xgbi

Way more technical, way smaller. HN is a lot more wide in terms of audience and subjects, Lobsters is mainly programming.

I love it.

How is Rocketlab valued 57B? They made $500M of revenue in 2025. This is 100x their entire balance sheet.

One of the founding books that really blew my mind and drove me on the path of software and hacking.

I was 17 in 2005 and discovered it by chance, and I’ve been binging on hard sf since then. Matrix and this were really transformative for me.

Also, for the longest of times I thought lobste.rs was a reference to this book :-)

Charles has very interesting takes on the modern world on his blog. I still read it with great passion.

What are you talking about ?

It says that the parental settings (when enabled!) are just letting children do whatever they want by default:

- buying overpriced objects - chat without any restriction online - play without interruption for long time

I think the first one is probably the most poignant: piping children into disguised gambling addiction by default seems like a major fault. Borderline illegal, if you ask me.

It looks a lot like a phony feature "let's add a parental control, it will make people feel like we're trustworthy and bring back more revenue. And please don't disable ingame purchases by default, this is our cash cow".

I just can't.

We shouldn't be saying "if an individual chooses to do so, we could achieve political harmony".

At what point does the government says: twitter/X has attained a critical mass and should adhere to strict political neutrality and enforce net-neutral policies, otherwise be dismantled ? I know, your current US government benefits from this. But in general, a government should be working towards neutrality. Otherwise this is a power grab.

Apart from very specific people, that want to manipulate masses, having such a great power over opinion by manipulating what people see should be strictly controlled.

I want to subscribe to your AI wars news please!

Joke aside, the strategic choices here and there hint at the blood lust of all other actors to dethrone Nvidia, it’s fascinating.

She's her own boss and shares her office space with 4 other people in medical space, no shadow IT there.

Since her desk is far from the internet router, I added this little guy for her to have less cables and allow more connectivity.

My wife’s work WiFi is handled by a gl.inet 150 (https://www.gl-inet.com/products/gl-ar150/) which is tucked behind her desk since at least 2019. Vanilla openwrt on it, provides WiFi from an Ethernet slot in the wall.

Uptime is in years, it’s invisible and chugs along without visible power draw. All her devices connect to it, including her Cisco voip phone. It autossh to my ovh server with remote port forward for remote admin. Cost me 15€ in 2016.

Same, I use LLMs to figure out the correct options to pass in the AZ or the AWS CLI, or some low-key things. I still code on my own.

But our management has drunk the Kool Aid and has now everybody obliged to use Copilot or other LLM assists.

Cause the codebase wasn't in my scope originally and I had to review in emergency due to a regression in production. I took the time to understand the issue at hand and why the code had to change.

To be clear, the guy moved back a Docker image from being non-root (user 1000), to reusing a root user and `exec su` into the user after doing some root things in the entrypoint. The only issue is that when looking at the previous commit, you could see that the K8S deployment using this image wrongly changed the userId to be 1000 instead of 1001.

But since the coding guy didn't take the time to take a cursory look at why working things started to not work, he asked the LLM "I need to change the owner of some files so that they are 1001" and the LLM happily obliged by using the most convoluted way (about 100 lines of code change).

The actual fix I suggested was:

    securityContext:
  -    runAsUser: 1000
  +    runAsUser: 1001

Rant mode on.

For the second time of the week this morning, I spent 45 min reviewing a merge request where the guy has no idea what he did, didn’t test, and let the llm hallucinate a very bad solution to a simple problem.

He just had to read the previous commit, which introduced the bug, and think about it for 1min.

We are creating young people that have a very limited attention span, have no incentive to think about things, and have very pleasing metrics on the dora scale. When asked what their code is doing, they just don’t know. They can’t event explain the choices they made.

Honestly I think AI is just a very very sharp knife. We’re going to regret this just like regretting the mass offshoring in the 2000s.

Meta: github is now requiring a login to see gists?

Hey robinhood, any feedback on Talos?

We've been using Talos for our internal clusters for a while, but with quite small ones (3 kube node, 5 worker nodes).

Upgrading has been generally a non event, and we're quite happy with them.

How do you deploy Thanos ? In one of the clusters ?

I have so many questions, it is a very good article!

My most important one is this: can I build a distributed k8s cluster with this?

I mean having fly machines in Europe, US and Asia acting as a solid k8s cluster and letting the kube scheduler do its job?

If yes then it is better than what the current cloud offerings are, with their region-based implementation.

My second question is obviously how is the storage handled when my workload migrates from the US to Europe: so I still profit from NVME speeds? Is it replicated synchronously?

Last but not least: does it support RWM semantics?

If all the answers are yes, kudos, you just solved many folk’s problems.

Stellar article, as usual.

Can't you simply `proxy_pass` the traffic with any load balancer or reverse proxy (that you probably have anyways if you use TLS)?

From what I read somewhere, Tesla was able to do that because they have remote ssh capability.

In at least one instance, they fixed the cars manually by running a massive remote command on all cars after a messed up update: https://lobste.rs/s/v42zil/former_tesla_employee_ssh_d_as_ma...

I wouldn’t call that very reliable , but they indeed do it regularly

Reading your comment, my first thought was "how entitled??", but I see that this meeting was set up by the DO founders? Aoutch.

Say what you want about docker compose, but when I see the amount of scaffolding necessary for this, with so many catch words like butane or ignition, I’m happy with my good ol’ docker composé file where everything is neatly organized. In one glance I can see what is deployed, depends on what and what net / volume is used.

For simple projects, it’s hard to beat

I need to store on Azure for DRP reasons: we would DRP to azure and need the bandwidth.

Also, storing the ZFS snapshots on Blob storage would still require us to retrieve the entirety of the 80TB before being able to use it. I need native ZFS at Blobstore-competitive prices

We are trying to use it for large backups of a production item, and it has not been a complete smooth ride all along.

We have many files (millions) and lots of churn over ~80Tb total.

Kopia has exhibited some issues:

- takes about 120GB (!) of ram to perform regular maintenance & takes about 5hrs to do so. There are ideas floating around to cherry pick the large inefficiencies in the GC code but it’s yet to be worked on. I’ll try to have a internship accepted to work on this in my company.

- there’s a good activity on the repository but the releases are not quick to come and the PRs are not very fast to be examined

- the local cache gets enormous and if we try to saddle it, we have huge download spikes (>10% of repo size) during maintenance. Same as above: pb is acknowledged but yet to be solved

- the documentation is very S3 centric, and we discovered too late that the tiered backup (long term files go into cold storage on s3) is only supported on S3, while we use azure. We contributed a PR to implement it in June, yet to be merged (see point 2)

So, not too bad, especially for a small-ish project maintained by mainly one person (from the looks of my interactions on slack and seeing the commit log). The maintainer is easy to reach and will answer, but external prs are slow. If I could use zfs cheaply on azure via s3, I’d use it over kopia, but as of now, it works.

Ah yes, rituximab, the R in R-CHOP. Welcome to the club!

The very first infusion in my case triggered such a nasty allergic reaction that they settled on a 24h infusion duration, thank god I had a port. Fun times

Story time.. I coded a similar thing when I started working after graduation for a DTV software company; we needed to have an IPC on Linux and I whipped up some very crude equivalent of protobuf (which I didn’t know about), based on an RLE lib i stumbled upon and without any form of discovery.

It was circa 2009, I had only been exposed to plain text protocols and didn’t know about json. In hindsight, we might’ve been better using standard dbus, of protobuf, but I was a rookie and it provided the performance we needed (for DTV metadata).

I’m happy to see that these can still thrive, and I just recently figured out that discovery is a net multiplier in these projects; doing this really proved me that any problem is solvable if you can have some time to think about it and can prototype.

I long for these moments now, I feel like nearly all computing issues have been solved and we are now just plumbers, connecting libraries and software modules through config files instead of building things.

The topic is interesting and the paper looks well sourced, but I find the title to be quite the mouthful!

Do you think it could be adapted so that we can use an rpi and pwm on a gpio to have a rx/tx without the RF component?

I don’t understand how you can spend so much. Even with inflation we (family of 4, cooling all meals at home) are at less than 800€/$ per month.

And we eat meat nearly all meals.

I've been using Asahi based on Arch [...] and should mention that the experience has been stellar overall.

I fully plan on moving entirely to Asahi Fedora [...] Fedora has a stronger team and less issues

Why do you plan to change if the experience has been stellar? You mention issues, what kind did you encounter?

I'm sorry if this feels rude, but I read your comment as a praise to Arch and then took an 180° turn to "I will switch to fedora ASAP".

Anyone with experience scaling Prometheus horizontally ? We are reaching the limits of our instance, memory and cpu wise, and I’m yet to choose between scaling it myself with sharding or using thanos/Victoria/cortex.

The form factor is way bigger than a NUC that fits on a vga mount!

why not both?

and check your logs to see who is not complying

I had an interesting issue with that the other day:

Someone on my company thought it good to launch background process from within a cron, and start it with `&ˋ at the end, so that it detached from the cron and goes as child to init .

Right, but the program opens std out and dumps its output in it. Guess what, the fd handling the std out ends up in a temp file which is deleted by cron, but still held by the program.

I ended up with a full / in all production servers, with no culprit when running ˋdu`. Only running ˋlsof |grep deleted` did I find these huge temporary files.

Killing the process and switching to a saner systemd service was the savior.