tomas789

Most of them are. Including the biggest ones.

Imagine somebody being able to track almost anyone using road side cameras. Good thing such things happen only in eastern Europe wink wink

Having an objective score is quite difficult. Maybe it would be better to do a pairwise comparison and calculate ELO?

I don’t think this is as simple as it seems. For example, we have our own CA and issue several mTLS certificates, with hundreds of them currently in use across our machines. We need to check every single one (which we don’t do yet) because there is an additional distribution step that might fail selectively. And that’s not even touching on expiring CAs, which is a total nightmare.

The batteries in MWh range cost around 160 EUR/kWh all in. Including grid connection and BOP.

I like the Rust being there. It is a reality check to keep the numbers in perspective. (C, C++ or something like that would work too)

My guess would be they are running some dummy app like quote of the day or something and it records the screen at 1fps or so.

OP should consider changing their alias.

I just thought you should know.

Yes. It is common on leetcode.

I used to write actively on LinkedIn. Nothing big but still something. I couldn’t align with myself with the original post but I couldn’t tell why. This is aligned with my understanding of how LinkedIn works much more. Thank you for explaining that to me.

I love your setup. The ingenuity of such solution is outstanding. I have so many questions but I'm afraid to ask.

What surprises me the most is that there is no benchmarks table right at the top. Maybe the improvements are not to call home about?

Yes. I’d rather read the prompt than this.

They are almost the same batteries. Different form factor but same thing. They are rolled from the same lines.

Python 3.13 has finally introduced free threading, which is a big deal. I initially thought we wouldn't be able to use this in production for ages since packages would need time to adapt. Surprisingly, it looks like it might be ready for real-world use much sooner than I expected.

I don’t know where the 800W is coming from. It is quite common to have 16A breakers on a circuit. That is 3,8kW. My regular apartment has 3x20A breakers and for example my induction hob is 7,2kW two phase.

I'm no electrical engineer but would that really be so much cheaper? A buck converted is essentially the same thing where the transformer is replaced by an inductor. You get rid of a bit of copper but all the power electronics to do the high-frequency switching is still there. Together with most of the components.

They "estimate" this would save 60% on galvanic link, and on the rest, you can save another half. So that would mean you are at 20% cost plus some ground wire fault detection HW.

I wouldn’t want to relay on random caffee’s router to do this for me. I would still end up running firewall more carefully on my end devices. Which for my iPhone I’m not even sure if I can. So probably a personal VPN would be a must.

This assumes that we reverse the reaction. That does not need to be the case. We can put it back into the ground. It can either stay where natgas or crude used to be or we can basically turn that into rock (don’t know details but there is a process to do that). Or it can be put in salt caverns.

Either way we still need to capture the CO and CO2 from flue gas which is costly even from point sources and way worse from DAC. Unless we do oxyburning or fosil fuels which is a hack but you need a source of cheap oxygen which we don’t have unless we dramatically scale up electrolysis.

This is certainly a step in right direction especially with proliferation of AI based assistants there will be a greater need to have readily available information about the codebase. This could easily take those copilots yet another level up.

For example my workflow now with Cursor is to keep relevant code in spearate tabs even though I don’t work on the files. I found it makes the autocomplete better as at seems to me that all the active tabs are fed to the model. That means less space for me and more distraction. Glean might here.

In my, arguably limited experience compared to some other folks, completely agree. It feels lite the correct solution here is to treat it as another kind of backend. I always keep the truth-keeper backend separate and with much higher quality requirements.

I don’t think that is the case. Surface roughness and directional changes are big losses. For example a single valve on a high pressure line can have a bigger pressure drop than 2 bars. So 160 km of a pipeline will drop a lot more.

I add my +1 to this. I often come across this blog posts while working as a OR professional.

Claiming that the hardware is capable of running Apple Intelligence just because it can show up a settings and missleading at best and a lie at worst.

Once we know it can be done, there is much higher chance somebody will look really hard into it and with a bit of luct, they will come up with a practical algorithm. Or at least good one for some edge cases.