactors, how much less efficient, also less safe? Does a prototype even exist? Some reference please.

Will we call the rare anomalies rare until it they are the norm? And then?

A lot of these pitfalls and boilerplate, we've removed in https://devenv.sh :)

The best talk I know about this is https://www.youtube.com/watch?v=t6goF1dM3ag

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact