Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
BenoitP
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
38 ms
·
391.
▲
by
BenoitP
5y ago
Here is the paper detailing the preparations, which were about mimicking an ITER-like wall: https://iopscience.iop.org/article/10.1088/1741-4326/ab2276/...
392.
▲
by
BenoitP
5y ago
From the last comment, on 2021-11-04: > Like it or hate it, there is a sea change happening in how governments treat cryptocurrency.
393.
▲
by
BenoitP
5y ago
> accelerate the construction of renewable energy Accelerate what? Wind and solar are at capacity, and storage solutions do not exist at that scale.
394.
▲
by
BenoitP
5y ago
Otherwise known as the Shapley value [1], which also takes into account the interaction between participants. > The Shapley value is the average marginal contribution of a feature value across all possible coalitions. [2] EDIT: There are
395.
▲
by
BenoitP
5y ago
Why are group chats more exposed?
396.
▲
by
BenoitP
5y ago
It actually really runs slower as is it first an interpreter, and the compiler is yet to be optimized to GCC levels; but is has a great compiler infrastructure that could eventually JIT to faster code (it being a runtime compilation, that h
397.
▲
by
BenoitP
5y ago
> Since they buy thousands of cars they get the lowest prices possible from manufactures They also don't get much to choose in which models, and that's the actual point of the thing. The overall problem is fitting low elasticit
398.
▲
by
BenoitP
5y ago
Java can run C/C++: https://github.com/oracle/graal/tree/master/sulong What's your point?
399.
▲
by
BenoitP
5y ago
Java is not being rewritten to C though. It is being rewritten to assembly (by a C++ program). ---- Also, by the same argument you're using, git should be faster if written in assembly; since assembly can do more than C can. The tradeo
400.
▲
by
BenoitP
5y ago
> Germany is a net exporter of electricity Oh, it is. Right now it exports all its coal and gas and wood burning production to neighbors; while only using 7.5% (= 8.94/(63.2+56)) of its installed wind and solar: https://a
401.
▲
by
BenoitP
5y ago
Well, good on you I guess. What's so sad is wind and solar being so concentrated in Germany. If it were more spread out over Europe, it would have lower correlation; and help with production peaks. There's egotistic, and there
402.
▲
by
BenoitP
5y ago
More data, for right now: https://app.electricitymap.org/map Germany is not doing that well.
403.
▲
by
BenoitP
5y ago
> Short term, gas is great for that Still an unacceptable CO2 emission in my book. Natural gas is at 55,82 [kg CO2 / GJ], when coal is at 94,6 [kg CO2 / GJ]. [1] So that's only 40% less pollution, and there are still huge
404.
▲
by
BenoitP
5y ago
With the current technology, you can't transport electricity more than ~1000 km with acceptable losses. Atmospheric depressions are bigger than than, and the distance for solar are much bigger as well (other than cloudy, there's d
405.
▲
by
BenoitP
5y ago
1. There is very little energy storage 2. Sun and wind only produce a fraction of the time (40% [1] and 25% [2] respectively, and 7.5% (= 8.94/(63.2+56)) of installed capacity this very second [3]) 3. Production capacity without sun an
406.
▲
by
BenoitP
5y ago
Breeder reactors make that concern go away: https://en.wikipedia.org/wiki/Breeder_reactor
407.
▲
by
BenoitP
5y ago
> I'd imagine that somewhere there is always enough wind and sun in Europe. * wind: wind power production scales with the cube of the wind speed [1], so actual production only occurs about 25% of the time [2]. Also huge areas produc
408.
▲
by
BenoitP
5y ago
Germany's wind and solar is only possible because neighbouring countries compensate for when there's no wind and sun; which happen a lot of the time. On top of that wind and solar production must be bought first on the EU market,
409.
▲
by
BenoitP
5y ago
Well, Germany's wind was done without a thought about how to implement storage; it did not have to; it was a perfect egotistic game-theoretic play: The EU energy market mandates that wind and solar production must be bought before ever
410.
▲
by
BenoitP
5y ago
The most depressing thing is that it doesn't end there. It continues on in real life here on HN where it gets flagged.
411.
▲
by
BenoitP
5y ago
I don't see the J Extension on here. Does anyone know what's the state of work of that group? (J Extension is about dynamic languages acceleration; stuff like code caches, and maybe providing GCs some help. I guess that's new
412.
▲
On Parallelism and Concurrency
(inside.java)
4 points
by
BenoitP
5y ago
|
1 comments
413.
▲
by
BenoitP
5y ago
I wish Google would bury the war hatchet together with Oracle. Their J++ take with Android is really getting far behind and starting to split the community. Also, they've been doing very interesting things for the platform (a thread sa
414.
▲
by
BenoitP
5y ago
> Each Antminer takes up roughly one TSMC wafer. That's highly surprising to me. Have they gone the Cerebras way?
415.
▲
by
BenoitP
5y ago
At last! Now they need to admit full responsibility in their management pressing engineers to move the center of thrust, fundamentally changing the flight envelope.
416.
▲
by
BenoitP
5y ago
Yup. I guess I'll use Youtube less, then. The downvote ratio contains so much information as to what video is garbage. Less than 80% approval -> I don't watch. I use a Ratings Preview chrome extension to show me the downvotes n
417.
▲
by
BenoitP
5y ago
An addition I'd like to make is that this circle is a high dimensional sphere (>300 dimensions IMHO in our current age). Another one is that an alternate representation of this circle can be made by having theta as the x axis, and r
418.
▲
by
BenoitP
5y ago
> All the nuclear supply has already been sold and is being constantly used 100% of the time. This couldn't be further from the truth, at least for France: hover over the lower graphs to animate the source upper graph here: https:&
419.
▲
by
BenoitP
5y ago
Oh yeah. Definitely doing the decision-taking but not the decision-executing.
420.
▲
by
BenoitP
5y ago
RDBMS have statistics to try addressing that. IMHO just like machine learning is statistics on steroids, I would not be surprised to see some neural networks and embeddings try to model statistics for RDBMS in the future. These models could
More ›