Hiring can remain irrational longer than you can remain unemployed.
One manager no-hires you because you don't post enough. Another doesn't like what you post. A third thinks you post too much. A fourth is pleased you seem to pay more attention to shipping products than hot takes. A fifth loves your hot takes.
So you get a call and are asked to do a coding thing. One person no-hires you because you wrote fizz-buzz by hand and didn't use Claude. Another wants to see that you know how to code by hand, but although your solution is fast, compact, and correct, it isn't the solution they had in mind.
At the end of the day, it's a highly inefficient, mostly irrational process dominated by social factors rather than objective feature detection.
> Turns out being a Jesus nerd was a secret requirement. Wish they could just put that in the job requirements.
Title 7 of the Civil Rights Act, in making religious hiring discrimination illegal, sometimes just drives it underground. Over the years it's done more good than harm, but at a certain point it may be time to let those who want to hire only Jesus nerds self-select.
Now, what we’ve been told about models is that they’re only as good as their training data. And so languages with gargantuan amounts of training data ought to fare best, right? Turns out that models kind of universally suck at Python and Javascript (comparatively). The top performing languages (independent of model) are C#, Racket, Kotlin, and standing at #1 is Elixir.
Generous of you to assume that someone who walks in, sees something somebody else has written and immediately calls it shit... Has something of value to say.
If they did, why did they hold it back just to speak so contemptuously of a subject that is actually interesting and reasonably well explained?
I think I see where he is coming from. Using math to prove that you can’t tune stuff, will to some, sound like using a laser leveling tool to prove that you can’t make a perfect pizza.
There is a quantum of earned generosity. Someone saying, "This doesn't seem right" has jumped to a conclusion, but they aren't getting personal about the author or the work.
Whether it's testes or testy language, getting personal and insulting does not meet my personal standard for assuming good intent and being worthy of an open-minded attempt to create constructive dialogue.
But I applaud you for wanting to lift the standard of discourse!
And the late Jaco Pastorius with the bass harmonics song that would have broken the Internet if we had had the internet when he released his first solo album:
Speaking as a person who owns basses... I like the sound of harmonics on a bass better. I think it's something to do with the longer strings giving more play to the overtones.
Last year my team had to remove all the phone numbers from 50,000 Amazon web page templates, since many of the numbers were no longer in service, and we also wanted to route all customer contacts through a single page.
Let's say you're on my team, and we have to identify the pages having probable U.S. phone numbers in them. To simplify the problem slightly, assume we have 50,000 HTML files in a Unix directory tree, under a directory called "/website". We have 2 days to get a list of file paths to the editorial staff. You need to give me a list of the .html files in this directory tree that appear to contain phone numbers in the following two formats: (xxx) xxx-xxxx and xxx-xxx-xxxx.
How would you solve this problem? Keep in mind our team is on a short (2-day) timeline.
In Yegge's case, he explicitly does NOT want a hand-written program, he wants the candidate to suggest a CLI tool, e.g.
These questions aren't good or bad unto themselves, but when the person asking is engaging in "Guess the answer I'm thinking of," don't beat yourself up if you guessed wrong. Your answer might be prized by someone else with an enormous amount of experience hiring engineers.
A former coworker was evaluating the cleanliness and structure of some non-free GeoIP data that took the form of several large CSV files. He was writing nested loops in Go that parsed the CSV and evaluated the predicates that interested him, and it was arduous and not going as quickly as he would like.
I told him to slurp it all into a sqlite database and to express his data integrity questions as SQL queries.
It was still a pain in the ass for him, but leveraging that tool made things go a lot better.
A few of the remaining newspapers I'm aware of run business awards (Best restaurant, etc), and the way to win is via wining and dining them, even though the paper claims it's based on people's votes.
Is that how it works where you are? Because over here, the best way to win an award from a publication is to advertise in that publication. Advertise enough, and you'll also become their go-to when they need a quote about anything vaguely related to your restaurant or other business, and once a year or so they'll print some hagiographic article about the amazing things going on under your leadership.
A very famous application of QuadTrees was Bill Gosper's HashLife algorithm for computing Conway's Game of Life. The Life universe is implemented as a quadtree, taking advantage of precomputed smaller squares to compute larger squares.
I have nothing to say about Quadratic Programming, so you tell me.
What I can say is that every reference I've found to Bill Gosper's algorithm describes the data structure as an immutable quadtree with canonicalized nodes, id est, there is extensive structure sharing in a Game of Life quadtree. That in turn facilitates heavy memoization.
I worked for GitHub for a time. There was a cultural abhorrence of the diaeresis, it was considered reader-hostile and elitist. I refused to coöperate with that edict internally, although I grant that every company has the right to micro-manage communications with the public.
It exists to indicate how a word is pronounced. Naïve is a better example IMO, cooperation feels too familiar.
Non-native speakers might see something like "nave" instead of "nigh-eve" unless it is clear that there is a stress that breaks out of the diphthong.
I don't think style guides are (usually) about absolute correctness, but relative correctness. A question is asked, a decision needs making, someone makes it, and now a team of individuals can speak with a consistent voice because there's a guideline to minimize variation.
IIRC it's use is to distinguish vowels that belong to separate syllables with vowels which form a diphthong. I think this could be beneficial to language learners, to give them a hint that cooperate is pronounced "ko ah puh rayt" instead of "ku puh rayt", and likewise naïve as "nah eev" than "nayv" or "nighv".
I encourage people to discriminate against me because I write like an educated African who works annotating AI training material.
Why not? I am a descendant of Africans. I am a mildly successful author by tech nerd standards. I was educated in the British Public School tradition, right down to taking Latin in high school and cheering on our Rugby* and Cricket teams.
If someone doesn't want to read my words or employ me because I must be AI, that's their problem. The truth is, they won't like what I have to say any more than they like the way I say it.
I have made my peace with this.
———
Speaking of Rugby, in 1973 another school's Rugby team played ours, and almost the entire school turned out to watch a celebrity on the other school's team.
His name was Andrew, and he is very much in the news today.
One manager no-hires you because you don't post enough. Another doesn't like what you post. A third thinks you post too much. A fourth is pleased you seem to pay more attention to shipping products than hot takes. A fifth loves your hot takes.
So you get a call and are asked to do a coding thing. One person no-hires you because you wrote fizz-buzz by hand and didn't use Claude. Another wants to see that you know how to code by hand, but although your solution is fast, compact, and correct, it isn't the solution they had in mind.
At the end of the day, it's a highly inefficient, mostly irrational process dominated by social factors rather than objective feature detection.
reply