kevinrineer

Is your org at the point where it would make sense just to upstream patches for FOSS projects and pull rather than patch and push to just your infra?

Just a matter of curiosity for me as my career progresses, where I think that is the ideal, but I don't work in a heavily privatized nor regulated space at the moment of posting.

I did read your latest blog post looking for an answer, but it wasn't clear to me from that reading.

So interesting to see Darpa represented in different labs. I wouldn't have bother to look for that interesting tidbit otherwise, so thanks for that.

Know any harnesses that explicitly don't have a "yolo" mode?

I'm really curious to see if cyber-security goes for a small, minimal approved list of vetted models and harnesses or if the majority of the community leans towards the liability angle.

One could compare it to the immediacy of an actor's overnight success shaped career.

Most business security cameras are powered by something and are most likely not tied to the emergency power lines.

This correction is not an endorsement.

Thanks for the rec. That repo is young, but the owner seems quite experienced.

I'll happily try it.

I am curious about your confidence for NFS for the nix store.

My thoughts, ignoring my technical disagreements, would be that you'd want to setup clients to have a specific set of software per their government role. A nix store could expose more software than you planned on supporting (e.g. forgotten software or licensed software).

> Sextant: a graphical interface for NixOS management as an alternative to Intune, by Bram Buijs [1]

Now that's a fairly interesting project and a way to actually bring technology like NixOS to the administrators in the way they expect. From my experience (sample size one, yes) there are more systems administrators that use web interfaces for management than DevOps IaC folks. This would be a good bridging solution if it is a successful project.

[1] - https://members.dawo.community/en/blog/blog-1/dawo-nixos-wor...

There's also the risk that VS Code is a "trusted" application in many enterprises because developers force it to be. VS Code's node runtime (and plugin system) executing somewhat arbitrary Javascript means that any dev servers become vulnerable to exploits you otherwise might not have prepared to defend against [1].

Otherwise, you could see Javascript running on a server and instantly know something was odd, depending on the server.

[1] - https://www.darktrace.com/blog/darktrace-identifies-campaign...

I guess I'm poor then because my homelab has nodes with these specs.

Wonderful link. Just commenting to add: in the first video, Grant mentioned it would be a trilogy series.

> me: That's enough setup, you're exactly right.

So that's where in the training set current models get that phrase! /j

AlphaGo, Watson, Stockfish, Eliza. I can name so many of the interesting stuff and each new LLM just comes and goes.

To be honest, I feel that the ceiling is higher in Windows. I just today dealt with a faulty HP Printer driver that caused wuauserv (Windows Update) to report that I had to reboot every day. Finding which files were corrupted in `C:\Windows\SoftwareDistribution\` was not an easy task.

I've also this prior week fixed double digit Wi-Fi issues, my institution is dealing with the deprecation of MSCHAPv2, and so on.

Not going to say MacOS is free of issues either - from my experience with screen flickers / blackouts on full screen video playback with quick transitions.

In my experience this year, the grass is not greener in any desktop OS.

Why are so many people invoking Thundering Herd without the RCA from the vendors? I just find it so intriguing.

> According to The Information, OpenAI has reportedly purchased tens of thousands of Mac minis and Mac Studios in recent months for reinforcement learning and training computer-use agents.

I do hope they verified this claim and didn't just read the headline. I'm not interested in subscribing to that source to find out unfortunately.

> I don't even have a pricing page, i try to see how much i get can from each client, from $5/mo to $100/mo

That seems ethical. Vibes based fees.

I too like running the same kernel for months on end. Just kidding, I restart periodically, especially when there's a kernel patch.

To those who worship uptime, how many kmods do you have loaded? Must be nice not subscribing to openwall/NVD/CISA/cve.org

I'm still rocking my nvidia 2060, which I had purchased for $400 at the time.

I struggle to imagine purchasing multiple 1k+ cards on my own dime.

I'm a bit surprised that this kind of incredibly engineered hyperconverged rack system isn't doing some form of liquid cooling.

Not only that, but do you really want to be handling the contract negotiations, sales, business taxes, insurance, and everything else?

Being a sole-proprietorship means doing things you aren't skilled in and, most likely, not interested in doing. When I did it, I had a horrid work-life balance and I would not recommend it.

The referenced gap is the supposed numeracy of developers 5 years ago and developers today, whether they be junior or senior developers.

There is a strong downward trend toward decreasing numeracy in high school graduates and it has existed, supposedly, since 2013. This observation seems to be extrapolated to developers in the quote you reference.

For an MDM managed computer (JAMF I know for sure), it can be configured that way per a company policy. I am not 100% sure of the answer for a computer not managed by JAMF as I have not experienced a forced update while using a non-MDM managed Mac in ~1.5 years of using a pre-owned M1.

You just don't have to think about it too hard:

jjplace/jjhub/codetown, whatever. Doesn't matter.

Names don't matter that much for brands. Names just have to be simple enough to remember (ideally two syllables or less). What the heck does Nike mean, for example? Boeing is just someone's name. Microsoft is just two words smashed together. A brand's name literally doesn't matter.

And, interestingly for Oxide, rack scale solutions from HW vendors might not be suited for your company's workload. Their solution isn't all-in on AI, from what I understand.

You mean you don't have that issue anymore? I'm jealous.

Agreed. Its more reminiscent of Cloudstack or Openstack from what I gather. I'm thinking of Jetstream2, but for you buy it rather than rent some of it with an NSF budget.

> ... as a synonym for "incident report"

People should stop using it as a synonym, then. The Latin effectively means "after death", meaning its a poor synonym for "what happened wrong recently".

> If the estimator falsely declares someone an adult, is Discord legally liable?

Not until a court case on the topic gives us precedent.

I found myself really frustrated trying to use MacOS at work, because I'm a heavy user of virtual desktops. Turns out, I couldn't find a way to disable animations to switch between virtual desktops on MacOS. If there is a way, I'd be surprised.

Shortening the animation to minimum was not sufficient for my preference.