HNHacker News
TopNewBestAskShowJobs

joshuamorton

7,723 karma · joined February 19, 2017

In case I forget to disclose in a thread, I currently work at Google. That said, opinions I express belong to me and all that jazz.
submissionscomments
joshuamorton··on Florida judge rules red light camera tickets are unconstitutional
> In some ways the government bringing civil charges against you is rather bullshitty

I think there are circumstances where this is true, but I don't think it's true in the general sense. And I really don't think red light cameras, which are incredible for public safety and a really fair enforcement tool, are a good example of a civil rights violation.

joshuamorton··on Florida judge rules red light camera tickets are unconstitutional
> It is enough to say absolutely nothing, and request the government to prove its case.

Only in criminal contexts. In civil contexts your silence can absolutely be an adverse inference. Usually these red-light cameras are civil penalties, not criminal (fines with no points). The judge here seems to be saying that these are "quasi-criminal" because, uhh, I guess there are penalties.

joshuamorton··on Show HN: Skir – like Protocol Buffer but better
How does this work if, for example, you persist the data in a database?
joshuamorton··on Don't become an engineering manager
I mean, this is true and also not true. The management ladder does start higher and arguably does end higher with C-suite titles (though even that may not be the case, e.g. Google now has a Chief Technologist or two, who are SVP/C-suite level). But at the major tech companies under discussion in the parent, yeah the levels do map basically 1:1.

You can maybe argue that it's easier to advance on the management side, and I think that was certainly true pre-2023, but since then there's been a clear tightening of easy empire building advancement, and at least in my neck of the woods, I've seen IC-track advancement has been faster from staff to principal than from manager to director.

joshuamorton··on Don't become an engineering manager
The question to ask of course is do you see how your manager interacts with their management?

Do the things that get done stay consistent, or are your priorities constantly changing? Do you and your team control your priorities or are they out of your control?

I think it's very difficult to see what managers are even doing (I'm not one, I'm a senior IC) until you have a particular level of visibility or exposure to planning and prioritization. But once I started to see that and see why it was hard as I began to interact with it myself, it becomes much clearer.

joshuamorton··on We Will Not Be Divided
> Imagine if a gun manufacturer sold a gun that you couldn't use against X or Y country.

The point here, of course, being that Anthropic is very specifically claiming to not be a gun manufacturer, and Hegseth's response is that the DoD (W?) will force anthropic to build guns.

joshuamorton··on Open Letter to Google on Mandatory Developer Registration for App Distribution
> My custom youtube clients would never be approved by google. My (free) apps for watching anime and reading manga would never get approved by Google.

Google isn't approving apps though. A developer provides identity verification and a set of apps (apk names & keys) they are responsible for. There is no verification process or approval from google. The entire process as outlined in https://developer.android.com/developer-verification is that you prove you own signing keys for an apk name.

joshuamorton··on Open Letter to Google on Mandatory Developer Registration for App Distribution
> Laundering millions is a huge amount of work already. You need to hide your criminal activity from banks investigating fraud. Presuming the banks are doing their jobs right, at least, but if they don't, then that'd be the place to start solving this problem.

It's not really clear that this is money that needs to be laundered, it's often irreversible transfers that are legit.

> People are already effectively faking addresses for something as stupid as Amazon reviews. Apparently it's that cheap to fake an address, because those crapware spam stores that rotate their name/products/listings aren't exactly the size of the mob.

I already responded to this below. Those don't involve scammer controlled addresses. If I send you a piece of physical mail with an OTP code, you can't use a random faked address.

> clearing the field for "professionals" while at the same time making identity fraud, address fraud, and (money) mules more lucrative.

The majority of this kind of fraud is already organized. That's why raising the cost is impactful, see my comments below. It's a tool to raise the cost of revenues to an ideally unsustainable amount.

joshuamorton··on Open Letter to Google on Mandatory Developer Registration for App Distribution
> You could also pay randos a $30 gift card to sign up for a developer account and share access. Enough people will do it.

This could work, but the issue here is that a lot of these scams rely on the "zero cost"-ness of turnup and use that as a asymmetry. If it costs you nothing to turn up new scam-accounts, and it costs me something to investigate and remove them, you win. If it costs you $10 to create new scam accounts then as long as I can get the EV of a scam account below $10, the scam isn't worthwhile.

joshuamorton··on Open Letter to Google on Mandatory Developer Registration for App Distribution
You now need to have a variety of fake addresses you can use, since scammed addresses will get banned. You also need fake IDs. So again, the bar has now been raised from "run a bot to make fake Facebook accounts" to "I have a large number of physical addresses and the ability to create arbitrary fake government IDs".

> Amazon has a huge problem with packages being sent to fake people at different addresses.

This usually involves those people getting weird packages and not doing anything with them, it doesn't require attacker-controlled addresses.

joshuamorton··on Open Letter to Google on Mandatory Developer Registration for App Distribution
But the proposal here, requiring developers to register their identities, doesn't actually impact consumers at all. They still have the ability to make the decision about whether or not to trust someone.
joshuamorton··on Open Letter to Google on Mandatory Developer Registration for App Distribution
> Someone will manufacture and sell bulk identities

How? You've now moved the level of sophistication required from "someone runs some bots on the facebook website" to "someone is now committing complex fraud against a government".

If the only people who can run scams are state sponsored, that's still vastly better than the status quo.

joshuamorton··on Open Letter to Google on Mandatory Developer Registration for App Distribution
> Now they'll need to pay off a local mailman to give them all of Google's letters with an address in an area they control so they can register a town's worth of addresses, big whoop. It'll cost them a bit more than the registration fee, but I doubt it'll be enough to solve the problem.

Yeah, this is a huge amount more work than, like, nothing.

joshuamorton··on Open Letter to Google on Mandatory Developer Registration for App Distribution
should I be confident that thunderbird.net is the real one, or could it be hosted at thunderbird.org, thunderbird.com, or thunderbird.mozilla.org?
joshuamorton··on Google restricting Google AI Pro/Ultra subscribers for using OpenClaw
> This all revolves around implied ability... Google should not have this capability

The entire "ability" here is, as far as I can tell, is that it's possible to connect accounts to IP addresses. This is something that practically every system does. HN does it to stop abuse and ban-dodging. Wikipedia does it. You're reading an incredible amount of bad faith into the concept of IP bans.

> but effectively you are defending your employer from a claim of gathering data without consent.

I'm very specifically not doing that. I've made no comment on what the practices of any particular company are. Ultimately I don't know. What I do know is that the comment you're replying to doesn't say the things you keep implying. If your goal is to silence any disagreement, please feel free to continue speaking like this, but if your actually interested in engaging, I'd implore you to appreciate that I'm speaking for myself and not threaten me with..whatever you're implying here.

> This all revolves around implied ability, I don't give a rats ass about whether or not there is an actual implementation of that ability

I mean this is critical. If all you care about is the "ability" every site on the internet that you can log into has the "ability". All of them, every single one. The stance you're taking here is that a website supporting the ability to log in inherently violates your privacy because, whether or not it does, it has the ability to track information that could correlate your account with other accounts (and many of them do!)

joshuamorton··on Google restricting Google AI Pro/Ultra subscribers for using OpenClaw
I'll be more explicit,

> google would be tracking who people's partners are

is a misunderstanding of that comment. Nothing they said implies that Google is tracking who people's partners are. You're welcome to have whatever opinions you are about companies, but I'd also hope that you're careful not to read conspiracies into places where they aren't stated, especially in about institutions you have preconceptions about.

joshuamorton··on Zero downtime migrations at petabyte scale (2024)
I will simply point out that I'm aware of larger zero down-time migrations, like this one: https://www.youtube.com/watch?v=ih97gwNmkRA
joshuamorton··on Zero downtime migrations at petabyte scale (2024)
The code is open source though, you can read it. The cli examples point you towards the relevant bits of the actual database code to read.

For my own sake, I'm not sure what is so surprising here. "Turn up a hot second replica and fail over to it intentionally behind a global load balancer." Is pretty well trodden ground.

joshuamorton··on Google restricting Google AI Pro/Ultra subscribers for using OpenClaw
I think you're vastly misunderstanding that comment.
joshuamorton··on Trump's global tariffs struck down by US Supreme Court
They don't delegate the policymaking. Tax code is always congressionally approved, and I'm unaware of any even remote argument that changing tax policy is delegated to the executive.

OTOH enforcement of congressional policies is basically always the role of the executive, so the fact that the IRS exists and does things doesn't really impact delegation.

joshuamorton··on Zero-day CSS: CVE-2026-2441 exists in the wild
> A comment does not automatically make code safe. What even is that argument? There have directly been examples of Rust code with SAFETY comments that later were found to be memory unsafe.

I did not make this argument. I encourage you to reread my comment and do so with the HN guidelines in mind!

joshuamorton··on Zero-day CSS: CVE-2026-2441 exists in the wild
> but its code has lots of unsafe.

And every instance of unsafe that I could find (except one, in test-only code) was a call to libc with a clarifying comment on why this particular use was safe. That is, all (or at least, all of it that I could find) was wrapping an unsafe API with documented (and usually straightforward and local) invariants that maintain safety, such that the calling code is safe.

I'd say that the fact that miri's trophy-shelf[0] has 39 entries and is 7 years old and still regularly updated is a pretty good indicator that memory bugs are sufficiently rare in rust so as to be notable. That is the opposite of "regular"

[0]: https://github.com/rust-lang/miri/blame/master/README.md

joshuamorton··on Zero-day CSS: CVE-2026-2441 exists in the wild
But not "routinely".
joshuamorton··on Show HN: CEL by Example
Correct, but you can also reject filters that will take longer statically. The point is not "any arbitrary CEL program will run in less than 10us", it's that I can encode "do not allow filters that take more than 10us to evaluate" an then have a very high degree of confidence that that will be true for any user provided filter that is accepted (and if I'm wrong it'll be...11us, not 5s)

In the common use-cases for CEL that I've seen, you don't want to skip evaluation and fail open or closed arbitrarily. That can mean things like "abusive user gets access to data they should not be allowed to access because rule evaluation was skipped".

You also may have tons of rules and be evaluating them very often, so speed is important.

joshuamorton··on Show HN: CEL by Example
What you really want is "can be completed after a certain amount of time", not "can be cancelled". You don't want iam policy rules to be skipped because they took too long.
joshuamorton··on I'm building a clarity-first language (compiles to C++)
This is very language dependent. People coming from python or Java would call them lists.

Vectors are a mathematical concept unless you use c++.

joshuamorton··on The Waymo World Model
I think there's two important things here:

1. You want to be forward looking, not backwards looking. Cutting services means less ridership means less revenue means cutting services means...etc. Bart is super useful for me during the week because headways from SF to West Oakland are often 5m. As I'm writing this (11 on a Friday) I missed a train and had to wait 20 minutes. Every seat on the car is also full, and while not packed, it's standing room only. If my choice is to wait 20 mins for the next train, other ways of getting places become a lot more appealing.

2. Government services should be good. This is good both because it makes people interested in using them (see 1) and because people who don't have other options deserve good services. The point of government is, at least in part, to serve those who can't serve themselves. I don't expect Bart to be revenue neutral for the same reason I don't expect CalFRESH to be.

joshuamorton··on The Waymo World Model
> You can't say that BART "is facing bankruptcy is precisely because it was mostly rider supported and profitable, and not government supported" when it is very obvious that BART would be in a much worse situation if it had had more government support... because all those governments are facing massive budget deficits right now.

I don't think this follows. Government budgeting isn't zero based. A hypothetical Bart with 2x the government funding in 2019 would have faced cutbacks, but likely has more money today than what we have now!

> or seek other forms of revenues while we wait to see if ridership returns.

Yes, this is called "taxes".

> If ridership never returns, then we ought not continue throwing good money after bad

Agreed if it was stagnant, but ridership is up more than 10% y/y and that was also true last year. It's on track to be revenue neutral again in a few years. Gutting services today would be exactly opposite of what you'd do for something like a startup showing clear path toward profitability.

> Obviously the main problem here is that it's literally illegal to just build high density corridors directly adjacent to the transit stations... which is what we ultimately need to prioritize

While sure it's hard, there's lots of these that exist. There's new stuff in oakland basically constantly, and were even seeing midrise stuff along Bart in SF, but it's units being built now, so they won't be available until 2027, which is when your proposed service cuts would hit.

joshuamorton··on The Waymo World Model
It's worth noting that, at least for bart, the reason that it is facing bankruptcy is precisely because it was mostly rider supported and profitable, and not government supported.

When ridership plummeted by >50% during the pandemic, fixed costs stayed the same, but income dropped. Last time I checked, if Bart ridership returned to 2019 levels, with no other changes, it would be profitable again.

joshuamorton··on Common misunderstandings about large software companies
> * Defining directly responsible individuals / single-threaded leaders so that every choice doesn't involve massive coordination

Obviously doing this makes sense when possible, but in cases where you see it, it's usually due to a fair amount of work to have made this possible.

To give a simple example, I work on a class of problems that require a VP level approval for new instances of $thing. I've gotten this down to a pretty straightforward process where I can work with folks who are proposing a new instance, get things working, and then the VP usually is happy to stamp the work we've already done. Though in some cases they aren't, and they have (good, probing) questions or changes they'd like us to make. But that's only possible because I have a longstanding relationship with the set of folks who ultimately approve this kind of thing, I have worked with them on a process that they like, and I have the experience to shepherd things effectively, and their trust and buy-in. And I'd argue that my case is pretty simple, because while there are a handful of responsible people that I could go to, there's rarely any individual concerned.

Consider a different, relatively common case, of a client and a server that are owned by different teams. The client wants a feature in the server, and perhaps is willing to do some development and loan headcount to implement the feature. Who is the single responsible party here?

It should probably be either the manager of the client server team or the mutual manager of both if the teams are fairly close in the wider org, but then you have multiple indirect layers between the people doing the work (client team members) and the accountable person (server team manager), and that's the state you get to after you've done a bunch of coordination work and gotten everyone to agree to that division of labor and accountability.

← PreviousPage 3 of 34Next →