FreeBSD may have less bloat, but I'm running a web server in a Linux container (debian-slim - so not even as small as you can go with linux) with only 256mb RAM (of which it is only using ~64mb).
> I never claim that 100% coverage has anything to do with code breaking.
But what I care about is code breaking (or rather, it not breaking). I'd rather put effort into ensuring my test suite does provide a useful benefit in that regard, rather than measure an arbitrary target which is not a good measure of that.
> Isn't the simple solution to use detached debug files?
It should be. But the tooling for this kind of thing (anything to do with executable formats including debug info and also things like linking and cross-compilation) is generally pretty bad.
> As an example: I'm autistic and I learn inside-out, building larger new concepts out of smaller existing ones; those with Asperger's on the other paw, learn outside-in instead, breaking down larger existing concepts into smaller new ones; both are part of the "autism spectrum", but differ very fundamentally.
To me this just sounds like the interaction of autism with other variances in neurotype. You can also reasonably categorise non-autistic people into people who learn outside-in and those who learn inside-out.
I want to say that cognitive strategy doesn't meaningfully impact what I'm talking about, but I struggle to define where the line is between cognitive strategy and neural architecture. It appears to me that different neurotypes have different ways of learning and processing information, and that makes for different most effective strategies, and that becomes evident in how the learnings are eventually used. There's a sort of wavy abstract concept called "derived meaning" that sort of helps explain this, but that's even less understood. But basically, how I understand it is that different neurotypes have different frameworks of derived meaning; when they learn something, they persist differing parts of it and in different ways; this results in a different organizational structure, different forms and mechanisms of association between different ideas, different approaches of using the ideas, and so on. I don't know for sure that there isn't a non-autistic neurotype that learns outside-in like in Asperger's ... but if one were discovered it'd probably become part of the autism spectrum for being so uncommon. :D
Inside-out versus outside-in also doesn't really have to do with the order of learning. I could very easily learn the basic metadata of an overall system before I start diving into the details. Black boxes are very common for me to avoid going too deep into research or reverse-engineering rabbitholes (or to paper over an inability for me to learn more about inner workings). It has to do with the structure of learning. For me to model something as a black box, I'll accept that, though there may be parts inside it, I don't need to care how they're implemented. For someone with Asperger's to model something as a black box ... I imagine they already do that by default. How much of that is cognitive strategy or not, I can't really tell, but outside-in learning seems to prefer to pick up entire pieces that have a desired value rather than digging into every detail contributing to that value. While I could mimic that or show behavior that resembles something like that, it's just not really how my brain works on a fundamental level like that? I don't really know how to draw the line. You do raise a decent point.
Quit some apps probably. I often have a bunch of stuff running in the background that I haven't bothered to close yet. It also sounds like it'd be good for detecting software that's gotten stuck in a busy loop or similar.
And/or possibly take a tea break while it chills out.
IIRC, Chrome now uses CoreText/DirectWrite for system fonts on macOS/Windows, and Skrifa (FreeType rewritten in Rust) outlines rasterized with Skia for everything else (system fonts on Linux, web fonts on all platforms).
I believe Firefox leans on the system raserizers a little more heavily (using them for everything they support), and also still uses FreeType on Linux.
> This approach is different to how many text layout engines approach this problem e.g. by adding "one word at a time" to the line, and checking at each stage if it fits.
We found it was roughly on par performance wise for simple text (latin), and faster for more complex scripts (thai, hindi, etc). It also is more correct when there is kerning across spaces, hyphenation, etc.
For the word-by-word approach to be performant you need a cache for each word you encounter. The shape-by-paragraph approach we found was faster for cold-start (e.g. the first time you visit a webpage). But this is also more difficult to show in standard benchmarks as benchmarks typically reuse the same renderer process.
reply