jacobgold

Can't complain too much about 86 years but I'll miss him. I worry how it'll feel to watch him now.

Whether they can compete on decision models or not, TypeSafe showed that a lot of the market had missed something important. With this much money, they have a lot more chances to discover other important things that are missing.

Cross posting to Bluesky too would be cool, and reach even more people. There are many ways to post, but lots of people use https://typefully.com

Alright, so you didn't implement it. And you're describing a "pacemaker task on an event loop"? That doesn't sound like a Go program at all, where we don't have a single event loop to measure and we have built-in runtime metrics.

Just seems kinda weird to give advice on forums when you're not speaking from actual relevant experience. I'm getting meat proxy answers and confident replies from people who've never done what they're suggesting others do.

Your Go binary was being swapped out due to memory pressure, causing latency, and using mlock was your fix for this problem?

Or did someone else do this work and you assumed it was effective? Or did you personally implement this change and then validate that the before/after metrics improved as a result?

I didn't claim it was hard or risky.

Have you actually done this yourself for Go programs that you've deployed into production? If so, what were the circumstances? Was it your solution to latency spikes?

I see that you've rewritten your AI generated reply there. So when have you used mlock in production for Go services, the way you're advising others to do? What were the circumstances?

FYI: This `delamon` account seems to be meat proxying us. Just relaying garbage AI answers without understanding his own words. Doubt he's ever used mlock himself for anything.

His reply below was presumably killed by mods for doing this: https://news.ycombinator.com/item?id=49970841

We're trying to solve a problem: we don't want huge latency spikes caused by swapping memory to disk. The only good solution to that is to have sufficient RAM. It's that simple.

Introducing mlock is operationally annoying and trying to solve the wrong problem. It requires elevated container permissions, it's not something your SRE team is going to expect, few programs do it, and it has all kinds of technical implications and interactions with kernel OOM system, forking, the Go GC, etc.

And nothing about mlock solves the problem of having insufficient RAM for your workload.

There are very specific scenarios where mlock is exactly the right solution but the typical Go HTTP service is not one of them. The KISS principle applies.

No, do not try to use mlock to avoid swapping, except in very specific scenarios. That way lies madness.

To be more explicit about my advice: disable swap and ensure you have sufficient memory for your workload.

Your 14 MB Go binary is not the source of your memory pressure. Your memory pressure is coming from the allocated anonymous pages supporting your programs data structures on the heap.

Have sufficient memory for your workload, monitor it, and let the Go GC and kernel manage memory normally. This will work exactly as you'd hope.

"It hurts when I do this"

"Stop doing that"

If you care about latency, disable swap. System wide or for the specific the cgroup.

You're inferring things I did not actually imply.

In Oct 2026, if you're optimizing for productivity, then yes, you probably should just use both Claude and Codex in a GUI agent multiplexer and forget about everything else.

If you care about other things, then have fun, no one is stopping you. I'd disagree with you that using alternatives mean you take more "care" and have more "pride" in your tools, but it's hardly worth arguing over.

The important thing is just to a GUI agent multiplexer, they're all mostly doing the same kind of thing. IMHO it's very important to have the official claude and codex harnesses (not just models). I also like having each session in a isolated container (or VM) so the agent can go wild and run in yolo mode.

I usually recommend Conductor to most people. Personally, I use the one I built, but it's got a few ergonomic issues for most people which I still need to fix.

And you're probably more productive than the people using Pi. Although you'd be even better off using a GUI agent multiplexer of some kind rather than juggling terminal tabs.

I love open source. I love the terminal. I spent the last 20+ years in a terminal w/vim every single day, and then claude/codex TUIs, and yet I care more about my own productivity so I don't use any of that now.

It's very possible to predict that certain technologies will be created. It's just very difficult to predict precisely when.

For example, humans will almost certainly be able to re-create the human brain in some kind of hardware+software computer. We just don't know when. It might be much sooner or much later than you'd expect.

Merely being skeptical of progress isn't a convincing answer to the AI x-risk argument.

I feel exactly the same way about the kind of LLM-based AI we have today.

But the AI x-risk argument is about what happens when we have "real" AI that can actually think like humans but a million times faster.

There are no real experts on this and yet it does seem worth worrying about in advance.

Assuming OP is a people, they're included, but I get your point.

And it's great they did find personal connection on Stack Overflow, but it was infamous for its hostility to noobs. It's sad to say, but Stack Overflow very possibly discouraged more people than it encouraged in this way.

I'd guess that both IRC and Discord have a much better track record of personal connection and helping people, and they're still thriving.

> An LLM would have done exactly what I wanted back then. But would it have given me what I needed? At the time, I needed to know somebody cared enough to help me.

Were StackOverflow, Google, and LLMs not all created by people who cared enough to help you?

There are much better places on the internet to find personal connection than StackOverflow ever was.

Personally, I'd like to hear the full pitch, and I suspect many others would too. If the ideas are interesting enough for you to dedicate the next N years of your life to, I'm probably willing to spend 10 minutes reading the blog post. Particularly from you, a known-good writer.

So I recommend it

Testing something, can you guys post useless replies like "cool man" or "...okay?"

> I really enjoyed the work, and for a while I told myself it was relatively harmless, but over time God forced me to confront the reality that the main impact of these tools will be to accelerate the design of a new breed of AI chips...

There's no arguing with someone who claims God forced them to do something, but I do wish this person good luck in whatever they do next.

I like and adopted the term "agent multiplexer" for these.

Just can't warm up to "ADE" (never even liked "IDE") and "agent orchestrator" and "control plane" are just not specific enough to stick.

At this point, no one seems capable of keeping a large database safe. I assume all medical and biographical information that exists is in the hands of the major state actors.

China hacked 22.1 million records of US government employees:

https://en.wikipedia.org/wiki/2015_Office_of_Personnel_Manag...

These counter-launches are starting to seem kind of tacky and boring. Just launch on your own scheduled guys.

I use the other 50% of my $200/mo Claude subscription by having Fable run Opus subagents for a lot of work. That way I don't have to deal with Opus directly.

I know very little about this topic, but this ultimately helps force a real solution to our prescription drug cost problem?

Most people won't be honest with you about these things because there's very little incentive. People tend to react defensively like you are here. But it might actually help someone reading this to know the truth.

Maybe it helps you to know that early in my career I was also sometimes promoted into roles that were beyond my abilities.

I genuinely don't want to upset anyone, but as someone with 20+ years of engineering and management experience, I think it might help some people to hear what is most likely really going on here.

This is probably an example of the Peter principle:

The Peter principle is a concept in management developed by Laurence J. Peter which observes that people in a hierarchy tend to rise to "a level of respective incompetence": employees are promoted based on their success in previous jobs until they reach a level at which they are no longer competent, as skills in one job do not necessarily translate to another.

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

There's a huge amount of title inflation going on in the industry. We're now seeing the "Principal Engineer" title being given to people with 3 years of professional experience.

Unfortunately, this is simply not enough time to be very good at all, no matter how much raw talent you have. It takes a decade of experience just to know how to manage yourself effectively, let alone mentor others and lead large projects.