Hacker Newsnew | past | comments | ask | show | jobs | submit | cwillu's commentslogin

Highly publicized, but not especially common. I would be shocked if abuse of AI wasn't rampant in the legal industry; a few high profile cases will have as much impact as a few high-profile piracy cases did on piracy.

Bread is flour, doesn't mean I want a cup of flour included with every loaf.

The newer satellites have a laser interlink, which alleviates the regional groundstation requirement.

OOM triggers when memory pressure is critical; it's not practical to further defer addressing it at that point.

Huh? Defer addressing? You are even more confused than the person I responded to. The memory would be reclaimed in this scenario, it is entirely “practical” in that sense, there is no contradiction.

Defer /addressing/fixing/solving/ the problem.

If the oomkiller is running, it's too late to wait and see: some process or kernel thread needs memory that the kernel can't currently provide


No one said we were deferring fixing the problem. Memory is reclaimed. We would only be delaying or avoiding the part where the process is killed.

Hell, a cable is vulnerable to a cord not being screwed in all the way somewhere (anywhere) in the house

That sounds like a problem with a transformer in the neighbourhood; nothing about the 100A service implies that the voltage should drop enough to flicker lights when you get near that maximum load.

It's the neutral return of split phase that is the problem.

In a modern system the hot, neutral, and ground are all running in the same jacket (like Romex) or conduit. Topology is consistent for all wires in the system.

In a knob and tube system, each wire is individually wired inside the wall cavities. So it's often the case that a neutral wire might feed one set of outlets on one leg of the split, then go on to feed another set of outlets on another leg of the split whose hot wire runs a different route. In fact, that's kind of the whole point of K&T: keep the neutral and the hot away from each other as much as possible because insulation (at the time) was unreliable and shorted wires cause fires. And definitely keep the two hot legs even further away from each other if you can.

Even with perfect topology of a modern system, consider what happens when you are running a 10A appliance on one leg. 10A is flowing through the hot, and another 10A through the neutral.

Now what happens if we turn on 10A on the other leg of the split? 10A flows through hot A, 10 amps flows through hot B, and ZERO amps flow through the neutral because the loads are balanced.

If you turn on a device that has high reactance, like a vacuum cleaner motor has a lot of inrush current, then it can cause lights on the other 120V leg to get an unstable current while the electricity finds its way through the new load instead of the neutral, but then not so much current is needed after the inrush, so it has to find its way back to the neutral.

Now, imagine the neutral topology isn't consistent with the hot, as in K&T wiring. It exacerbates the problem.

Plus, no matter what you do let's say your maximum service ampacity is 200A and you have a balanced 240V load like a permanent A/C. The inrush current can cause the entire rest of the house to brown out for that moment as the incoming service simply cannot deliver that many amps without a significant voltage drop.

That said, it's the unbalanced 120V loads like refrigerators, portable A/Cs, and vacuum cleaners that really cause the flicker, especially in a K&T wired home.


> That sounds like a problem with a transformer in the neighbourhood

The house was built in the 1890s before electricity was even "invented": there are (disconnected) vestiges of knob-and-tube wiring, and gas lighting pipes before that. A few houses around here have coal shoots (I think mine may have one in a corner someplace too). So I'm sure there are all sorts of 'things' going on in the panel and wiring.

The transformer that feeds my house (and a half-dozen others) actually failed 1-2 years back, and was replaced. Thankfully during the spring/summer, and not winter.


The claim they're responding to is that the supply for an entire house is 3.3kW.

Presumably for the tank farms, where they have significant heat-transfer requirements to keep the liquid oxygen and methane cool to prevent boil-off.

Please quote where the OP said “guaranteed”; all I see is a claim that email is guaranteed to fail.

Email as the sole notification source for a safety critical situation is criminal, in the literal “the responsible engineer or coder masquerading as an engineer should be in jail” sense of the term.

More people need to get arrested for shitty engineering

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

Search: