mNovak

>> There's no attempt to create a single, wide-ranging social media platform, with payments, video, shorts, DMs, and more

Umm.. Facebook? Just offering all those features doesn't mean much though, you really have to be dominant in each one, since there's much less moat to debundling consumer apps vs enterprise

Interesting, I guess I've never seen anyone upset about xkcd before.

I'm not sure if "author" here refers to the HN comment author or Munroe, but the latter is hard for me to view as "knowing better than to ever create anything", given authoring 1000+ webcomics and all.

I'm really curious to know how he selects the topic of each article; the archive has quite a range!

This is the type of blog I aspire to make, one day

As an RF engineer, this is a very creative application of mmW and I hope can be successful, but is not without very significant challenges which the company has made no claims to have solved. The issue is, mm-waves (and all high frequency RF really) are very sensitive to distance and losses. As a random example [1] gives some numbers for theoretical and precision machined waveguides in these frequencies. In the very best case we're talking about 0.03dB/cm loss. That means in the first meter of ideal waveguide you've lost half the power! And not just disappeared, you just dumped 0.5MW into the pipe itself, meaning it would likely melt within minutes. Meanwhile, their concept is to bore 20km into the earth.

The only way I could imagine this working is to lower the gyrotron itself down the borehole, but there's presumably tradeoffs of physical dimensions and output power that make it challenging. Again maybe not impossible, but there's no indication they've made fundamental advances in miniaturizing the source, or producing lossless waveguide.

Related, another interesting company I came across is Petra [2], who is doing something similar -- plasma/heat based non-contact drilling. But because they fracture the rock instead of vaporizing it, there's essentially gravel/sand leftover which is hard to evacuate from a deep vertical shaft, hence their focus on power/fiber trenches not geothermal.

[1] https://www.nrao.edu/meetings/isstt/papers/2012/2012151153.p...

[2] https://petra.cc/

Is concrete actually a good building material for homes? It doesn't seem like it'd be insulating, and you're definitely not getting internal plumbing/wiring. And concrete is actively bad for the environment. It all feels like the shipping container home -- seems great at a glance, but doesn't really pan out to be lived in.

Ikea style flatpacked prefabs still seems like a vastly better option to me.

>> you would think people in his position would know what private keys are and how to secure Bitcoin funds they pretend to "seize"

How exactly would you have handled this differently? They were well aware the account could be recovered via mnemonic, but they couldn't move the funds because they didn't have the PIN.

There's a certain irony to me that this discussion of cool html elements is all embedded as images

Ecommerce filtering for electronics generally is terrible. I even made a hobby website for used electronics, specifically inspired by the precise parametric search we expect from McMaster and Digikey etc

Yes, but buying and 'delivering' the gift are two separate actions. So I can buy $30k in bonds for you today (which benefit from the current rates), held in the gift box until I transfer them to you. But as you said, I can only transfer $10k/year less any purchases you make that same year.

In the 1 year hold case (buying direct + gifts for 1 year only) you can transfer and immediately sell the gifts alongside the original bonds, in the new year.

It works for me in any case, I've done it several times. Not sure what's different

I find it easiest right-click>inspect the password field in chrome, and paste my password directly into the html value field.

Another thing to note, if you're particularly trying to load up, is you can buy I-bonds as a 'gift', separate of your individual limit. Specifically you and your spouse can buy $10k each directly, plus each buy $10k as a gift to the other (thus $40k total). You can actually buy >$10k as a gift for someone, but you can't transfer beyond $10k a year to that person, effectively increasing the minimum hold period.

Also a general note - you can enter your payment account electronically, but ever changing it requires mailing a form! So choose wisely.

> This is solved off-cloud > You already do, if you store it off-cloud > This is solved off-cloud

Yes, but I think the goal is to have these things AND be on-cloud

All I want is an AI to give me beautiful color palettes. For some reason that is the absolute hardest part for me to get right.

I would recommend getting comfortable with some basic double-entry accounting software like GNUcash [1]. I found that extremely helpful in learning how to reason about these things, as well as grasp the purpose of different types of accounts (asset vs A/R vs income etc). They also have pretty good docs and tutorials as I recall.

Also then it'll quickly show you where the existing software falls flat (and it does) in terms of scriptability, scaling, etc.

[1] https://www.gnucash.org/

This was a great and informative article, but one thing it misses is that there is in fact another way to purchase a 'legitimate' peerage title (though the cost is much much higher). I recall reading about it a while back here [1]. Basically, you pay a (broke) noble family to legally adopt you. Apparently there's brokers and everything to get you connected.

[1] https://www.bloomberg.com/news/articles/2020-02-02/even-if-y...

Now we just need a slick UI for the full workflow: DALLE concept, DALLE scene extensions, and AI up-sampler

It's just a bunch of regex, nothing too sophisticated. Generally the people making listings are trying to convey all the data, so in a sense it's semi-structured. The most limiting factor is getting a CPU model; I'd like to eventually build up a taxonomy of major device brands/models to map to their specs (I do have this currently for Apple, which many people post year or A-model numbers for).

I like to buy used electronics online, but it's very tedious cross-referencing model numbers and benchmarks and such to determine what's actually a good deal. For example which is better, a 10th gen i5 or a 7th gen i7 at the same price?

So as a pandemic hobby, I made a website to scrape eBay laptop listings, extract all the relevant data, cross reference against benchmarks, and sort the result by "value" (aggregate performance / cost). I took inspiration from the parametric search on site like digikey for commodity components, where no one cares about brand.

You can check it out here [1]. Maybe eventually I'll write a showHN, but I haven't been working on it for a while, and there's a few features I want to add before that. Originally I planned to use affiliate links, which seemed like a nice all-parties-win setup, but eBay kicked me off the Partner Network program for not generating traffic. So presently it's totally free and non-profitable.

[1] https://bytebucket.co

Like a Zestimate?

I'm actually stunned just how effective that MARPAT pattern is compared to the NATO blob camo, in the second image.

Worth noting also that a lot of the more specialized roles that someone like Intel or the few other HW companies need is increasingly specialized, meaning MS and PhD. That makes it all the more frustrating to encounter these salary disparities others have described.

Part of what's happening here may also relate to the fact that EE PhDs are dominated by international students (70-80%), compared to ~10% in CS undergrad. That gives companies a lot of negotiating power to drag down the average salary.

That's a bit like saying SWEs just bang out fizzbuzz every day and poof, out pops the software.

So I think your complaint is really about the internal float storage truncating sig figs more than anything.

I agree that for all its "smart" pattern matching features, Excel should be able to figure out when to give you a better storage precision, and handle it fluidly in the background. Seems like we could assume one doesn't import a ton of extra digits for no reason.

I guess I'm confused what alternative you want, to display long numbers? Even more so than a sheet of paper, an excel doc has tiny windows to display a huge potential range of values, and something like 0.00000567 simply might not fit.

Seems like if it bugs you, just ctrl+a then set the number format, same way you might set the font and size at the start of a word doc.

Looks nice, but I'm confused how to read the Expected Returns y-axis. My assumption would be that's an annualized ROI, but it's displaying for instance that US Stocks return 0.7% ? Also no idea if Risk has any objective unit?

Running the 6 asset classic, it is slightly amusing that the result is essentially just a classic 70-30 stock/bond mix (risk dependent obviously), and the other 4 assets are <5% of portfolio.

Would be nice if you could include also representative ETFs of a given asset class (I assume they exist) to make the results more actionable.

Interesting, and congrats on launching!

On the one hand it sounds nice for me to be able to sell a side project and have finality. But also I've never had the patience to actually advertise/monetize them, so the valuation a pre-revenue service would get is too low for me to accept the sunk-cost fallacy. I suppose that's the small-project catch 22

At a glance this seems to show that bonds are in a historic drawdown. Can anyone ELI5 the implications of that?

It's because large, directive antennas are still 'dish' style and have to be mechanically pointed at the target (Earth based DSN receiving antennas). That pointing causes vibrations and potentially a shifting center of mass.

Phased arrays allow beam pointing without mechanical movement, but are very expensive for large high gain antennas.

Not sure about lichess specifically, but I know many game servers are a bit cagey about their cheat detection systems, since they don't want to give the cheaters any hints.

But as a broad concept, one technique is to compare moves against bot suggested moves. i.e. if a high % of your moves are those that a strong AI would make, something is fishy.

Related (-ish) is the free and open source Go server [1,2], OGS. A great community, and ad free. I've been playing there since the pandemic started; lately even some professional/tournament games were being hosted there.

[1] https://github.com/online-go [2] https://online-go.com