Snild

I'm not sure I understand what you're saying here. Are you saying that screen readers don't know which language to interpret text as?

Just to offer a more benevolent interpretation: I read that "Google thought patents aggressive tools" sentence not necessarily as any implication of Google's intent, but rather an implication that Google were worried about patents being used as weapons against them.

Ooh, that's a fun one.

Another con: you'd better hope your input contains no duplicates. :)

> linear-time

But... there are two nested `range(n)` for loops. Typo?

I mean, people still prefer to not have to interact with the tax authorities, so maybe heaven is a bit of an overstatement. And I have no personal experience with the US system. But yeah, I sure wouldn't want to switch. :)

My understanding (based on no research at all) is that it was introduced here for a similar reason: "homeownership is good". And it probably is.

But I don't know if the deduction helps with that. People buy what they can afford, so the "discount" is probably consumed by higher interest rates and inflated prices. It can also be argued that it's a form of regressive taxation: the rich are eligible for bigger mortgages, and therefore have more to gain from the policy.

Personally, I would like to see it removed. Any serious attempt to do that it's probably political suicide, though, since it would have a noticeable impact on many people's finances. And who knows how it would affect the housing market. Maybe if it was done gradually over many years.

Nope. Here in Sweden, your employer reports your pay to the tax authorities, and deducts your tax from every paycheck. Banks report your capital gains and losses.

In March or April, everyone gets sent their pre-filled tax form for the previous year. You can change it as needed, for example adding income from a side hustle, or making deductions -- though those are fewer and less complicated than the US from what I understand. Many common deductions (e.g. for mortgage interest) will be pre-filled too.

Most people will not need to change anything, and so can approve the pre-filled form through text message or digital signature (or phone robot for old people iirc). If you do need to make changes, it can be (but doesn't have to be) done online without too much trouble. Don't get me wrong, it's not always trivial, but it seems significantly more pleasant than what I hear about the US system. And, again, the vast majority of people just have to send a text message some time in March or April.

Noise-cancelling headphones, Ben Prunty's "Faster Than Light" soundtrack, and an interesting problem.

And a desperate hope that I won't be interrupted this time.

The vault does no good if you're passwords can be identically recreated by the attacker. Which the "crappy and predictable seed value", in combination with account creation time data, makes trivial. That creation time may be explicitly shown by the service, or at least hinted at through something like the timestamp of the user's first post.

https://github.com/SonyMobile/py-hprof is a Python library that exposes a Java heap dump's objects as Python objects, allowing convenient interactive and/or scripted analysis.

IIRC, I started out with Android Studio's dump viewer, but I wanted to check a "nested" attribute (.x.y.z) for all instances of a class -- and there were several thousand instances. I quickly tired of expanding and scrolling the nested attribute lists.

I think I also tried MAT's OQL, but still missed the ability to add whatever logic I needed on-the-fly.

My "Show HN" post got no traction, and I don't think very many people have used it, but I'm still proud of it.

I'm guessing they're referring to Linus' ranting on Intel here: https://lore.kernel.org/lkml/CA+55aFwOkH8RH12Dzs=hT3e7eS3Ckz...

> The whole IBRS_ALL feature to me very clearly says "Intel is not serious about this, we'll have a ugly hack that will be so expensive that we don't want to enable it by default, because that would look bad in benchmarks".

(I'm definitely not an expert. Grains of salt encouraged.)

There were no checksums, so incorrect ciphertext is possible.

But Enigma is a stream cipher, so if there was a missing character or someone bumped the wheel halfway through, it should still be possible to find two separate solutions that produce the plaintext before and after the error, respectively.

I don't know.

They can't allow obvious fraud, or there'd be money to gain in spinning up lots of VMs to inflate your stream count (and thereby your share of the total pie).

But people must be trying, I assume. Some will fall more or less clearly on the fraudulent side, and some are probably technically ok, just a little annoying or deceptive. Like duplicating popular playlists and adding your song to it: https://celebrityaccess.com/2020/01/08/the-dark-arts-of-spot...

That kind of suggests an interesting tradeoff: do I want the bandcamp money from more dedicated fans, or the convenience and discoverability of streaming services?

Admittedly, I may be making a false dichotomy there.

I've chosen the path that I hoped would make as many people as possible able to listen to my music. That includes not only Spotify, but also YouTube "videos" and free downloads from my website. The Spotify income covers my own Spotify subscription, which is nice, but I didn't make the music for money's sake. :)

And I realize the above may come off as judgemental, but that's not my intent. I hope it can be understood as an observation and an anecdote, not an argument about what others should do.

That's what I expected as well, before I tried. I assume Spotify just prefers to make a few big deals instead of lots of small deals. It saves them from having to manage all the individual payments, and probably the whole "is this really your music?" thing.

So the aggregator's fee is simply an entry fee. As you can see from my numbers above, it's not a lot of money. :)

To be fair, the aggregator I use offers a choice between 15% of the revenue, or a fixed yearly cost of $9.99 per release (first year may be slightly more expensive, depending on release size). Since I used to release a trickle of singles as I finished them, the per-release pricing definitely made no sense for me; the 15% thing was the best deal I could find when I started uploading in 2009-2010. Also, back then (not sure about now), some aggregators' terms actually demanded exclusive digital distribution rights, which was a complete non-starter for me: I wanted to put stuff on YouTube and free downloads on my website too.

Oh, and I assume that the big labels have better deals with Spotify than my small aggregator does -- earning a better per-stream rate.

> They priced an entire month of music at about £10, the same as a CD. That's 720 hours (30 days)

Yeah, but who listens for 24 hours a day, all month?

Also, that £10 CD can be played infinitely. That's infinity hours of music, which comes out to around 0p per hour. That's completely unreasonable! ;)

The convenience of Spotify converted me from piracy, getting my money back into the recorded music game. I don't think I'm alone in that. That 29% cut quoted in the article doesn't seem unreasonable to me, especially when some of that is "new" money. What were the margins of old record shops?

That was the main thing I wanted to discuss, but let me also question some assumptions you seem to make.

> the labels have less control to "dictate" or push specific artists.

Why is that a problem? The curation and popularity lists are not gone, btw: the streaming services have taken over that role.

> The availability of everyone to release their own music has diluted the market so now there is so much volume, nobody gets heard

On the flip side, more people get the opportunity to be a little heard. Seems like a win to me. Also, there are definitely still those who are heard more than others.

Every time I hear this kind of complaint, I wonder what kinds of horrible deals they've made with middlemen.

I have some music available for streaming. In November, I had a total of 5126 streams, which gave me 14.611 USD.

If we extrapolate that to 1 million streams, it would be $2850, but Gary is saying he's getting less than 2% of that.

The aggregator I'm using (my only middleman) takes 15%, already subtracted from the numbers above. Gary's middlemen apparently take over 98%. That sucks, but it doesn't appear to be the fault of the streaming services.

One caveat to note about these numbers is that the value of streams seem to differ a lot between countries. It's possible that Gary has more listeners than I do in less-lucrative countries. I don't think it could cover the whole difference. The most likely explanation is that Gary's record company is taking most of the money, and/or he has to split it with others who were involved in the creative process.

If anyone is interested in my earnings data (further analysis, charts, peer review, whatever), it's available here: http://dolkow.se/files/streams_nov20.xlsx

The xor swap truck is rarely better. It causes pipeline stalls, and so is likely to be significantly slower than the trivial swap. The temporary storage will be in a register, so it's (usually) not causing memory accesses.

Based on no knowledge at all about Telemetry, probably something similar to the amazing trace functionality in the Linux kernel.

That is: non-shared (per-cpu in Linux) ring buffers, small simple events, and formatting done after the fact.

Not much, IMHO (as someone who enjoys writing C). It would still be essentially the same language, and the popularity of other languages -- whether they be C++, rust, python, Java, haskell or whatever -- come from introducing more interesting concepts than that.

...meaning they end up pretty small: my debt was about 250 000 SEK -- very roughly $25k -- after 5 years of university. Combined with the monthly stipend that everyone gets, I was "self-supporting"; my parents didn't have to (and wouldn't have been able to) pay any money out of pocket for my education or living expenses during that time. And now I pay around $1k total on that loan every year. I could pay it off faster, but the terms are so good that I'd rather use it to lower my mortgage, or invest it.

Thanks to this system, I got the opportunity to do what I wanted, and society gets another well-paid software engineer that can pay lots of taxes. Win-win!

(Edit: corrected debt from 300 000 SEK to 250 000 after checking)

I loved Hellgate: London. The game definitely lacked polish (ran out of money, I assume), but the general gameplay feel was great.

In Swedish, it can mean both "poison" (noun) and "married" (adjective).

I believe the quote/screenshot is from Linux Foundation. But yes, confusing. And no actual argument or discussion of the problem.

I honestly don't see what the harm is supposed to be. What is the purported effect of a broader attribution in the copyright header?

No, but it's fun when it does. :)

I find that manual assembly is mostly useful when I know something about the data or desired operation that is not possible to communicate to the compiler.

> who want a platform completely unburdened with apple or google

I don't think KaiOS will be. It seems that Google is one of the main investors in KaiOS Technologies[0][1].

[0]: https://www.kaiostech.com/google-leads-seriesa-investment-ro... [1]: https://www.crunchbase.com/organization/kai-technologies#sec...