HNHacker News
TopNewBestAskShowJobs

tbrownaw

7,440 karma · joined December 21, 2009

Also tbrownaw@prjek.net on email. (Or @tbrownaw on Twitter, but I'm more likely to actually see an email.)

Whenever possible, teach the computer to do your work for you.

submissionscomments
tbrownaw··on Oregon gave homeless youth $1k/month with no strings
IIRC this is at least partly because we measure that differently (re-arrest in X years vs re-conviction in Y years I think it was?).
tbrownaw··on Oregon gave homeless youth $1k/month with no strings
> Based on responses from about half of the program’s participants

That "responses" is a link to the actual report.

1. The program was 120 people. 80 did an initial survey, 80 did a final survey, and there was an overlap is 60 who did both surveys. The survey was offered to all participants.

So, this is not a random / representative sample.

2. The program also included counseling sessions.

So, there's the potential for different results for money without counseling, or counseling without money.

3. I don't see any comparison to a control group.

For example, it's well known that homelessness is usually transitory. Without a control group, there's nothing to identify what was caused by the program vs being caused by the usual course of things.

tbrownaw··on Ode to the AA Battery
Well anything up to 240W can fit into USB-PD.
tbrownaw··on Ode to the AA Battery
I have some LiPo-based 1.5V AA and AAA batteries that have USB-C ports built in for charging. The AAA ones claim 1200 mWh (800 mAh) and the AA ones claim 3400 mWh or ~2250 mAh. (I say "claim" because I bought them off some random Amazon shop a few years ago and haven't bothered to actually check.)
tbrownaw··on FBI is investigating Minnesota Signal chats tracking ICE
> something that is perfectly legal

The goal is to prevent ICE / BP from doing their jobs. Which I rather suspect is not actually legal.

Thinking they're incompetent doesn't change that. Thinking the specific laws they're (nominally) enforcing are evil doesn't change that. Thinking that national borders are fundamentally illegitimate doesn't change that.

Perhaps the FBI had been ignoring this out of incompetence. Perhaps they'd been ignoring it as a form of protest. Either is interesting.

tbrownaw··on FBI is investigating Minnesota Signal chats tracking ICE
> Patel said he got the idea for the investigation from Higby.

This is confirmation that this wasn't being investigated until just now. This is surprising, I would have thought that "how are these people organizing" would have been an obvious thing to look into.

tbrownaw··on The '3.5% rule': How a small minority can change the world (2019)
> natural imperative

It happens because trust me bro.

tbrownaw··on The '3.5% rule': How a small minority can change the world (2019)
Well also the old school civil rights stuff didn't have supporters needing to engage in linguistic gymnastics in order to drum up support. Which makes just drawing attention to the issue rather more effective.
tbrownaw··on The '3.5% rule': How a small minority can change the world (2019)
> Regarding the “3.5% rule”, she points out that while 3.5% is a small minority, such a level of active participation probably means many more people tacitly agree with the cause.

The more effective you are at getting people to participate, the less effective that participation will appear to be. Because it's just a proxy for what actually matters.

tbrownaw··on Adoption of EVs tied to real-world reductions in air pollution: study
Tires yes, but EVs tend to have regenerative braking which will reduce brake particulates significantly.
tbrownaw··on Adoption of EVs tied to real-world reductions in air pollution: study
AIUI there are still disagreements about how to calculate that exactly. This study doesn't (and doesn't try to) provide any input towards settling that.
tbrownaw··on Adoption of EVs tied to real-world reductions in air pollution: study
This study is about air quality in neighborhoods. So it would show the same thing even if EVs just moved pollution from where people use their cars to where power plants get placed, because that's not the question it's addressing.
tbrownaw··on [dead]
https://www.merriam-webster.com/grammar/when-to-use-can-and-...
tbrownaw··on Recent discoveries on the acquisition of the highest levels of human performance
...so some people develop their sills using something like RAD* tooling which lets them develop skills quickly, and some don't and end up taking longer but getting better eventual results?

Also, the ungated part doesn't say how they're measure "top" high-school vs university students. It doesn't match what I've heard about the persistence and consistency of basically all standardized tests; are they using within-school rankings for this? If so, that would fit perfectly with students being sorted during university selection.

.

* https://en.wikipedia.org/wiki/Rapid_application_development

tbrownaw··on Stunnel
Is there anything that isn't horribly outdated that still needs this?
tbrownaw··on The challenges of soft delete
There are tables at $dayjob with both (begin, end) and also (incept, expire) fields. It's "on such-and-such date, X was true", but also allows for "as-of Z date, we believed that...".

Also you can have most data being currently unused even without being flagged deleted. Like if I go in to our ticketing system, I can still see my old requests that were closed ages ago.

tbrownaw··on IPv6 is not insecure because it lacks a NAT
> And indeed, if a firewall is off NAT can still function (if NAT is separate).

Well technically you can translate your /16 to look like a different /16 from the outside. IE each internal address gets turned into its own separate external address.

But that's not how NAT gets used in practice. How it actually gets used is to but many hidden addresses behind one or a few public addresses. And that multiplexing necessarily implies that incoming connections must be specifically told where to go; ie that there's a firewall.

tbrownaw··on IPv6 is not insecure because it lacks a NAT
> NAT isn’t actually a security feature—it’s an address conservation mechanism that became necessary because we ran out of IPv4 addresses.

> But the security benefits people attribute to NAT actually come from the stateful firewall that’s typically bundled with NAT routers.

1. It requires a stateful firewall.

2. It isn't possible to accidentally a default-allow rule on that firewall.

It may not be intended as a security feature, but it can't not act as one in practice.

tbrownaw··on IPv6 is not insecure because it lacks a NAT
> IPv4 is from the era of local computer networks, which feature clients and servers. Clients talk to servers, but servers are not supposed to care or even know about clients unless clients decide to reach out to them. Client-to-Client communication is generally discouraged.

No, it was meant to be a global address space where anything could talk to anything. That became unworkable due to scale and the limit inherent in using only 32 bits for the address space.

Some older protocols (ftp) don't play nice with NAT and need special handling, because address multiplexing was never intended to be a thing.

tbrownaw··on Bypassing Gemma and Qwen safety with raw strings
> It’s surprising how much society apparently thinks merely being above 85 IQ is sufficient to gate all kinds of things behind.

Doing the thing just needs to be at least as hard as automatically recognizing (ie without deliberately spending effort on it) that it's a bad idea to do the thing.

tbrownaw··on Level S4 solar radiation event
Make sure you have a surge protector or ups, in case it makes the power grid go funky. Which you should have anyway.

Also, it could be a convenient excuse to upgrade to fiber internet service if you haven't already. (Yes, excuse. Equipment should have more than good enough isolation to not care.)

tbrownaw··on Level S4 solar radiation event
But they should mostly be in the same part of the cycle rather than spread evenly.

It probably wouldn't make sense to calculate "average snow days per month" across an entire calendar year (in most places...), this is the same thing.

tbrownaw··on Porsche sold more electrified cars in Europe in 2025 than pure gas-powered cars
"electrified" is full-electric plus plug-in hybrid.

Does this mean that a non-plug-in hybrid would be in the "pure combustion-engined" bucket, or that they don't make those?

tbrownaw··on Prediction markets are ushering in a world in which news becomes about gambling
> I’d call that a pragmatic approach, not a conservative one.

The other meaning of "conservative", the one that's opposite "reckless".

It should in theory be possible to take a conservative approach to being progressive.

tbrownaw··on Prediction markets are ushering in a world in which news becomes about gambling
> Gambling isn’t a new problem, but apparently we thought it would turn out differently this time, for some vague unclear reason.

The things I've seen have been either "it's not right to tell me not to" or "non-participants can get useful information by observing the odds". What I haven't seen is claims that it won't be net-harmful to participants.

tbrownaw··on The guide to real-world EV battery health
> get like 250 miles or less range per charge. Still dramatically less than gas cars.

My old '07 Focus would get about 260 or so miles between visits to the gas station.

tbrownaw··on The guide to real-world EV battery health
> doesn't need neck-snapping acceleration

I thought this was mostly* a side-effect of electric motors inherently behaving differently than combustion motors.

* Not that it can't be deliberately turned off since everything goes through a computer.

tbrownaw··on What twenty years of DevOps has failed to do
> You are far better off with people that can work together across the bridge, but that requires actual mutual trust and respect, and we’re not able to do that.

Are you claiming it's fundamentally impossible for people to get along, or just that positive interpersonal relationships can't be reliably forced at scale?

tbrownaw··on The recurring dream of replacing developers
And you do this honestly, by negotiating reduced hours for the same pay or by negotiating piecework rather than time-based pay. Right?
tbrownaw··on Scott Adams has died
There's also the problem that the logical basis is "two wrongs make a right" and the factual basis is unquantified personal anecdotes and disparate impact.
← PreviousPage 10 of 34Next →