3 ms·
Okay, I already know that most people about to comment on this post will complain about "yet another css framework", "not again", and "isn't this just Skeleton?
by ajusa 9y ago
Okay, I already know that most people about to comment on this post will complain about "yet another css framework", "not again", and "isn't this just Skeleton?".
This was originally based off of Skeleton, with the intent being how small I could get the source. I looked online for CSS code golf, but I was unable to find any article, or even examples. The only thing I was able to find was CSS shorthand.
This is my code golf experiment. My goal is sub 500 bytes, anything lower probably will not be possible (then again, I said that about sub 700 bytes). My rules are pretty much full feature parity with Skeleton, and looking similar to the way lit looks right now.
The main way I was able to make lit this small was by optimizing for the gzip algorithm. I learned a ton by doing that, including some counter intuitive things when compressing files.
I would love for people to join me and help out, because I am running out of ideas to make the source smaller.
- tmzt 9y agoWould optimising for brotli and/or a custom gzip encoder result in improvements?
- ajusa 9y agoThat is an interesting question. It should result in pretty significant improvement, however lit would lose practicality. Optimizing for normal DEFLATE/gzip makes lit more practical, since brotli doesn't have widespread use. Sometime in the future, when brotli has replaced DEFLATE, lit could be further optimized. In general though, the differences would likely be minimal.