HNHacker News
TopNewBestAskShowJobs

filleokus

3,558 karma · joined June 6, 2012

[ my public key: https://keybase.io/filleokus; my proof: https://keybase.io/filleokus/sigs/X3sy2-d75jviQGx30hY5coCfZDG_epBylXhEwlSSHls ] filleokus_at_gmail.com
submissionscomments
filleokus··on CrowdStrike's impact on aviation
Hmm. I think this is a pretty shallow take.

My experience from the airline industry is that the vast majority of systems classified as flight safety critical are not connected to the internet, or large networks at all. Which is good.

But unless we want to drastically change how airlines operate, the rest need to be online.

Today, you can purchase a ticket (or rebook an existing one) on your phone really close to the departure time. When that happens, a gazillion interconnected systems, across legal entity borders, need to cooperate to take you (and your luggage) to the destination.

To put all of this in a large non-internet network seems pretty pointless.

If we wanna go down that route, the only real "security improvement" I can think of is to dismantle the digital systems and go back to paper. Like Ryanair did during this incident. Handwritten boarding passess, verified against print-outs of passenger manifests.

filleokus··on Show HN: SwimOS Rust – A framework for real-time streaming data applications
"real time" as in figma multiplayer not as in control system for a car (I'm assuming, never used SwimOS)
filleokus··on Schengen ain't what it used to be
> Passengers going through without document checks is a minor element of Schengen

In a pure economic efficiency sense you might very well be right. But the "illusion" that borders without border control used to give shouldn't be completely disregarded either. Pre 2015 (and Corona) when you could just roam around without ever showing ID just felt very different. It didn't feel like I was moving between distinct and sovereign countries, but more like moving between states in the U.S: "Oh now we crossed in to Texas cowboy noises".

Especially if you live in a border region (e.g Malmö-Copenhagen) and commute across the border, you really feel the difference and might even consider stop doing that because of hassle and increases in travel time.

During corona the closed Finish/Swedish border in Happaranda/Torneå just felt surreal, most warlike experience I've ever had. These cities are essentially one, and all of a sudden families, friends, businesses and their employees were split with fences and armed people stopping them.

filleokus··on Using S3 as a Container Registry
I've started to grow annoyed with container registry cloud products. Always surprisingly cumbersome to auto-delete old tags, deal with ACL or limit the networking.

It would be nice if a Kubernetes distro took a page out of the "serverless" playbook and just embedded a registry. Or maybe I should just use GHCR

filleokus··on The economics of a Postgres free tier
I guess that's true now with these "serverless" databases where compute and storage is separated in a neat way. Otherwise you either need to be ok with keeping the whole ≈vm alive, or have a really old school shared database setup alá 2004 web hosting… Maybe it's no problem, but feels kinda icky?

But regardless, I agree with you. I can spin up the cheapest fly.io machine for 2 usd / month, or some DO droplet/function, but neon and these guys are either free or 20. It always feels better paying a bit than nothing.

I know CockroachDB has some serverless > 0 (but cheap) pricing https://www.cockroachlabs.com/pricing/

filleokus··on Ask HN: How do browsers isolate internal audio from microphone input?
Not sure if it's the whole story, but the latest response in the linked Chrome ticket seems to indicate that the api's were used incorrectly by the author

> <he...@google.com>

> Status: Won't Fix (Intended Behavior)

> Looking at the sample in https://chrome-please-fix-your-audio.xyz, the issue seems to be that the constraints just aren't being passed correctly [...]

> If you supply the constraints within the audio block of the constraints, then it seems to work [...]

> See https://jsfiddle.net/40821ukc/4/ for an adapted version of https://chrome-please-fix-your-audio.xyz. I can repro the issue on the original page, not on that jsfiddle.

https://issues.chromium.org/issues/327472528#comment14

filleokus··on Region-specific Machines pricing
Interesting take.

At one side, the big cloud providers have their very fair profit margins. But I've always assumed that comes mostly from that they can, and do, charge extra for the beautiful software (not seldomly through ridiculous egress bandwidth pricing).

I'm skeptical of how much flexibility and optimisation the big cloud providers really get compared to a very large VPS provider?

All the compute products (different form of container hosting, serverless) are essentially still priced at CPU+RAM + a premium. Some additional abstraction for nouveau databases ("capacity" or "billing" units). Many of the highly abstracted offerings also have the scale-to-zero concept, while the provider still have to have physical capacity ready (to some extent).

Sure, you can probably crank out some percentage more utilisation, but that doesn't really matter for the 100+% price-add on the market bears.

But I of course have no data to back up my claims, just speculation.

filleokus··on Google Sheets ported its calculation worker from JavaScript to WasmGC
AFAIK I think Igalia is/was working on WasmGC support for Webkit/Javascript Core [0]. Not sure about the status though, it's tracked here I think [1]. It says it's updated earlier this month, but don't know of the progress made.

[0]: https://docs.webkit.org/Other/Contributor%20Meetings/Slides2... [1]: https://bugs.webkit.org/show_bug.cgi?id=247394

filleokus··on Denmark to charge $100 per cow in first carbon tax on farming
I saw some numbers about 30x and 10 years, but yeah it's definitely an externality that should be taxed.

But if we have two situations (A) where 1 tonne of fossil Co2 is added to the atmosphere, from e.g fossil fuel and (B) where 1 tonne of renewable methane is added, from e.g cows, it doesn't feel obvious to me how to compare them.

(A) will increase the Co2 PPM / have a GH effect for centuries, if not thousands of years. So on any human timescale it's essentially permanent. This has a pretty clear price in many tax frameworks.

(B) will have a 30x GH effect for 10 years.

What's the cost of externality (B) compared to (A)?

filleokus··on Denmark to charge $100 per cow in first carbon tax on farming
But the methane will oxidise down into CO2 in 8 years or something (according to Google). So essentially you pay the tax for the "GWP-bump"? It's resonable but maybe a bit hard (or at least non-obvious to me) to calculate fairly?
filleokus··on Denmark to charge $100 per cow in first carbon tax on farming
Yeah, but how is the "net" greenhouse effect calculated? Over a long time period it should be zero:

Co2 from atmosphere -> Consumed by grass / grain -> Cow eats -> Methane released in atmosphere -> Broken down to Co2

filleokus··on Denmark to charge $100 per cow in first carbon tax on farming
From where does the (extra) greenhouse gas emission come? The cow itself consume renewable feed (of course [1]). Is it the transportation of feed, production of clean water, random diesel tractor stuff? Maybe deforestation due to feed production?

[1]: Or is there feed where the carbon is taken from e.g petroleum somehow? Sugar is a hydrocarbon after all...

filleokus··on Loop: Open-source macOS window manager
Hmm. I interpreted "[not publishing source code] goes also against the GPLv3 license that you're using" as meaning that the author is "bound" by the license to provide the source code.

Of course, the terms in GPLv3 that puts limits on the licensee / user (like that derivate works must be GPLv3) are not relevant.

But section 6 in the license stating that the source code must be distributed still holds true. Either OP distributes the source code to people who have the binaries, or the work is not licensed under GPLv3?

filleokus··on Loop: Open-source macOS window manager
Maybe I'm missing something? But my assumption was that if I "acquired" (downloaded…) software licensed under e.g GPLv3, me as the user at least have the theoretical possibility to pursue the rights granted to me by the copyright holder in the license.

Of course the copyright holder can license new distributions under a different license (or different licenses at the same time). But are you saying that a copyright holder post hoc can revoke the license?

Wouldn't that be kinda weird? Suppose I buy some proprietary software (without any dependencies) because it's GPLv3, and then when I wanna exercise my rights, the sole copyright holder just goes "nah, when I think about it, this is actually not GPLv3".

(If you are correct, I guess that situation then might just be a civil case unrelated to the actual license where I essentially didn't get what I paid for)

filleokus··on The Backbone of Cybersecurity: Hardware Security Modules
> Operation Time

> RSA key length (bits) Average time (seconds)

> 1024 16

> 2048 124

> 3072 600

> 4096 ~1000

That must be a typo, that they mean milli seconds - right? Otherwise this seems too slow to do anything useful?

filleokus··on Private Cloud Compute: A new frontier for AI privacy in the cloud
Don't know if they use it (or if it would somehow weaken/break the privacy claims you cited), but Apple has an SDK called DeviceCheck[0].

Essentially, your server send a nonce which the client signs using a key pair derived from the Secure Enclave. The server can then verify the signature by an API provided by Apple's servers, and they respond whether it was signed by a Secure Enclave resident key or not.

I'm guessing this could be helpful to make it hard(er) to write a Python client.

[0]: https://developer.apple.com/documentation/devicecheck/establ...

filleokus··on Ultrasonic Coffee
As a Swede, I've never been so deprived of good coffee as when I was in Italy. So heavily "roasted" (more like burnt or torched) that no taste except maybe tar or char is left. Served either in ant sized cups or with milk to mask the horrendous taste.

No, the Nordic way of light (or medium) roasted coffee prepared in a drip brewer is the only way.

If this breaks into Italy, I will make sure to stay away from it ;-).

filleokus··on DuckDB Doesn't Need Data to Be a Database
How stable is the DuckDB file format nowadays? One annoying thing I've experienced (like a year ago) was when I upgraded the rust bindings that produced my DuckDB files, which then couldn't be opened with DBeaver because they hadn't released a new version yet.

I guess you could run into something similar with this solution?

filleokus··on Big data is dead (2023)
I think I've written about it here before, but I imported ≈1 TB of logs into DuckDB (which compressed it to fit in RAM of my laptop) and was done with my analysis before the data science team had even ingested everything into their spark cluster.

(On the other hand, I wouldn't really want the average business analyst walking around with all our customer data on their laptops all the time. And by the time you have a proper ACL system with audit logs and some nice way to share analyses that updates in real time as new data is ingested, the Big Data Solution™ probably have a lower TCO...)

filleokus··on Going Dark: The war on encryption is on the rise
> They need to understand that this miracle they want does not exist, too! > I just feel like you complain about the fact that I split my argument in 3 points and would rather have it stand on one point :-).

Which is your point 2). :-)

As succinctly as I can: I think my argument now is that 1), redefined by you (or reunderstood by me) during our conversations is not worth making.

Initially in my original post I interpreted 1) to mean that "since E2EE is used by 80+% of people today, it can't be made illegal" which I argued against (because I think this is false). Then you expanded upon it / I understood better to be "the technology exists, it can't be made to disappear", which I agree with - but think is a meaningless or naive way to look at legislation. Australia e.g famously enforced strict gun laws in the late 90's, telling legislators then "the first thing you need to understand is that guns will still exist" is true in a meaningless way that doesn't convey anything.

I think a reason why I felt the need to argue about this is that I think a lot of techies / HN people have a very black-and-white view on the legislative process, "if it's still technically possible it's not worth/possible to outlaw". Not saying you fall into this category though!

filleokus··on Going Dark: The war on encryption is on the rise
If the app is not available in the regular app stores and requires side loading, (or possibly rooting) the threshold is quite high.

> You just seem to keep repeating that legislation like this is dangerous

And you keep repeating that it's a problem for the legislators that E2EE technology exists, when it's not. If a person doesn't think this is ethically problematic, even only 10 caught sex offenders every year would be consider a win. Whether there are ways to circumvent the law or not (hint: there are always ways) is irrelevant.

> Meaning that "this law targets child pornography" makes no sense

It makes total sense! I can almost guarantee that there would be more than 0 child sexual abusers who will use WhatsApp post this law, and will be caught. The aim of this law is to limit the spread of CSAM, and if more than 0 people are caught then the legislators will have succeeded (to some extent).

Then when ChatControl 3.0 comes, targeting terrorists and organised crime within the same framework, again more than 0 will probably be caught. With ChatControl 15.0 worthless discussions online about legislation will be reported to employers and I will be caught ;-).

Therefore, as I wrote in my original post, we need to explain WHY this is dangerous and not worth it. The fact that "E2EE exists and you can't catch all bad guys" is a very weak argument.

filleokus··on Going Dark: The war on encryption is on the rise
I would assume those users were somehow highly motivated to use the app. "In this gang we only use An0m, install it if you wanna deal our drugs". I doubt any of my non-techie friends (or even my parents) would side load an app if it was required to text me.

WhatsApp has 2 billion users, Signal 40 million and An0m had 12k. Legislation like this will make E2EE communications unavailable to everyone but the most motivated techies and criminals.

filleokus··on Going Dark: The war on encryption is on the rise
New thread ;-). This is the cloudflare-is-a-big-man-in-the-middle-operation meme. When a site uses Cloudflare the traffic is encrypted between the visitor and Cloudflare, and then (possibly) re-encrypted between Cloudflare and the site.

As a large percentage of all internet traffic is going through Cloudflare, they could spy on us or let $GOV_AGENCY spy on us.

There is no evidence and Cloudflare is (or at least used to be) somewhat of a HN darling, so it's perhaps not so popular to say, but I wouldn't be hugely surprised if turned out PRISM 2.0 involved them.

https://en.wikipedia.org/wiki/PRISM

filleokus··on Going Dark: The war on encryption is on the rise
We've now passed the threading limit so I'll make this my last reply (maybe) :-).

I think we are in agreement that laws like this are bad and not something we want in our societies.

Where I think we might have differing views is that I consider the "it exists" argument as being really weak against a proposal like this. If someone truly believes that this will only be used for "good" purposes and that it has near-zero cost on society ("Honest people don't have anything to hide"), arguing that "criminals can still use illegal E2EE" is not fruitful.

"So what? We will probably catch many or even most criminals. They aren't known to be the sharpest tool in the toolbox, it's going to be cumbersome to use illegal Signal and criminals are lazy. And by the way, child sexual abuse material will also always exist, does that mean we can't make that illegal either??"

No law is ever 100% effective, and policy makers know that. They just have a different cost/benefit calculus, or don't see all the costs (or both).

filleokus··on Going Dark: The war on encryption is on the rise
I don't fully understand the talking about "scale" then. Scale is irrelevant to criminals. OpenPGP has been around longer than I've been alive, and OTR for soon 20 years. The fact that WhatsApp and iMessage has been per default for the last 10 does not push the needle for a motivated criminal.

The fact that it IS deployed at scale now is the "problem" and what will be "solved" by something like Chat Control. We would go back to how it was before E2EE became the norm.

What this legislation would allow is (even more) mass surveillance of normal people.

filleokus··on Going Dark: The war on encryption is on the rise
Hmm. Well yes, the crypto wars are over, and strong encryption is readily available. That will never disappear. But that's been the case for decades now. The change (and possibly what's driving these legislative initiatives) is that a huge chunk of communication is E2EE by default.

If WhatsApp/FB Messenger and iMessage added the government as a hidden group member to all our conversations tomorrow, and the day after that Signal was removed from the app marketplaces due to not complying, >99% of people would loose E2EE.

We would go back to how it was in the early 2000's, 99% of people used Yahoo/AOL/MSN "normally" and people like us would dabble with Jabber/XMPP off-the-record. In some ways it would be easier today, the cryptographic libraries and standards are better and invented. In other ways it would be much more difficult. Undoubtedly Ylva Johansson would be happy!

(I hate to bring up crypto currencies. But it's similar. It's impossible to make crypto currencies disappear, but it's fairly simple to make them marginalised and/or not effective as privacy tools by introducing stringent KYC measurements at the intersection between crypto currencies and the normal economy. Sure, you can always buy drugs on the darkweb and food at anarcho capitalist festivals, but it will never have any real privacy impact on real people in their everyday lives.)

filleokus··on Going Dark: The war on encryption is on the rise
I strongly agree with 2) and 3), but sadly I think 1) is overplaying our collective hand.

I would guess that the largest deployments of end-to-end-encryption today is Whatsapp and iMessage by a quite large margin. E2EE for "real people" is provided by the grace of two massive publicly traded companies who have to follow local regulation. If Apple complies [1] with dubious requirements in China, I wouldn't bet against them doing it elsewhere either.

Sure, we have Signal, but how many normal users would start delving into side loading if the application simply was banned and not allowed on the marketplaces? We can always use PGP-over-whatever, but that's an argument for 3) - not 1).

I think politicians / police (and honestly many normal people) believe that the government have the right to do lawful interception of private communication and see E2EE as a step to far. The US has been wiretapping phones for a century already.

We as privacy arguers have a pedagogical challenge of explaining why regulation like this is bad and not equivalent to 1930's style phone wiretapping.

[1]: Chinese iOS users have their iCloud data (that for the vast majority includes the decryption keys) on Chinese servers, subject to the Chinese legal system. For the rest of the world the situation is the same, but s/China/U.S, which is arguably problematic as well.

filleokus··on 4K 1000Hz Monitor Exhibited
I've also heard some handwavey explanation about how certain parts of the panel manufacturing process is essentially separated by DPI. That you can reuse certain aspects of a ≈ 100 DPI manufacturing line across varying refresh rates (and physical sizes).

Then those line can probably serve a large part of the gaming / normie market, while the ≈ 220 DPI (e.g 5K @ 27") have a very small market.

filleokus··on GxHash is a fast and robust non-cryptographic hashing algorithm
GP meant that all collisions contained 's - a possessive apostrophe (if I remember my grammar correctly?). oncogene's, toadstool's, Carib's.

The /usr/share/dict/words on my macOS machine doesn't seem to have any 's words in it.

filleokus··on Telegram has launched a pretty intense campaign to malign Signal as insecure
I know a quite a lot of people (including me) who use it as a kinda RSS reader to keep up to date with Russian cyber security threat actors.
← PreviousPage 3 of 23Next →