News for hackers. Stuff that matters.

mhh__

I do worry about communications in disasters and so on but I think the plan was always to cut the phone lines in that case anyway
Re:Python's experimental JIT to be removed on Tuesday October 27, 2026 @12:49PM (#50261970)
Why is this controversial? We spend hours waiting for python code when it could be much faster if it was treated like a serious programming language
Re:Python's experimental JIT to be removed on Tuesday October 27, 2026 @08:40AM (#50258239)
It seems like such a failing and waste of human life that we spend all this time waiting around for python when the JS engines are so cutting edge.
I don't know anything about the polish electricity system but if supply or demand isn't very elastic then excess of either can lead to pretty huge price fluctuations and knock on effects thereafter.
Graun can still be very good at times but i think it's a fair comment at this point that they're best assumed to be wrong about most things.
Re:DeepSeek just open-sourced its inference engine on Thursday October 22, 2026 @07:36AM (#50193971)
Interestingly doesn't seem to be slop coded
Re:C# now beats Rust on WASM size on Wednesday October 21, 2026 @07:56PM (#50189267)
The JIT adds basically nothing to the language other than complexity and also forces a boundary between dotnet and it's peers that I think hobbles it via harder interop.

The performance in theory can be slightly better in some niches but my point is simply that it's a huge chunk of complexity for not much gain.

Re:C# now beats Rust on WASM size on Wednesday October 21, 2026 @07:53PM (#50189244)
numerics is I think OK but very expensive
Too large, if I may say such a thing.
Re:C# now beats Rust on WASM size on Wednesday October 21, 2026 @07:58AM (#50179434)
Yes I think it's a bit underrated. The problem I find is that no one writes C# libraries that actually do anything (e.g. even basic linear algebra is pretty crap compared to python) and that the JIT is a pointless hindrance but the experience of using the stack is pretty good.
Re:A skeptic's first contact with Kubernetes (2024) on Sunday October 18, 2026 @04:45AM (#50139208)
My own version of this is that I was actually quite pleasantly surprised using k8s for the first time, expecting it to be much more complicated, but then pretty shocked at how complicated the ecosystem around it is.
The sun is usually quite dark at night
Unless nuclear gets dramatically cheaper one should expect to see basically just the minimum amount of it to maintain grid stability / heavy industry, or something in that direction (I think). Obviously this is quite path dependant.

The one thing I really wish people would take away from thinking about energy / grid composition and so on is that its a portfolio construction problem.

If you ever, ever, see anyone banging on about how expensive or cheap a particularly type of generation is please just write them off as either an idiot or a liar - what matters is the cost of constructing a portfolio of electricity production that meets the requirements of a nation e.g. running an industrial economy solely off solar power is probably impossible but it seems realistic to imagine a "modern" (politics aside) economy that runs off sending emails on a mixture of solar power and storage.

Maybe. Volatility I think, ceteris paribus, forces the clearing price up and wind is extremely volatile so there is only so much you can get with having more of it.
If it makes no sense economically there's a good chance (in general) that it was an intelligence agency of a superpower
I'm seeing a lot of totally unevidenced slop that [guy I dislike]'s support has collapsed after Ukraine blew up a datacentre.

In general most accusations are confessions, pay attention to people who use words like disinformation and misinformation. They want to control other people's speech.

We have expensive, potentially unreliable (if an interconnect fails we are completely fucked so certainly fragile), energy.

Energy is what allows modernity to function, we have made it so expensive that no one can make anything in Britain just in time for hard power and manufacturing to be really important again.

Chemicals, metals, pottery, glass,fertiliser etc are all basically impossible to do competitively in Britain.

We may genuinely have the worst energy market in the developed world. It's going to get cheaper and reliable any day soon - George monbiot told me so it must be true.
No it's a catastrophic failure of planning and institutional nerve.
Balancing auctions are pay-as-bid, the alternative is lights-off.

Capacity auctions are iirc pay as clear. Maybe day ahead as well.

One thing to consider as well is that without some external stimulus scheme all of the fossil/nuclear (even ignoring CFD schemes) would immediately go bust IIRC if the market was solely pay-as-bid because wind and solar are cheap but extremely spotty so the price would either be 0 or 1000, probably not the latter long enough to keep the gas plants we currently rely on in business => no more power on cold calm nights.

These guys are all going to end up in prison at this rate. Dems are currently predicted to win next election no?
They're picked (mostly) using very mechanical rules though so e.g. hedge funds actively speculate on which stocks will be added to the index
No, the investment thesis is that there is a risk premia associated with investing in equity (as opposed to cash) that you will be compensated for - if this risk premia exists the way to harvest it is to be as diversified as possible (hand waving).

The relative performance of different baskets of equity is of course much more subtle / prone to behavioural effects and so on.

Sure and there are billion dollar systems at every bank that runs entirely on VBA macros in excel doesn't mean its not structurally flawed.
Re:Release of Polars 2.0 on Tuesday October 06, 2026 @11:29AM (#49979988)
Avoiding pandas developers is a great reason to use polars imo
Re:Gitframes on Monday October 05, 2026 @11:04AM (#49965807)
"But I didn't write it" is usually the defense. Just following the sat nav honest officer!
They have people in sim almost all the time / as much as allowed during the weekend.
Because the marketing teams don't do numbers and it's a high risk strategy
Re:Show HN: Germany's new sovereign AI model Kolibri on Saturday October 03, 2026 @10:42AM (#49944718)
"It's a German joke, it doesn't have to be funny"
Re:Show HN: Germany's new sovereign AI model Kolibri on Saturday October 03, 2026 @10:41AM (#49944716)
Stereotype accuracy is one of the few social science results that replicates.

Older »