Hot take: Current C and C++ compilers are hilariously wasteful and inefficient in a ways that are completely orthogonal to the preprocessor, C++ Modules, etc. I plan on writing a blog post to support this claim. I also hope to try and experiment with solutions to these wastes.
-
-
Replying to @vector_of_bool
I’ve been wondering about caching optimizations. Whereas ccache is coarse-grained, if I build the same project repeatedly, even if headers change a bit, at some point it’ll be spending a lot of time optimizing the same functions it did last time and getting the same results.
1 reply 0 retweets 2 likes -
Replying to @benfrantzdale
Caching is a huge thing I want to address. CCache is great but flawed. Partial blog spoiler: I want the compiler to remember the token sequence after preprocessing and skip compilation if it sees the same in a later compiler invocation.
3 replies 0 retweets 3 likes -
Replying to @vector_of_bool @benfrantzdale
There was an IBM C++ compiler that did what you want.
@cmuratori mentioned it once in his Handmade Hero stream. The name of IBM’s C++ compiler escapes me.1 reply 0 retweets 0 likes -
It was called "Montana". The released product name was "IBM Visual Age". The product itself was awful because it was a giant bloated install that only worked on, like, the _previous_ version of Windows NT or something when it came out :(
1 reply 0 retweets 1 like
But the compiler work was forward-thinking. A lot of the things people today are only starting to do, they did. You can read about it in their paper if you are so inclined. https://dl.acm.org/doi/abs/10.1145/291252.288284 …
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.