Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
manfredo
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
28 ms
·
211.
▲
by
manfredo
6y ago
Perhaps a better question is, what are compiler authors optimizing for? A compiler that runs in 3 seconds that produces code that's half the speed as a compiler that runs in 3 minutes is a worse option than the latter in plenty of use
212.
▲
by
manfredo
6y ago
Right, and while this technology is set to start iterating, fission has been iterating for more than half a century. France generates the majority of it's power from fission since the 1980s, recently joined in this achievement by Bel
213.
▲
by
manfredo
6y ago
Hydrogen is probably a feasible power source for container ships. Its energy density by unit of mass is very good, though energy density by volume isn't as good as methane. The cost of hydrogen containment is a function of surface area
214.
▲
by
manfredo
6y ago
Sideloading amounts to downloading a .apk file and clicking yes on the prompt to install it. It's no more difficult than installing applications on Windows. I know firsthand that elementary school children are capable of sideloading ap
215.
▲
by
manfredo
6y ago
But the point is, you can just download a .apk file and install like like you would on a desktop. You can even download a .apk file that installs an alternative app store. You do have to click through a couple screens with scary warning mes
216.
▲
by
manfredo
6y ago
The problem is that synthetic methane requires a source of carbon. Extracting the fraction of a percent concentration of carbon in the atmosphere is not feasible. Hydrogen is one of the inputs for the sabatier process, so if the lower energ
217.
▲
by
manfredo
6y ago
You can side load apps. You don't need to go through Google Play.
218.
▲
by
manfredo
6y ago
Consoles are much, much narrower in use than smartphones. Most consoles have a few thousand titles, if that. By comparison the App Store had millions of applications. And as you point out, manufacturers are building these consoles at a loss
219.
▲
by
manfredo
6y ago
Statements that genuinely are an affront to human dignity are already prohibited under anti-harassment rules. Such positions are also rarely controversial: most people do not want to disrespect human dignity. A substantial segment of people
220.
▲
by
manfredo
6y ago
These anecdotes are backed up by empirical observations: https://nces.ed.gov/pubs2010/2010015/indicator5_19.asp https://nces.ed.gov/pubs2010/2010015/tables/table_19_1.asp https:&#x
221.
▲
by
manfredo
6y ago
If we're being pedantic, the whole point of the admissions process is to discriminate. The question is, discriminate on what basis? Undoubtedly they discriminate on the basis of grades and test scores, but also on the basis of things l
222.
▲
by
manfredo
6y ago
In order to determine whether a certain race is disfavored, you need to look at admissions data. The data for Harvard was staggering [1]. Non-ADLC Asian applicants in the 90th percentile were half as likely to be admitted than non-ADLC Blac
223.
▲
by
manfredo
6y ago
Stances that genuinely are an affront to human dignity are already prohibited under harassment rules. Banning politics in the workplace has no bearing on this. Other political issues may very well be important to people, and they're fr
224.
▲
by
manfredo
6y ago
This comment is a perfect example of why banning political speech in the workplace is a very good idea. Every workplace that I've worked in that tolerated political speech ended up with a highly vocal segment of people making statement
225.
▲
by
manfredo
6y ago
No, it's possible to support changes to the status quo while simultaneously maintaining an apolitical workplace. I most certainly support changes to the status quo, and I would simultaneously take the same stance as Coinbase with regar
226.
▲
by
manfredo
6y ago
Right that's why you send random bytes to other people. You're forcing everyone else to receive or store random bytes. Thus rendering any actual enforcement of a law forbidding possession of encrypted data (or random bytes) nigh-i
227.
▲
by
manfredo
6y ago
Then start sending random bytes of everyone. Random bytes are undistinguishable from encrypted data. If everyone is in possession of what appears to be encrypted data, then it's no longer reasonable cause for suspicion.
228.
▲
by
manfredo
6y ago
It's almost always a combinations of: * Searching algorithms * Sorting algorithms * Dynamic Programming (memoization) * Graph operations (traversal, best path, minimum spanning set) * Maps & Hashing. * Heaps or trees. E.g. finding
229.
▲
by
manfredo
6y ago
The way some companies work is that there's a set list of question archetypes, but interviewers come up with their own methods of presenting the problem. For example, a seam carving problem might get presented as finding a horizon in a
230.
▲
by
manfredo
6y ago
Wouldn't construction of an orbital observatory consist of putting together prefabs? Not to mention, modern construction often does consist of putting together prefabricated pieces so saying we have no experience with construction in z
231.
▲
by
manfredo
6y ago
> Meanwhile - AFAIK - we have zero experience in constructing anything in 0G, beyond assembling some large prefabs. Weren't the ISS and Mir constructed in zero G? Well, in free fall but that's still zero gravity in effect.
232.
▲
by
manfredo
6y ago
The ability to better visualize this info, is the value. Governments are willing to pay for Palantir because doing things tracking down pill mills is easier if you can perform operations like filtering for people that bought the same drug
233.
▲
by
manfredo
6y ago
I don't think it's correct to say that they sell surveillance services. When I think of "surveillance" I think it means data collection. Palantir doesn't build data collection tools, they build data visualization to
234.
▲
by
manfredo
6y ago
I've found that people who make this kind of statements often aren't familiar with the kind of work Palantir does. Palantir doesn't collect any information. They build tools for the government to browse information they alrea
235.
▲
by
manfredo
6y ago
I interned at Palantir some time ago. I had essentially no idea of what their software did until I had a demo at the onsite, and it wasn't really clear what the product was really for until I started working there. The product is simul
236.
▲
by
manfredo
6y ago
They tested for that: https://interestingengineering.com/crashed-jet-nuclear-react...
237.
▲
by
manfredo
6y ago
Because we need truly fantastic levels of storage to make renewables even remotely feasible. The US has 8 seconds of battery storage, and ~4 minutes of hydroelectric storage. We need 12 hours of storage to get to 80% renewable generation, a
238.
▲
by
manfredo
6y ago
It's less that we've lost a technological capacity, and more than we are no longer leveraging economies of scale. France and the US built a lot of nuclear power plants in the 1960s, 1970s, and 1980s (well, mostly just France in
239.
▲
by
manfredo
6y ago
Where are you getting this claim? Natural gas costs somewhere between $40 and $70 per MWh [1]. Storage costs for batteries range from $100/KWh (very optimistic) to $300/KWh [2]. But that's per kilowatt hours. That's $1
240.
▲
by
manfredo
6y ago
> Does that mean nuclear is just a totally great idea? No; the issue with nuclear is cost. Historically it has been quite expensive, but subsidised in opaque ways. In a zero carbon world, it might make sense for baseline generation even
More ›