0xpgm

Also, even though it called itself purely object-oriented, the earlier versions lacked inheritance.

Nowadays object oriented programming is almost entirely associated with inheritance hierarchies.

Perhaps inheritance isn't as central to object oriented programming as initially envisioned.

For a long time Guido wanted CPython to remain simple, as the reference implementation.

Probably also during the 2 to 3 migration, but that goal appears to have now been abandoned.

I wonder how the authors of other implementations like pypy will be able to keep up as the reference implementation becomes more and more complex.

How I've done it is to use some kind of observer pattern, where you register listeners to state changes.

Whenever state is updated, the listeners are all called with the updated state and can update the UI and what not based on the new state.

I haven't profiled this and compared it to frameworks, but feels adequate for the relatively simple UIs I've done.

On Margaret Hamilton from Chapter 5 of Hackers by Steven Levy

"She was well aware of the hackers’ playfulness around the ninth floor, and she was moderately friendly with some of them, even though they would eventually blend into one collective personality in her memory: one unkempt, though polite, young male whose love for the computer had made him lose all reason."

> I'd rather my language surface problems at compile time (via type errors) for all possible code paths rather than a particular codepath at runtime.

The kind of problems caught by compile time type checking are less fine grained than those that can be caught at run time. For more exploratory programming, one would prefer the latter.

At compile time, one checks that they are passing around correct types, at run time, the actual logic of the program is exercised.

Awesome! thanks for sharing these :)

Right, I get it.

People tend to underestimate to what extent pre-LLM programming experience contributes to their ability to steer LLMs in the right direction and produce decent results.

In 5 to 10 years, we'll see what kind of engineer is produced from pure LLM steering without years of hand coding experience. Even many of the more experienced ones would have some of this skill atrophied.

Presumably a DSL would allow you to do more heavy lifting in your specific domain with less tokens.

Let's way you wanted to do a web application using LLMs. Using a web framework would cost less tokens than using the vanilla underlying language (Python, PHP, whatever..), which is again way less tokens than building up from assembly (an LLM should be able to do this given enough time and compute).

Makes sense.

Why not let the shell do the expansion of `somedir/*` to refer to the content of the directory instead?

Running a company is hard work. If the current leadership in these companies cannot act responsibly, they need to make way for leadership that can.

There are many companies that compete but are careful not to break laws or cause obvious harm.

Why should a billion-dollar funded corporation still want to externalize the costs of its actions?

So perhaps the news here should be that OpenAI didn't take security seriously in their experiment, rather than the narrative that AI agents are a looming danger to the world.

As heavily funded as the top AI startups are, how is it that they cannot fill every single role with the best expertise available?

Is tech hiring so badly broken? Or do they have such broken processes / misaligned incentives that even people who could be doing a better job in these companies are unable to?

Also, was something lost in the transition from the traditional 'sysadmin' role to 'platform engineer' in the 'cloud native' environment?

If humans were directly making these attacks law enforcement would already be onto them. But because the humans routed these attacks through AI agents, it only makes news and contributes to AI hyperbole.