Fixed link:
https://nitter.net/awesomekling/status/1508953394836353024I haven't forgotten. I just haven't had much time to work on it as more important things have occupied my time.
How much RAM does it use? How big is the binary?
At least we have some source to look at...
https://github.com/SerenityOS/serenity/blob/master/Userland/Libraries/LibWeb/HTML/Parser/HTMLTokenizer.cppNearly 3k lines of
insanely bloated C++ for the tokeniser alone. Mine is ~600 lines of x86 Asm. What happened to HLLs being more efficient to write!? Perhaps they have quantity but not quality.
https://github.com/SerenityOS/serenity/tree/master/Userland/Libraries/LibWeb/HTML/ParserThis looks like a WebKit clone. Parser itself is at least another 3k lines, and that's not even counting the awfully-enterprise-smelling crap like ListOfActiveFormattingElements(!?!?) And yes, let's create another file just for a single fucking HTMLToken class because our brains are too tiny to comprehend it together with the rest of the code. At least it doesn't look like they're allocating and freeing them every time. Mine is around 2kloc of good old C89, and it's still too bloated to my liking.
The idea of SerenityOS itself sounds good on the surface. They say they're rewriting everything from scratch. Yet they still need to use GCC, and being ADHD trendchasers, are using "modern" (=bloated) C++, which means they're going to be far off from writing a compiler of their own.