In my experience you just slap minikube or k3s on your dev machine, and treat it as any other environment. Argo, helm, kustomize, whatever can all work against a local single-node cluster just fine. It takes some effort to make sure your configs are overridable per-environment, but it’s worth doing. (And something you’re hopefully doing anyway if you’re doing any kind of integration/test environment.)

It also requires that each of your services can scale down as well as they can scale up… none of them should be so huge that your whole cluster can’t fit on a single machine, if you’re just simulating one “request” at a time. (Single instances of everything, don’t reserve memory, etc.) There’s use cases where this isn’t practical, but in most cases it’s very doable.

I certainly know how to debug each of the services in my environment, but how do you step-through debug a single request across services? Like, if service A make a gRPC call to service B, are you saying you can “step into” the call from A and your debugger is able to break on the corresponding call in B? And frames from the call in A are there in a logical “stack” from the breakpoint in B?

(Honest question… if such a workflow is possible I’d love to hear about it. Debugging just a single service at a time in isolation is boring and obvious, but if there’s something I’m missing I’d be really curious.)

If you’re gonna let them just build up over time, why not just open the app and see the latest videos in there?

Time sensitive stuff I… kinda get. But I would get stressed out if that was actually the norm. I really don’t like the idea of “engage now or you’ll miss out” in my entertainment.

I can see that being useful if it’s important to you to start watching someone’s videos within minutes of them posting it, but I’ve never understood why that’s desirable for anyone.

To me, I watch YouTube when I have some time to do so and make the active decision to open the app… then let me know about which of my subscriptions have recent videos. I just can’t imagine being in the middle of something else and dropping everything because someone posted a video. But different people are different I guess.

I was curious if there were any interesting bit-masking patterns in ASCII for vowels that could be exploited, so I shamelessly asked ChatGPT and it gave a pretty nice insight:

- Setting bit 5 high forces lowercase for the input letter

- masking with `& 31` gives an index from 0-25 for the input letter

Then you can the `bt` instruction (in x86_64) to bit-test against the set of bits for a,e,i,o,u (after lowercasing) and return whether it matches, in a single instruction.

It came up with this, which I thought was pretty nice: https://godbolt.org/z/KjMdz99be

I'm sure there's other cool ways to test multiple vowels at once using AVX2 or AVX-512, I didn't really get that far. I just thought the bit-test trick was pretty sweet.

Chat transcript is here (it failed pretty spectacularly the first couple times, tripping over AT&T syntax and getting an off-by-one error, but still pretty good) https://chatgpt.com/share/684c8b39-a9c4-8012-8bb6-74e1f8b6d0...