You can argue either way about the demolitions, but thats not why the article is on HN. It's trying to play off the nefarious and scary, "Weaponises AI" when in reality this is something that could have been automated a decade ago.
I wonder what it would take. For things like Java, JSA, and c# there would have to be things like a parallel utf8 api (yes please), but the really hard parts is the things that are as old as time (windows).
I don't think it will ever happen, but one can dream.
But I had once expected ReiserFS to be permanent, especially since this Hans Raiser dude who was driving the ship seemed to be sharp AF, and this seemed doubly-true when his filesystem got mainlined.
But it was not permanent. It did not last forever.
This idea of permanence, or rather the lack of it, was the whole of the point that I was responding to and also trying to impress upon.
At the end of the day: We do not know the future. Things can change.
I've outlived this one high-performance Linux filesystem in my life that I was using. This does not in any way mean that I will be the last to outlive other Linux filesystems.
What I wish more languages did was what the guile optimizer does. There is a source->source optimizer which does inlining, DCE, CSE and partial evaluation.
That is very handy, especially when writing macros. I only have to look at assembly when I want to know about optimizations that are not visible in the source->source optimizer.
If I ever want to know if something is reified (which I never do) I can always look at the ASM.
I'm not sure if this will surface as an "enabled by default" feature in .NET 11 - indeed there may not be a final decision on that yet. But it is an active area of work that will arrive sooner or later.
`AsyncMethodBuilder` is a way to declare async methods that return your own task types. AFAIK it is not supported with the current runtime async, and maybe never will be - it's quite hard to support, I believe.
Taking bike.jpg and bike-repackaged.jxl in that demo, on my Pixel 10 Pro in Firefox Nightly, the size and decoding times are:
bike.jpg - 147 kB - 10ms
bike-repackaged.jxl - 126 kB - 33ms
So, I save 21 kB, but pay 23ms.
That extra 21 kB isn't part of a new request, so when it comes to download performance, we're talking about an active response. At 3g speeds, in 23 ms you can download somewhere between 20-118 kB, so at best, it evens out.
If it's a lower end device, the decode delta increases. If it's a faster connection, the 21 kB saving matters less. If the image was cached, then I'm just paying the decode overhead.
People really, really like to try to save up on the matches, so to speak. Sometimes literally: there was a short period of time during Khrushchev's rule in the USSR when the natural gas was made free for the populace. Well, the matches still cost money (1 kopeck for a box of 50 matches), so lots of people would just not turn the stove off, instead always having it burning the smallest possible flame when not in use.
All that to say, "I save 21 kB, but pay 23ms"? You'll save more if you stop sending your favourite web font over with your page and use the system fonts, and trim down your site's JS.
You are probably the guy to ask. I have always found dns over TLS to be the fastest, but with the quic versions making an entrance, maybe things have changed. Which is the one that gives me the fastest replies?
We actually quite like DOQ and DOH3, both of which use QUIC.
All modern Apple systems use QUIC. For instance if given 9.9.9.9 via DHCP they will automatically upgrade to encryption via DDR, then then upgrade to DOH3 with no intervention by the end user. We find that encryption load is lower with QUIC-based transports, but we need to really get a formal research paper together on that "in our spare time."
Faster? Probably, but is it measurably "better" remains a question for others to answer. Due to parallelism in most query sets, minor wins with DNS latency matter less than you might think.
Not OP but I don’t think DoQ is very interesting. DoH is nice because it blends in with HTTPS. Both do the protocol level things to avoid connection handshakes. The HTTP overhead isn’t that much really.
DoT is more complex than you’d imagine because it has to try hard to implement its own solutions to avoid handshakes.
In some ways I feel that caring a lot about fonts is on the opposite side of the spectrum to whatever it is making a population feel their lives have more meaning during wartime or crisis.
As a visual designer, sure, it is you job. As a computer user whose ultimate goal of using a computer is something else... I am kind of an old fart, but I often find that the best content is often found in conjunction with "font-family: sans-serif".
I'm a mechanical engineering PhD who works on numerical computing, data analysis and the like. I've spent several hundreds of dollars on fonts. Because I like things being just so, because I like seeing the artefacts I produce being nice to look at, and because they're just plain beautiful.
Day to day, I use open source fonts that Butterick recommends. But once in a while, I like using just nice stuff. It's like a nice fountain pen for others.
I believe everything you say. I hope you didn't feel attacked. It is just a feeling I often get when I'm online. Some people seem to spend more time on form than actual content, and while I agree that how things look is a part of the expression, I am somewhat of a brutalist because I feel that it rarely matters much, and most of the time it is a distraction.
but again, I am an old fart who is sad gopher lost.
reply