Skip to content

Comment on Using machine learning to choose compression algorithmsparent

Comments

The downside with switching over to Jekyll is that gwern.net is increasingly dependent on Haskell integration with Pandoc, which would be difficult if you were compiling with another language: you'd have to refactor all of the additional rewrite passes (for interwiki scripts, image dimension specification, affiliate links, link annotations, inflation adjustment...) to make it at all possible, and you'd be using a lot of Haskell anyway. So it goes - the Law of Equivalent Exchange.

AboutSource Built by g1lg1l

Hackerly is an independent reader for Hacker News, built on the public HN API. Not affiliated with Y Combinator.