HNHacker News
TopNewBestAskShowJobs

asveikau

15,237 karma · joined September 5, 2012

submissionscomments
asveikau··on Police are using AI to pull over drivers for 'suspicious' behavior
I think there is a rise of "this is legal because we used a computer to do it." A HN audience will not appreciate this sentiment, but honestly, a company should not provide tools to do this because they should have a filter. They should say "I won't provide tools to break constitutional rights".

Somewhere along the way, tech stopped asking questions like that, and now, people are hostile to the suggestion that the question should be asked.

asveikau··on Wikipedia's Cleanup of AI-Generated Content
The issue is when people flag neutral information as having a political bias.
asveikau··on The Mythical Man-Month in the Age of AI
Unpopular opinion for this site: Most true facts are not affected by the existence of AI. Most things in life do not need adjustment for "the age of AI".

For decades, eery tech boom cycle is filled with people who don't understand the mythical man month, or that LOC is not a productivity metric. The AI boom is no exception. AI has not invalidated this.

asveikau··on In Leaked Audio, Flock CEO Tells Cops Backlash to the Company Isn't Real
Yes, but in the case of prosecutors it can be very messed up. When someone is wrongfully convicted and there is very unambiguous exonerating evidence, prosecutors will fight overturning convictions.

It's bad enough to wrongfully send someone to prison, but to keep fighting for wrongful imprisonment when it's shown they made a mistake is downright evil.

asveikau··on In Leaked Audio, Flock CEO Tells Cops Backlash to the Company Isn't Real
This attitude is also very common for law enforcement or prosecutors.
asveikau··on A 10-year-old CNN outperforms frontier LLMs for skin cancer diagnosis
I agree with what you said above, but not this. Jev was discussed "all over" by people chasing hype. I think that is actually a narrow set of people.
asveikau··on OBS Studio 33
I'm sure that obs can export to something that can be used to stream to rumble.
asveikau··on OBS Studio 33
How was latency for all of this?
asveikau··on Docker is weird
It's because from a kernel perspective this part literally needs root:

> just trying to execute the command within the child container's filesystem

There is no such thing as changing your root filesystem or your uid from an an unprivileged process. There is no "just do it in user mode" either.

So they must have a root process somewhere doing it for you.

I just started googling, apparently for podman the setuid root binary is newuidmap

asveikau··on Docker is weird
That cannot be true. You need to tell the kernel about the namespaces, cgroups, and the new root. Otherwise it's not a container, and there is no reason or mechanism for a child proc to respect it. And if you let any old unprivileged process do this, you open up security problems.
asveikau··on [dead]
Based on the headline I expected to have opinions about this, but this is about Japanese elections, and I don't know anything about the political factions of Japan. These things don't always translate exactly across international lines.
asveikau··on Docker is weird
I would guess that they do that by having a root daemon which authenticates requests from a unix domain socket, possibly using SO_PEERCRED.

Or a setuid binary

asveikau··on Docker is weird
Care to give some examples?

A lot of the more important syscalls involved do require root.

asveikau··on Docker is weird
If you're old enough to remember some techniques that people used before docker (like chroot(2)-ing daemons for security), the "containers aren't real" observations are not really new info. When containers came on the scene it felt like an outgrowth or evolution of that technique, which was extremely crude and simple.
asveikau··on Docker is weird
I think the only thing in the "weird things" list that don't necessarily apply to another container system on linux would be the go-specific issues.
asveikau··on eMudhra to exit global SSL/TLS certificate business
> using "some company in India" made it rather dismissive of the warning, actively taking the company's side(the bad "no keys were issued" defense), whereas just saying "some idiots" had it generally approve the message.

But you know your audience better than ChatGPT does. Whether people react better, worse, or indifferently to the India angle will depend on the biases of the recipient.

asveikau··on Every AI model has the same "controversial" take on humans
Yeah I came here to say something like this. The AI doesn't have a controversial opinion. It responds to text with related text. You gave it some text about controversial opinions and it returned some text.
asveikau··on Eight Bits of Trust: eMudhra's CA keys cracked in milliseconds
I guess everybody is lying when they say to patch nodejs and python's certifi in response to this issue. They should listen to you instead about how it doesn't work that way, and stay vulnerable to MITM.
asveikau··on OpenShell – A safe, private runtime for fleets of autonomous AI agents
> It wouldn't let me build any features that track/audit process executions

Sounds like waitpid(2) is dangerous!

asveikau··on eMudhra intermediate CA compromise
Update: The official patches for both OpenBSD and FreeBSD are now posted, no need for manual updates anymore.
asveikau··on Eight Bits of Trust: eMudhra's CA keys cracked in milliseconds
Update: The OpenBSD errata just dropped.
asveikau··on eMudhra to exit global SSL/TLS certificate business
I have explained it to a few normies.

I start by saying this is about the tech behind the S in HTTPS. I say that the main point of this is that when you talk to foo.com, it has to prove that the people behind foo.com are on the other end, so it's safe to send them your encrypted password or whatever.

Then I say a bit about how the system works by having a small number of trustworthy companies vouch for other people's servers. This doesn't need too much technical detail to get the point across.

Then I say one of those "trustworthy" companies put an encryption key of 4, where it's supposed to be a long number that's hard to crack. This makes the verification easy to forge.

Then I say that everybody who runs a server needs to remove this company's bad key, and Windows or Mac or your phone needs updates too. And that your web browser probably already quietly removed it, which removes a lot of the risk.

I feel like when I explain it this way people get it.

asveikau··on Eight Bits of Trust: eMudhra's CA keys cracked in milliseconds
> They don't. They use the OS certificate store,

It appears to me they do, and that is why everybody is saying to patch your python, go, and node infrastructure separately from /etc/ssl.

> which on all platforms has already had this certificate revoked from it.

I know of two systems which this doesn't seem to be the case as of yesterday.

1. I manually edited cert.pm on OpenBSD. I don't see the OpenBSD errata yet.

2. As of yesterday Mac safari was still vulnerable when I tried.

asveikau··on GitLab critical patch release
The module in that diff being called "onig" is quite unfortunate, the error codes look like a slur.
asveikau··on Eight Bits of Trust: eMudhra's CA keys cracked in milliseconds
This issue is highlighting a few things for me:

* It's absurd that Python, Go, or Node need their own certificate stores that need separate updates. (Probably OK for browsers because they seem very on top of updating them without user intervention.)

* This issue is especially bad when you consider deployment through containers, docker, etc. More parallel copies of vulnerable certificate stores. With docker volumes you could mount the host's certificates into the containers and get "update once" behavior but I doubt people are doing this.

* It wouldn't be too hard to implement CRLs against the OpenSSL certs as a script that runs from cron. I might end up doing this.

asveikau··on eMudhra intermediate CA compromise
In case anybody else is curious, since the workarounds are only offered for Linux and NetBSD, I was able to fix this on FreeBSD with:

    sudo certctl untrust /usr/share/certs/trusted/emSign_ECC_Root_CA_-_G3.pem
And on OpenBSD by editing /etc/ssl/cert.pem, removing everything starting from the line:

    === /C=IN/OU=emSign PKI/O=eMudhra Technologies Limited/CN=emSign ECC Root CA - G3
Up to but not including this line:

    === /C=IN/OU=emSign PKI/O=eMudhra Technologies Limited/CN=emSign Root CA - G1
asveikau··on OpenBSD 8.0 release will have inteldrm/radeondrm/amdgpu drivers from Linux 6.18
You tell me I'm part of a swarm of people like me, but we haven't met, and I don't think you have enough information to know what a person like me is.

On the other hand my life experience is often that people who accuse others of doing activities in secret, and being dishonest about it, that is often, but I will give you not always, indicative of where their own head is at.

asveikau··on OpenBSD 8.0 release will have inteldrm/radeondrm/amdgpu drivers from Linux 6.18
Because you assumed they would direct an LLM to do it and lie about it. Seems like projection to me.
asveikau··on OpenBSD 8.0 release will have inteldrm/radeondrm/amdgpu drivers from Linux 6.18
Just because you don't know how to do it doesn't mean they don't. As I said, the same people have been doing this for years, long before LLMs. They need to re-sync with upstream periodically so that's why you see announcements of new stuff.
asveikau··on Android apps running native on macOS
I think you meant to say:

It's not just overused — it's a cliche now.

← PreviousPage 2 of 34Next →