HNHacker News
TopNewBestAskShowJobs

EraYaN

129 karma · joined March 7, 2024

submissionscomments
EraYaN··on Company claims 1k% price hike drove it from VMware to open source rival
And most importantly some clients will not communicate that they are unhappy and just move on quietly. Especially for something like virtualization which is essentially a commodity.
EraYaN··on Australia proposes ban on social media for those under 16
That is why you always regulate the provider/seller (ie. the social media site) and not the consumption side. We already learned this in the war on drugs, prosecution of consumption has never really worked all that well.
EraYaN··on Australia proposes ban on social media for those under 16
If you put the parent in jail for 6 months and fine the company 1 million per user account, they for sure will start thinking real hard about fixing the problem.
EraYaN··on Australia proposes ban on social media for those under 16
I mean Apple even added a native feature to iMessage with full on tracking, exactly because of safety. Sure someone needs to send it manually but the whole point of Check-in (tracking someones progress home) is safety. Just too many shady blokes out on the street at night.
EraYaN··on Australia proposes ban on social media for those under 16
They are not propaganda in the classic sense since that requires intent and a goal.
EraYaN··on Trump wins presidency for second time
The difference between what they are and what you should call them. Getting voted in asks for coddling your potential base.
EraYaN··on Why Companies Are Ditching the Cloud: The Rise of Cloud Repatriation
Depending on you size, even just having some people from all the open source products on staff is probably cheaper anyway. And gives you pretty good control. And if it used to work the only one that can have broken the config is you, which means you can also fix it. Sure maybe you need to rollback a kernel update or whatever but in the end it's on you.
EraYaN··on A comparison of Rust’s borrow checker to the one in C#
Quite a few C libraries are using it as their build system. Even in "production-ready" libraries.
EraYaN··on A comparison of Rust’s borrow checker to the one in C#
They don't even use Razor Pages but a custom RazorSlices package to do the templating [1]. Yes, that is much faster because it removes MVC and a ton of infrastructure but it's also kind of gross. Also the use of stuff like UnsafePreferInlineScheduling has some downsides (running application code on the IO thread) and honestly I'd never use in production.

The custom BufferWriter stuff is pretty neat though, although also not really something most people will reach for. And there is more, like the caching of StringBuilders etc.

But it also doesn't use the actual HTTP server to build headers, but they just dump a string into the socket [2], feels a bit unrealistic to me. In general the BenchmarkApplication class [3] is full of non-standard stuff that you'd normally let the framework handle.

[1] https://github.com/TechEmpower/FrameworkBenchmarks/blob/mast... [2] https://github.com/TechEmpower/FrameworkBenchmarks/blob/mast... [3] https://github.com/TechEmpower/FrameworkBenchmarks/blob/mast...

EraYaN··on A comparison of Rust’s borrow checker to the one in C#
Well the amount of tricks that basically nobody can implement in production that those implementations do to get the numbers that they do is ridiculous. Meaning the numbers they get are wildly optimistic at best and misleading at worst.

If you take the standard template for any of these frameworks (both Java and C# and any other language) and you add authentication etc, the real performance will be 5-10% of the numbers reported in those benchmarks. Have a look through some of the weirdness in the implementations it's wild (and sometimes educational). The .NET team especially has done stuff specifically to get faster on those benchmarks.

EraYaN··on Chain-of-thought can hurt performance on tasks where thinking makes humans worse
It kind of does though, because it means you can never trust the output to be correct. The error is a much bigger deal than it being correct in a specific case.
EraYaN··on A comparison of Rust’s borrow checker to the one in C#
Those are notoriously cheat-y for just about all languages on that list. Any actual project is never going to get close to the efficiency that those number would require. Both Java and .NET over promise by A LOT with the numbers they get there.
EraYaN··on OpenZFS deduplication is good now and you shouldn't use it
digiKam can dedupe on actual similarity (so different resizes and formats of the same image). But it does take some time to calculate all the hashes.
EraYaN··on Proposal: JavaScript Structs
At least in C# you can ignore most of it and the complexity doesn't really come from numerous foot guns. You can still write Java 1.6/.NET 2 style C# code just fine, it's all there. The rest of the features can be fully ignored and they won't hurt you.

But then again the newer features they do make writing code a lot nicer, giving more compile time analysis warnings etc hopefully resulting in slightly better code. And the new features also enabled a lot of performance improvements in the runtime which is nice. .NET 2/4 wasn't all that fast, .NET 8 can be a lot faster.

EraYaN··on The Age of Depopulation
At that point the older generations will already be wielding their political power with an iron fist to get the last drops of all the money and resources that do not get replenished. This was never an apocalypse type issue, but always purely political and the younger generations will suffer for it. THAT is the main issue, if you are old enough now you'll be fine, you'll get to "eat" the last little bit of stuff and then you pass away. It's the folks that have 60 years to go that are in trouble. Convincing the elderly to not vote or to take "their" stuff away is always going to be unpopular when 60% of the voting block in that age.
EraYaN··on Congress fights to keep AM radio in cars
Both of those things can be true, the smaller the bandwidth is you need the more extra information you can pack in your channel width to make it possible to reconstruct a lot of data errors. And modern digital codecs can squeeze data much further than plain analog audio, hence it's easier to get it to go long distances if the system were to have that as a design goal.
EraYaN··on Congress fights to keep AM radio in cars
Disturbing the supply chain for both hydrocarbons and electricity is actually scarily easy and in most places those are closely linked anyway. But it really takes a tiny push and the whole system comes crumbling down. Especially if you have adversaries that know what they are doing. There is very little "offline" infrastructure.
EraYaN··on Congress fights to keep AM radio in cars
But at the same time there are many many ways to make the stuff that makes that EV mobile. And you don't have to modify the car.
EraYaN··on Microsoft: 'ever present' AI assistants are coming
Most modern iPads do get written on them though, it's just not Windows based I guess. And iPad get more and more features to really write on them a lot.
EraYaN··on Employers Say Students Need AI Skills. What If Students Don't Want Them?
Mostly because most project ban you if you start auto generating PRs and patches or even issues. Most LLMs are just not good enough yet. We have seen this with curl and a bunch of other projects where LLMs were used to do security issue reporting and well the maintainers were not too happy talking to a bot that didn't know what the hell it was doing.

It's essentially a DoS attack on the maintainers.

EraYaN··on FFmpeg 7.1 release: a tons of codecs
For your Bluray collection HEVC is still advisable at reasonable bit rates. AV1 really only beats it in fairly low bitrate scenarios. Plus the HDR story is much better with HEVC.
EraYaN··on CouchDB 3.4.1 Released
We are using it internally as just a document store (with some erlang views but that is it). It scales very nicely, but for smaller documents postgres does the job for us so we are (slowly) moving away to simplify the tech stack.
EraYaN··on TSMC execs allegedly dismissed OpenAI CEO Sam Altman as 'podcasting bro'
TSMC has way to many competitors for this to make sense. And besides one very important key supplier who has huge wait lists for their EUV stuff, ASML.
EraYaN··on Microsoft's new Outlook client moves your email to the cloud (2023)
Thunderbird will support EWS, which if you use a cloud based Exchange account will be unavailable in 2026.
EraYaN··on Gaining access to anyones Arc browser without them even visiting a website
Firefox seems to be borrowing some of the UI features slowly (at least the vertical tabs). And at least the Mozilla Foundation is very public with their wants and goals.
EraYaN··on Gaining access to anyones Arc browser without them even visiting a website
At least Firefox seems to be borrowing some of the UI features slowly. At least the Mozilla Foundation is very public with their wants and goals.
EraYaN··on 0day Contest for End-of-Life Devices Announced
The idea being that the security updates would then also have to keep coming as long as copyright is held.
EraYaN··on Warning: DNS encryption in Little Snitch 6.1 may occasionally fail
But you are already on their OS, so they would always be able to do that. They make the kernel, the hardware and it's firmware, so it's a moot point and needless paranoia. Might as well use the API that gives a better user experience.
EraYaN··on A bottle of water per email: the hidden environmental costs of using AI chatbots
The biggest problem there is that that laptop is pretty darn efficient. And multiple GPUs and other servers running 100% to run that query for you are not quite as efficient. It does so much more math to get the LLM to do anything useful, the scale is really staggering compared to registering your keystrokes and updating the screen, even if that all happens in a bloated Javascript code base. You can run that laptop, which is not using more than 40 watts in spikes when just typing in a text editor for a long time compared to the full inference and processing load of that LLM. And that ignores the cost of training and getting the training data amortized over all queries.
EraYaN··on Warning: DNS encryption in Little Snitch 6.1 may occasionally fail
Not at all actually, passing hostnames means they can fully handle happy eyeballs for you and all other performance optimizations that you can do if you resolve and connect in one call.
← PreviousPage 2 of 3Next →