pnt12

Some software gurus really grind my gears, and Robert Martin is one of them. When confronted with bad advice he gave, he's quick to say it's not meant to be taken literally. Then, gurus like Kent Beck, say that you cannot criticize their approaches if you don't implement exactly as they say. So, while this is not exactly a paradox (different people with different opinions), I feel like gurus make their livings on unfalsiable claims while shaping the world of software engineering.

So, some kudos to Robert for accepting criticism and discussing it, but no cigar for downplaying his own advice when confronted - I also recall a different discussion, where someone confronts his statement "you don't practice tdd you're not a professional", and his answer "it was not meant to be taken seriously".

These people had great ideas but they should be more critical of themselves, eg "here's when not to apply this", "here's where to bend this", not "you're doing it wrong" or "don't take it literally".

The GPU main advantage is its parallelism - thousands of cores compared handful of cores in CPUs.

If you're training models with billions of parameters, you're still gonna need that.

Some counterpoints:

> when work is done, I leave, and it's done.

My experience from WFO is worse: task may be done earlier, but feels wrong to leave earlier, and colleagues may not take it badly.

>not using my resources (electricity / broadband / etc) for work

Gas is more expensive than electricity, so I'll take the tradeoff. Even better if you can reduce the number of cars in the family - purchase, annual fees and insurance, etc.

> easier interaction with colleagues

A pro and a con: good for collaboration, but also easier to be distracted when trying to focus.

Wow - that is the fastest loading site I've seen!

The story about pc stores is way off. First there was steam, and for many years it was the only store.

Many launchers came to be, with half assed community features. They didn't want a thriving marketplace, they wanted a piece of the pie. People hated them and they died.

Recently epic store came, with a billion dollar Corp behind, and they are fighting the long battle. The free games are a nice thing (usually older offers or curated indie stuff), but the exclusive business sucks. Luckily it seems unpopular, and fewer brands esnt to adopt it - some are so strong they can do it, but others found out that they can lose their loyal costumers and fade out, eg Borderlands.

Their vision, which I disagree with, of customer choice, is that its store is an alternative to Steam. Probably they justify their exclusives by not being a monopoly, which gets weird: steam has the monopoly numbers, but epic has the monopoly practices.

They got a few customers from their giveaways - they should stick to those and further improving their store, maybe some people will actually want to use it.

> If it's so good, why haven't all C libraries adopted the same tricks?

A man and a statiscian are walking down the street when they see a 50€ bill. The statistician keeps walking but the man stops and says "hey, look at this cash on the floor?". But the statistician, uninpressed, says: "must be fake, or someone would have already picked it up". And continues walking. The other man grabs the cash and takes it.

Can you elaborate on what's TOAST/DETOAST?

Of course he doesn't want the money, he's only getting money to shut down those conspiracy theories.

/s

Very disappointing article.

It's true that anime (as in: TV shows and movies) is more popular than ever, but its business has lots of problems. I'd love for someone to improve this situation, so that we could have more episodes of beloved sagas that are somehow paused, eg Berserk and One Punch Man.

But I'm completely uninterested in AI generation or parasocial stuff. Let art be art, even across mediums (comics, TV, games), but the rest just sucks the humanity out of the art.

LOTR was a work of passion , The Hobbit was a mess.

With The Hobit , there was a change in directors (originally Guillermo del Toro) and lots of studio meddling (eg pushing for 3 movies when the original is a 200 pages book, inclusion of LOTR actors for wider appeal).

I think Peter Jackson was brought in to try to salvage what he could.

I concede that intent matters.

Maybe a better analogy is a bank with open lockers and no vigilance: if someone enters and steals money, the police will look for them, because "the coffers were open" is not a valid defense. But customers will also demand answers from the bank - why were they so negligent and incompetent that someone can just enter and get their money?

We should hold similar values for digital systems.

Was the author's intent on stealing private data and causing harm? Did he gain from this abuse? Did the company take enough measures to safeguard their data?

Companies have been mostly not held responsible for their fuck ups, and no matter the law, that's wrong to me.

Seems outlandish. Citation needed? I'm aware of a couple of cases in the US, but not all over the world.

Secondly: can consumers be blamed for gross negligence? It's not reasonable for a bank to post account balances in public billboard and ask people not to look at others. We should contest when private data is available publically, hidden only by small obfuscations, not professional security practices.

I agree with all your points, except list comprehensions. I used to like them, but now refer the callback style from Javascript much better.

Cannot auto complete at start:

My auto complete doesn't work at the start of a list expression. In [for obj.abcdef...], what is obj? It cannot know. In a callback, using typescript , it knows: list.forEach((obj) => obj.acdef...): obj is an element of list, which, if typed, we can infer its value.

It's minor but it's not optimal.

confusing multi line comprehensions

I don't know how to format this in HN, but this is pretty weird in multi lines: [obj.y \n for obj in my_list \n if obj.x >3 ]. Something like this reads much better: my List.\nforEach((obj).\nfilter((obj) =>obj.x >3).\nmap((obj) => obj.y))

I'm comparing with Javascript which I don't enjoy much either. But I liked using this style in C#, with the LINQ nodule.

The point is that Ballmer is an adversary, and may choose the worst cases for binary search. As I understood, the algorithm in TFA holds against any choice.

As others said, if you don't expect adversary behavior in your data, it should be good enough.

Poetry does a lot of good stuff: lock file, dev dependencies (and other customizable groups), easy to use virtual env, dependencies pulled directly from git.

I do not recommend pip in serious projects, as it lacks the above. Here's things that I've seen:

- add dependencies and not save them in the requirements file - forget to activate cenv and install packages globally - different package versions across deployments due to absence of lock file - main dependencies mixed with dev dependencies

tl;dr: you keep the copyright but give inconditional rights for elastic to distribute your code, eg under different licenses.

This article is very forgiving towards tech companies, suggesting that users are techno phobic and the solution is to remove such labels.

My perspective: consumers have seen new tech emerge a thousand times and are favoring reliability instead of the flavor of the month new tech, especially when it was designed to sell instead, not to solve real problems.

This seems just like TVs, where more people long for dumber TVs with quality display - they are faster and more reliable than what the market of smart TVs is providing.

I think that's a very forgiving perspective.

Product teams are often wrong, and A/B tests can lead to inferior long term solutions.

As always, it's best to criticize the idea, not the people.

Does Nvidia negotiate deals for all gaming + crypto + AI companies to use their cards, or do they just provide good hardware + software that consumers want to use?

That would be the difference of an anti-competition monopoly vs just a monopoly.

Veritasium has a video on this:

https://youtu.be/rhgwIhB58PA

I accept them.

I've dealt with problems from python-dotenv and very much prefer it as a command than as a library.

As an example, once I changed a .env fikr and unit tests started failing. Digging deeper into it, and lots of code was checking for .env to load its configurations, and would break without it. I'd prefer this not to happen, as our tests were executing based on outside, not version controlled, configurations.

After removing dotenv as a library and using it only as a command, we were able to separate configuration and logic, and not have .env files affecting our unit tests - we simply ran the application with dotenv command, and the unit tests without.

> Generally, ad-tech and friends cares very little about small cohorts of users.

Sure, and it will keep being that way. But if this gets improved at the transport layer, seems like a win.

As an analogy, if buses are late because roads are bumpy and drivers are lousy, fixing the bumpy road may help, even if drivers don't change their behavior.

When I write code, I'm concerned about two things: - it does what I want - it's easy for others to understand it

If I'm doing red green refactor, the red and green phases are all about logic and constraints, but the refactor step is all about communication - I focus on making the code easy to read for others.

But if I'm reading code, I just want to know what it does.

Só, I didn't make a study, but I'd argue that writing code is more similar to writing natural language than reading code is to reading natural language.

That's a bit of a strawman: you're twisting the scope and arguing for it.

A more similar context would be: they ask Tom Hanks to create a voice similar to Woody, the cowboy from Toy Story . Tom Hanks says no, Disney says no. Then they ask you to voice their cowboy voice. It's obviously related: they tried the OG, failed, they're going for a copycat after.

But if never approached Tom Hanks or Disney, then there would be room for deniability - without mentions to real names, it would require someone to judge if it's an unauthorized copycat or just a random actor voicing a random cowboy voice.

It was a bad play from their part.

Companies want to have their cake and eat too: the code is free, but they're complaining there's no warranty.

If companies want better guarantees, then they should contribute more.

It's an interesting question, but I doubt it will happen. Companies gain even more (their free code is now more reliable) and FOSS maintainers lose more (now they're liable for the code they give away).

More questions: How can FOSS maintainer be compensated for this? Are they liable in every country? Etc etc.

Alternatives: companies could do public audits of specific software/library versions.

I can see his point: deleted code is really hard to discover. You can fetch it if you know it's there, but how do you find it in the 1st place?

With that being said, commented code is rarely useful and it cna be a liability indeed.

This approach is full of footguns.

Forget to install virtualenv? Have fun getting rid of packages in your global system.

Sub-dependencies? You'll have to think of a strategy for it. Most people don't and the reproducibility of their project suffers.

Developer vs production dependencies ? You'll have to think of a strategy for it.

And I think I'm forgetting one, but that's already enough.