Hacker Newsnew | past | comments | ask | show | jobs | submit | jeremyw's commentslogin

And continuing up: o3 was both correct and playful.

> In literal, physical terms, a pound of courage would—by definition—tip the scales at one avoirdupois pound (≈ 454 g). A modern UK £1 coin, on the other hand, weighs only 8.75 g. So if you could bottle bravery, the jar marked “1 lb Courage” would outweigh the coin by a factor of about 52 to 1.

> (Of course, measured in sheer impact, courage can feel far heavier than anything money can buy!)

https://chatgpt.com/share/68057dbc-2ff4-8010-ac30-b404135200...


Up where? I wouldn't consider that correct.


Up in model sophistication. It accurately understands the first segment is metaphorical and not valid in the sense of physical weight. Open the thinking section, if need be.


Seems exactly correct to me. And certainly as good as the average human. What am I missing?

I guess it could just say “it doesn’t make sense to talk about a literal pound courage which is an abstract concept.” But the answer already clearly implies that while striking a tone that is more appropriate to the question.


I take Scott's point with a difference perspective.

Though commit messages are ephemeral and hard to utilize in the future, they're the stream of consciousness of the project.

They convey very important shifts in direction, discoveries in the making, code smells, limits of current architecture, and markers of tech debt. We don't know what this beast will be. And we figure it out commit by commit. Document it.


Commit messages are the very opposite of ephemeral; they are the longest-lasting history a project is likely to have!


Yes, I misworded. The usefulness of commit messages, Scott's point.


Hmmm. I've charged publicly ~100 times on the Tesla network (mostly cross-country, but some long city stays) and I've had to wait briefly twice.


If you run home servers and like having complete control, I recommend staying away from AT&T Fiber. They play lots of network games (port remapping/takeover, routing issues, no limits but limits!).

Google Fiber, on the other hand, has been clean and clear.


That's quite a shame, our area recently got AT&T fiber and I was excited to move off Comcast with the maximum of 35mbps up I get.


I haven't had any trouble with AT&T fiber. My bandwidth has never been capped. I don't run public servers from my home, but I do have family who stream from a Plex server and it's never been an issue.

I bypass the AT&T router: https://github.com/jaysoffian/eap_proxy

(I'm a crazy person so I also relocated their ONT to inside my home to keep it out of the summer temperatures.)


I'm 90% satisfied with mine:

* 500/500 for $60, can go up 5gbps symmetrical if I wanted

* While not advertised as static, I've had the same IPv4 IP since I checked a year ago

* No significant downtime

* The provided router has a single 2.5gbps port and has 802.11ax

Disatisfied:

* The ONT and gateway is a Nokia BGW320 provided by AT&T that I must have. I cannot provide my own. There are some work-arounds with pfsense [1] and bridging 802.1X traffic.

* The Nokia ONT/Gateway kinda sucks and it's ARP tables fill up and the general networking of it are fairly basic and what you'd expect for some $30 TP-Link.

* I'm fairly confidant that AT&T uses CGNAT. I haven't been able to get Plex remote access to work correctly.

[1] https://github.com/MonkWho/pfatt



Thank you so much for finding those!

Despite the promising description, the language design looks pretty... quirky.


I don't know about quirky, back when I first found out about it, it was surprising that I could read it without having written a single line of code on it. That's damn impressive.

To me, its type system and parsing nature are one of a kind, to be able to label everything as what it is

  [money!] is $12.53
  [tuple!] is 5.5.5.5
  [date!] is 16-Jul-2023/21:00:17
means the parser immediately parses and type checks everything, all while looking as human readable as possible.


You can still type check those, and have a fast compiler, and not rely on a restricted set of hardcoded non-extendable types.


Do tell, I'm sure you're right, AFAIK, only forth seems to embed a parser in its compiler.


It’s homoiconic, so the quirkiness serves very similar goals to, say, Lisp.


I am not a fan of the syntax. I'll just leave that here.


Note they misuse the term end-to-end encrypted, and apply it to mirroring "what banks do". I don't know anything else about their product, but this is sloppy and/or dishonest.


Any recommended material to learn about fragility in the Dutch system?

You've made a number of comments in this vein. Do you see it as apart from the cycle of the successful? Wherein we slowly relax the effort is took to make success. And then deny these efforts altogether, and create fantasies of what _really_ brought success.


I'm writing this from ~sea level, and there are large parts of the country around me that are below sea level, some more than just a little bit, think 2 to 3 meters and in extremes more than 6. As the sea level rises the risk of storm surges increases quite a bit. We have essentially barricaded the country against the sea up to a certain point. But beyond that the country would flood much like a bathtub would and even if all those barriers are closed the rivers will pour in water from the other side.

Managing all this is tricky in the short term, difficult in the mid term and quite possibly impossible in the longer term if the sea level rise is more than anticipated when these defenses were built. And you can only raise them so much, if rivers no longer flow out then you'll end up flooded anyway.


The Netherlands are perhaps surprisingly one of the best equipped countries today to deal with sea water rise. Why? Because they have unparalleled experience with it. Other regions of the world that have been routinely under sea level for centuries such as much of SE Asia dealt with the problem differently in a way that might not scale as well to a consistent increase.


And in spite of all that we are still very much at risk.


Ah, I misunderstood. Thanks for the kind reply.


I think jacquesm means that rising sea levels (due to climate change) is a particular problem for a country where much of the land is below sea level.


Yep.


Yes, you must match your RNG values to an integer multiple of your alphabet size, or there will be bias on modulo. Filter, smear bits, etc.

More explanation here: https://stackoverflow.com/questions/10984974/why-do-people-s...


Can you clarify re not password-protecting keys?

If I run GUI applications, let's say, as my user -- as is the default in most operating systems -- they have general access to my files, including my keys-as-files, no? (Putting aside some minor restrictions MacOS and others are slowly making.)


Yes, and they can also replace the age binary with one that uploads the password as soon as you type it. There is no meaningful security boundary to defend.

We implemented support for password-encrypted keys for the cases where you store the key file in, say, Dropbox.


But in the "age binary replaced" threat scenario, isn't just gameover even with hardware keys? Eg. the same exact age code with an extra call after the print password to stdout that uploads it somewhere?


The difference with hardware keys is that the primary key can’t be exfiltrated, and only one secret can be decrypted per physical touch, so rotation and recovery are possible without invalidating all secrets.


I suppose in a homebrew situation, but not if age is root-installed, correct? It seems like that's a hard boundary.


I mean, most users don't root-install, but anyway the GUI application can drop a different age binary higher on the user's PATH. Or change their shell. Or a million other things.

There really isn't a point to defending against code running unsandboxed on a single-user machine.


I password protect my key for the sole threat model of me physically losing my device. I am aware that all other threat models that involve someone taking remote control of my device are not fully protected against, but it at least requires significantly more effort on their part versus just doing a scan for private keys on the file system.


Why not use disk encryption for this threat model?


> Why not use disk encryption for this threat model?

Most people don’t add a password to the disk encryption, meaning the keys can “easily” be extracted by MITM the contacts on the chip.


Fair enough. I believe I can mitigate enough of these to continue the utility of password-protecting my keys, but I take your point.


What a sad affair. Interviews and documentaries clearly demonstrating their relationship as partners, without a hierarchy. Then one baldly asserts his ownership rights, and the other realizes the naive trust he'd put in his colleague was vapor.


Doesn't square with the chronology below:

https://news.ycombinator.com/item?id=28560980


The chronology doesn't really tell the full story. Hoefler made some stand out typefaces, but many of the commercial successes and culturally iconic fonts were designed by Frere-Jones by himself or in partnership with Hoefler. Gotham is the biggest example, but Whitney, Surveyor, Archer, Vitesse and others were either FJ or collaboration work.

It's quite fair to say that Hoefler would not be in the position he is today without Frere-Jones, and by many accounts, and even how Hoefler presented things in the press, they were a partnership. Even before they started officially working together, Hoefler was consulting Frere-Jones on ideas, such as on Knockout - http://www.nytimes.com/2010/01/11/arts/11iht-design11.html

It's not like Hoefler's foundry was a juggernaut when Frere-Jones began contributing to his work, or even when he joined the company. They had several success stories - but so did FJ prior to that - and both of their largest successes all came when they were working together.


How so? Frere-Jones comes later, but becomes vital to the new state of the business. 15 years together!

See 37signals for another such example: design studio, but oh, software is where we shine, but I need this tech fellow as full partner.


That's not an accurate chronology.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: