Definitely to keep in mind when using sampling profilers indeed. They're useful, they help to get a starting point in the analysis but they're far from enough to find the root cause of a performance issue.
This definitely shows PyPy as a successful runtime.
Definitely this. In a world where LLM would actually be accurate and would never spit outright crappy code, programmers would still be needed. It'd mean spending less time writing but more time investigating and debugging the produced code.
Interesting research! Is reading code a math and logic task? Is it a language task? Well... it might be its own thing.
The words we use indeed matter. This is definitely a domain where we should avoid ambiguities...
Or why you should let domain simply expire, there's plenty of work to do before that.
Not necessarily my favorite governance model, but if you're on that scheme... those are good guiding principles.
Definitely a complicated history... this doesn't make the evolution or documentation of it easy.
Interesting results. It's especially nice to see how sched-ext allows to easily iterate and experiment with process scheduling strategies.
Since this particular fad apparently doesn't want to die... this is a good reminder about why you want to do something simpler.
Very nice piece. Hopefully it'll push people to remember that the big social media enclosures are not really the Web. We can have more democracy on the Web again if we collectively want to.
Good introduction to collision resolution inside of physics engines.
Looks like a nice CSS library for the semantic styling of web content.
Definitely this, it's not the first time we see such a hype cycle around "AI". When it bursts the technology which created it is just not called "AI" anymore. I wonder how long this one will last though.
Where WebAssembly is, and where WebAssembly on the server is going... let's hope it doesn't become another CORBA.
We're still fairly dependent on just two major web indices... time for an index built as a common for everyone to use?
The often forgotten history behind the creation of Git. This article does a good job summarizing it.
No, your model won't get smarter just by throwing more training data at it... on the contrary.
Fascinating bug... the fine details of mundane protocols like SMTP can sometimes be surprising.
Interesting discussion about UI density. What are we talking about? Is there value to is? Which aspects of a UI are impacting it? The conclusion makes it all very clear.