News for hackers. Stuff that matters.

up2isomorphism

The number seems too optimistic compared to my observation.
Such a waste.
Re:Remove pgrust: the entry is cheating on Friday October 23, 2026 @04:43PM (#50215050)
Your explanation makes the removal very reasonable. Also pointing out to a version after this is done surely does not help.
It is not contradictory at all. If your neighbor has a gun you probably buy a gun too in fact probably a bigger one.
Training is never a problem. The feedback is the essential requirement of human learning process, and AI cut it off.
Someone argue this is fine since there used to be no OPT long time ago and US is doing fine. But the key is here is NOT if there is OPT or H1B but rather about that 100K cost. It is much better to cancel the OPT/H1B than charge this fee. Because for the former you do not lower the bar, but the latter doesn’t stop the people coming through the program but you only get low tier of talent, which is the least you want in the competitive world. Also it does not reduce spying but rather make it easier.
Re:Homa: The End of TCP for AI Clusters [video] on Monday October 05, 2026 @10:50AM (#49965634)
Actually AI is a place where TCP works mostly fine, due to its bandwidth limited nature rather latency limited.
Re:Apple iPhone 18 Pro Camera test on Monday September 21, 2026 @01:06AM (#49783261)
The really problem of iPhone camera is not these benchmark things. The problem is that the photo captured by the “Modern iPhone” simply does not look nice , let say, compared to iPhone 4S.
Re:Rust Is Tier-1 Language at Microsoft on Thursday September 10, 2026 @06:58PM (#49651275)
So do we observe overall Microsoft software quality improvement so far?
Re:Don't use musl if you care about performance on Friday August 28, 2026 @01:57PM (#49482188)
Funny thing is that the major reason most people use musl is because glibc make it (artificially) hard to do completely static linking.
Re:AWS Acquires DuckDB on Wednesday August 26, 2026 @01:58PM (#49453235)
It is not misleading at all. For all these softwares, the team that behind is the most important part rather than the software itself.
Re:Zig's Io.Threaded Is Neat on Saturday August 22, 2026 @12:57PM (#49401543)
Just use C, these are well understood things 20 years ago.
The deep problem is that insurance is a highly profit driven business in the US which basically go against what a good insurance should look like.
Re:Investing in automated C to Rust translation on Wednesday August 19, 2026 @10:54AM (#49362487)
Ubuntu has not been famous for robustness as a distribution. It seems the tastes is a major factor as seen in this case.
Re:PostgreSQL for Everything on Wednesday August 19, 2026 @10:48AM (#49362400)
Hardware is so far nowadays to make people with little systems knowledge confident to make such claims at least from their use cases. However it is neither generally reasonable nor efficient.
300x if it is true you will be just busy dealing with you customers rather than pitching here. Also since it is a vibe coded project, if you are really that good , you should even need to related yourself with Postgres. Who will want related itself to something that is 300x slower than itself?
Coming from a country where almost everybody rides a bike. I would say US is the worst place to ride a bike. Also the utilization is inherently low due the city layout and weather is bad in winter in most US cities. It would be much more efficient to use the space for bike lanes for better pubic transport.
Re:How is the Bun Rewrite in Rust going? on Monday July 27, 2026 @05:23PM (#49075705)
If actually anyone noticed, then it is a pretty bad move. Also “going well overall “ does not answer the author’s observations either.
Re:Cruller: Bun's Zig Runtime, Continued on Zig 0.16 on Thursday July 23, 2026 @06:15PM (#49028784)
The statement is weird, of course it is aiming at replacing bun, and it might not be a bad thing. Not sure what that clarification serves.
1.5B is a joke
The results are great. They will still need to post the detailed chat session to show how AI is helping though.
This is not really true considering Apple and nvidia are two most successful hardware companies, and they are notoriously closed. Not to mention microsoft, oracle they are all pretty closed.
Re:How Our Rust-to-Zig Rewrite Is Going on Thursday July 16, 2026 @11:29AM (#48935904)
I think there will be soon a wave of rewriting rust to language X coming up.
This is a very wasteful way of getting communications to somewhat compensate the lack of competition in US telco market.
You are exactly depicting the problem of the rust culture.
300x is mostly a marketing term, especially without the test description.

BTW, showing no respect to what it is trying to copy looks uncomfortable.

It is very bad metric to measure a rewrite using the original tests. It is like to say you clone a cat by mimic 600 cat images.
Re:Python 3.14 compiled to metal – no interpreter on Monday July 06, 2026 @05:57PM (#48811026)
Then the question is why? Because that is an another way of saying donating tokens.
Who told you US telco sector is a free market?
It depends on how inequalities is defined. But from the wealth distribution point of view, today’s wealth distribution skews much more towards the top. Although the lowest living standards improve thanks to the technology advancement.

Older »