HNHacker News
TopNewBestAskShowJobs

lmm

47,766 karma · joined October 14, 2011

I'm known as m50d some places, including github; my blog (such as it is) is at https://m50d.github.io/
submissionscomments
lmm··on Outsourcing plus local AI will soon become more economical vs. frontier labs
> How much is Waymo burning a year? 3B on 300M ARR? Anthropic is what 5B on 20B ARR? Waymo is 3x older. Why don't we hear such confident statements about how subsidized their rides are?

We do. We hear it less often because no-one is talking about how Waymo changes how we all need to work or whatever, that's all.

lmm··on Opaque Types in Python
> the option to litter side effects throughout the file and having those side effects execute automatically on import.

You can actually do that in Java too with static initialisation blocks.

> 3.0 was a big missed opportunity to kill a lot of the deprecated cruft.

Maybe. 3.0 already came damn close to killing the language. If they'd made the changes any more radical it could easily have been another Perl 6.

lmm··on Opaque Types in Python
Wow, that's a much better reason to not take them seriously.
lmm··on Opaque Types in Python
> Why go through the ceremony of `public static void main(String[] args)` when Python just executes the script line by line at the top level? Oh wait, now you have things like `import` actually executing code instead of simply being a compile-time namespace convenience, and you need weird techniques like `if __name__ == "__main__"`.

You'll note that even Java now recognises that "public static void main(String[] args)" was pointless ceremony.

> Python scripts were meant to be lightweight and free of the tyranny of enterprise OOP which was epitomized by Java. But people found out that keeping track of types in your head is laborious and error-prone, and getting a compiler to check {that the shape of your objects and function calls match} is a huge productivity boost. And so Python 3 enabled static type hints... which, like I said before, Java had from day zero.

Java didn't have "hints", it had mandatory types everywhere. Which, again, they've now recognised was a bad idea and gradually removed. Having a type system is a good idea, and maybe if Python had taken more inspiration from OCaml (or more people had used OCaml instead of either Java or Python) we'd all be better off, but the early-2000s-Java cure of mandatory types everywhere was worse than the disease.

> To make matters worse, static type hint features were introduced progressively over the years, leading to things getting deprecated from the `typing` module and moved to things like `T|None` and `list[T]` and `collections.abc`.

Which happens in any language that evolves over time. Java, despite having a type system built into the language from day 1, now has about ten different deprecated sets of nullness annotations with overlapping functionality.

> I really don't think having top-level (module) variables and functions in Python is a good thing, especially because then they are duplicated as fields and methods in classes. In Java, fields and methods (whether static or instance) can only be placed in classes, and I think this particular straitjacket is a good thing.

Nah. No other language has adopted Java's approach, for good reason. Functions and values are easier than classes and the language shouldn't force you to complicate your code when there's no need to.

> Probably the most tragic example is the ways to build up strings in Python: `+` and str(), `%` operator, `str.format()`, f-string.

It sucks, but what's the alternative? Either a language dies young or it lives long enough to have a bunch of deprecated crap in it.

lmm··on What we lost when we stopped letting kids leave the front yard
> It's the most common sports related head injury by a wide margin

Interesting, but we're not talking about sport cycling, we're talking about ordinary commuting.

> helmets are quite effective at injury reduction.[1]

Your citation provides no support for your claim.

> Anecdotally I've flown over my handlebars and hit my helmet without serious injury, and I'm sure I would have been in much worse shape otherwise.

This is why we don't make public policy on the basis of anecdotes.

> On cars the law requires seat belts and airbags and a variety of other legally enforced safety measures.

All of which are a lot less unpleasant at the point of use than helmets.

> If you have a study suggesting that helmets would significantly help I'd be curious to see it.

The rates of head injury are comparable (something like a factor of 4 difference), head injuries are very common in car collisions, helmets are already routinely used in motor racing. The evidence is no weaker than that for bicycle helmet use.

lmm··on What we lost when we stopped letting kids leave the front yard
I'm sorry for your loss, but your school is very much an outlier; public policy should look at the overall rates, and statistically it is rare. Particularly if we're talking about an everyday school commute along surface roads (the risk profile for mountain biking or BMX-style stunt riding is quite different, and wearing helmets for those activities makes a lot of sense).
lmm··on Use boring languages with LLMs
Rather than "boring", this seems to be reaching for something like the concept of a "pit of success", or https://haskellforall.com/2016/04/worst-practices-should-be-... . I don't think the fact that the most common pitfalls in Go are well known should be taken as a sign that it doesn't have more esoteric pitfalls as well; it's just that the common cases (like nil) are the ones that everyone sees all the time.
lmm··on DynIP – Dynamic DNS with RFC 2136, IPv6, DNSSEC, and BYOD
> we accept RFC 1918 and CGNAT addresses in records

Doesn't that cause security issues by making it possible to put other people's private servers (that you want to do XSS-type attacks against) into your domains or something? I have a vague memory of it being a security no-no somehow.

lmm··on Ferrari Luce
Which suggests that a similar but worse product shouldn't sell either?
lmm··on What we lost when we stopped letting kids leave the front yard
> Bike helmets mitigate one of the most serious and common forms of injury while riding bikes.

A form that is still extremely rare. No-one seriously advocates helmets for car passengers, for example, even though the injury rates are very similar.

> be hit by a car

Cars don't hit people, drivers hit people.

> They don't prevent you from doing anything you would otherwise do.

They're annoying enough that they do, in practice if not in theory. To say nothing of the fact that drivers pass you closer and more dangerously if you're wearing a helmet.

> helmets are a fantastic example of reasonable PPE, not overactive safetyism.

Quite the opposite.

lmm··on What we lost when we stopped letting kids leave the front yard
> Its presence would likely lower the insurance premium significantly, due to a much lower chance of hitting someone right ahead of the vehicle.

Why? It's not like drivers have to pay up when they hit someone, as long as they weren't drunk. And in the unlikely event that a driver does get made to pay the big risk is medical bills, so the incentive is to make sure the car is set up to always kill anyone they hit.

lmm··on What we lost when we stopped letting kids leave the front yard
It's more cars per person yeah - more people working per household, more car dependency, and cars becoming cheaper.
lmm··on Exit IP VPN servers mitigation rollout
If you trust your browser it's fine, and if you don't then both CSS and SVG are significantly more risky.
lmm··on Nobody cracks open a programming book anymore
It was great 15 years ago before the Iron Law of Bureaucracy kicked in and the powermods took over.
lmm··on The Eternal Sloptember
Something that would disprove the author would be an increase in industry code quality. A reduction in bug rates might be a reasonable proxy measure.
lmm··on The Eternal Sloptember
It's not periodic so much as continuous. But yes, HN is constantly where Reddit was about 3 years ago and correspondingly declining. E.g. see the rise in "underrated comment" type comments lately.
lmm··on Jira Is Turing-Complete
That explains why it's impossible to tell whether any given Jira operation is going to halt or not.
lmm··on Australia Four-Day Work Week Study Data Shows Boosted Productivity
> Naive question but if it works best wouldn't companies that have a four day work week outperform theirs peers and because of that grow faster, and become more common?

Eventually, but what's the typical lifecycle of a company? And if e.g. Treehouse succeeds or fails, was that because of their 4 day work week or because of any of the hundreds of other reasons a company might succeed or fail?

lmm··on On The <dl> (2021)
> So then what's the point of making this distinction?

It keeps the semantic web people busy not doing too much damage. Otherwise we might get an XHTML3 or CSS4 based on RDF triples or some such horror.

lmm··on Python 3.15: features that didn't make the headlines
> And maybe calculate the Levenshtein distance between all terms if it’s not found, and just assume it’s the closest one.

So when a library adds a new method, it silently changes which method client code calls? That's a bit too magic IMO. I think the best you can do is be case-insensitive and ban methods that differ only in case (or, if you want to extend the idea a bit more radically, ban having things in the namespace within Levenshtein distance x of each other, and then you can autocorrect errors smaller than x/2).

lmm··on We're testing new ad formats in Search and expanding our Direct Offers pilot
> Anecdotally, the number of people who have asked me about VPNs over the last 12 to 18 months has skyrocketed. I probably get a message a month from people of all backgrounds.

That's probably more to do with the increasing popularity of China-style national firewalls/tracking (sorry, "age verification") than ads.

lmm··on Nobody understands the point of hybrid cars [video]
Is anyone even selling those in the US these days? Obviously there's a (dwindling) supply of used ones, but I can understand wanting to put your kid in a new car for safety reasons.
lmm··on Nobody understands the point of hybrid cars [video]
> on the Mediterranean islands usually rental is the only option

It's not like buying an ICE or hybrid instead of an EV helps with this scenario at all.

lmm··on Haskell Foundation 2026 Update
accursedUnutterablePerformIO or bust :)
lmm··on Incident Report: May 19, 2026 – GCP Account Suspension
> I can't for the life of me think of something that qualifies more for being a 125b company than actually having 125b in assets.

Which this company didn't. They managed 125b of assets belonging to other people, they didn't have 125b of their own.

lmm··on Google Declaring War on the Web
I want web pages to stand alone, not be part of a newsletter I'm meant to subscribe to. Maybe we could, shock horror, share links to individual good pages.
lmm··on Dumb ways for an open source project to die
> Business logic just works with data it has access to. Backup, encryption, access control can be separate concerns. A good programming stack would make sure you don't have RCE.

The problem is that the business logic tells everything else what to do and has authority to do everything that the program can do. Exploiting that way is fiddlier (it's like doing ROP only more so), but ultimately once the software gets into an unintended/unexpected state it's game over.

> It's just that we are used to coarse-grained permissions and abstractions defined back in 1970s. E.g. an app gets access to entire network stack and then can do anything - send telemetry, spam, download code, etc. If we had more high-level comms layer on top of app it could be much more inspectable.

Many have tried. If you try to require fine-grained permissions and user in the loop people just say yes to everything. Ultimately the user thinks they want to do the thing they said to do, and asking them again won't change that.

lmm··on Dumb ways for an open source project to die
Yes and no. In practice sooner or later you need to update something - you need it to work with a new OS/database/TLS algorithm - and then often there's a domino effect.
lmm··on Disney erased FiveThirtyEight
> Enough people did not vote against Trump is my point.

The argument is that there were enough people who voted against Biden-Harris, or against the DNC process, to swing it.

lmm··on Dumb ways for an open source project to die
Separating the runtime from the business logic doesn't really work because the business logic has the authority to do anything it has the authority to do. It's the https://xkcd.com/1200/ problem all over again.
← PreviousPage 7 of 34Next →