Viewing profile — robmccoll

Non-nullable reference types at the language level? Null coalescing operator? Safe navigation accessors? Record composition?

DXM may or may not suppress coughing relative to placebo - the study cited here appears to be have been written entirely by authors from drug companies, so perhaps there is some bi…

Not totally accurate - there are a handful of foaming agents and surfactants that are mixed and matched to make shampoos, so really it's nearly the same except that no one has ever…

Because I associate that style with "I have little to say, but I will say it in as many words as I possibly can"

What does the actual startup latency look like? Does it depend on the size of the resulting image?

The problem I have with Kunernetes in general is that everything is slow and configuring everything seems needlessly complex in common use cases.

For reference, a PlayStation 5 is $600-650 for the base models (lower performance than Steam Machine) and $900 for the Pro model (likely higher performance). I know this is a PC an…

That's interesting. I gave nearly the same task to Gemma4 31b as a test yesterday. Write a symbolic math engine in Typescript that can perform evaluation and simple expression redu…

I just don't want to have to have a pipeline going in order to fully occupy my time. I don't want to wait on the model to review the prompt, read the parts of the codebase indicate…

Yes - I want that and dramatically faster. Newer models don't seem to need any more or less guidance and iteration, so let's make the time-to-wrong-answer as short as possible.

Musk sold Twitter into xAI which he then sold into SpaceX as a financial engineering effort to lessen the impact of massive debts and cash burn. The IPO and some clever structuring…

My Nissan Ariya checks nearly every box. The controls for the AC have dedicated buttons, but they are capacitive, not tactile. Anyway, love the car. It's a real shame that they are…

Making these claims on a 2B parameter model seems a bit like seeing linear scalability from 1 to 4 cores and then assuming 256 cores will give you a 256x speedup. Or demonstrating …

* * *

Maybe - let's assume that we could simulate a human brain to a high level of fidelity starting from a known good state and emulate its input and output. Do you think we could poke …

Any solution that involves having to use a keyword to get the value returned from a function is such a poor design choice to me. Nearly every time I call a function I don't want to…

Being the least bad doesn't make something good. macOS is the least bad choice for the majority of people that just want a machine to mostly browse the Internet, look at their phot…

If they think macOS is one great idea, that's a terrible misjudgment.

Edge Gallery by Research at Google

It seems like Halo's rampancy take on the breakdown of an AI is not a bad metaphor for the behavior of an LLM at the limits of its context window.

Totally! But it doesn't have to be this way.

I care because every second of that startup time is lost productivity and focus. For me and any developer on my team. Hot reload only works if no class or method has changed so tha…

Does this app take 5 minutes to start? That's so much dynamic Spring magic. Also, how do you keep track of control flow when anything at anytime could have been overridden by somet…

Probably training data. The largest number of public repos are built on that stack. We recently picked React for new projects because LLMs seemed to be the most reliable when writi…

jq is probably faster than storage, the network, compression, or something else in your stack and not your bottleneck.