HN user

nzzn

48 karma
Posts0
Comments16
View on HN
No posts found.

It is not at all surprising that “The Soul of a New Machine” resonates with people in Hacker News. And it is a tremendous book. But for me, his book about Paul Farmer, “Mountains Beyond Mountains”, was even more meaningful. Tracy met Paul by happenstance and in a short encounter recognized the stubborn greatness that was Paul. He was an amazing character who sadly is no longer with us, but captured in a book worthy of his life. Read the book and contribute to Paul’s life work “Partners in Health”.

Lua uses the table type to represent both dictionaries (hash tables) and arrays of values. This seems to have been predicated on keeping the language “simple” with a minimal number of defined types. A laudable goal.

However, arrays of a single type are just enormously common in applications. Support for arrays is pretty much ubiquitous in other languages, including ones that are in the same general dynamic space.

Internally Lua does treat arrays in their own pathway to keep performance reasonable. There is also some user facing special syntax for arrays. Arrays should be part of the core language — some learning overhead for the newcomer but worth it.

For a more full featured Xoshiro/Xoroshiro implementation see

https://github.com/nessan/xoshiro

Documented at: https://nessan.github.io/xoshiro/

Handles the full family of these generators featuring arbitrary jump sizes, stream partitioning for parallel applications, and, like this library, a number of convenience sampling methods to shield the casual user from the complexities of using <random>.

Can be used with a companion `bit` library (https://github.com/nessan/bit) that performs efficient linear algebra and polynomial reduction over GF(2) for those that want to explore some of the mathematics behind these linear generators.

Why are you doing numerical code in single precision? If no one in your group did a numerical analysis course at college it’s time you moved on to think about some other problem more suited to your experience set.

People are lazy.

American students are mostly rich by any global standard and very, very lazy. Doesn’t matter a huge amount as they don’t matter — they will amount to nothing anyway. The States imports enough talent to make up for the lacuna. In the meantime, their parents pump cash into what will become their Alma Maters.

When I went to an Ivy League US university as a grad student I was astonished at the remedial nature of undergraduate courses. Content that students in my country mastered at 13 needed to be spoon-fed to US students that were 5-6 years older.

Even back then almost nobody failed a course in the US. It was a major deal to fail someone. I came from a culture where the standard was absolute. No curve. Get below z% and you spent the Summer getting ready for a retake. Fail that, and you were out.

Education was paid by the State so it wasn’t a business. Profs could fail 20 - 40% of a class and often did.

It is astonishing that a Philosophy prof is seeing this. Who the fuck does philosophy at Uni and can’t be arsed to read the recommended texts?

He/she is a full Prof. Almost impossible to fire. So fail the lot of those entitled, lazy, bums I say. Enjoy that tenure!

The pricing of all these AI models is so opaque. Can’t they hire someone with common sense who realizes that $x/token is absurd!!!

Totally brain-dead. Like pricing a car by the number of o-rings in the engine.

Very in-house definition of “better”. Better on paper and at producing paper but who cares?

Typst is technically interesting but it doesn’t appear to have a driver (a person, a Zuck, a Bezos, a …) with a business/power-hungry brain at the top table. Forget paper/PDF, and for the love of God forget TeX/LaTeX. All — absolutely all, resources should be focused on web/HTML. Rank-order your priorities. Using all cores in a CPU to produce something for a printer is not even on page 33.

Focus please and make the web a friendlier place for exploring/expositing science/knowledge. It’s annoying because you tapped into something but seem to not understand the little ledge you have carved out. Wishing you a explosive success!

Quanta is such a great resource! High value spend from Jim Simons vast pool of dollars.

I came across MTW’s “Gravitation” as a student in the 70’s and it inspired a positively unreasonable desire to own a copy just because it looked so beautiful. Couldn’t afford the doorstop of a book at that time but happily it is still in print 50 years later.

VC’s, hedge funds, etc. all run lean—none have 10k+ employees or anything close. Employees are under a much brighter spotlight by the owners all the time. Groups might lose focus for an interval but once they their score is down it’s goodbye & the score is P&L calculated in very short intervals.

Buy a Ring doorbell? Just don’t. Ring requires a login AND enforces absolutely strict 2FA. You need to re-login periodically. It’s a doorbell! WTF. Doorbell rings, you are now fumbling to try and get a passcode—-delivery guy has given up. Absolutely horrible. It’s a doorbell. Get lost Ring