Semantic linefeeds: one clause per line

I’ve started using “semantic linefeeds”, a concept I discovered on Brandon Rhodes’ blog, when writing content, an idea described in that article far better than I could. I turns out this is a very old idea, promoted way back in the day by Brian W Kernighan, contributor to the original Unix system, co-creator of the AWK and AMPL programming languages and co-author of a lot of seminal programming textbooks including “The C Programming Language”.

The basic idea is that you break lines at natural gaps between clauses and phrases, rather than simply after the last word before you hit 80 characters. Keeping line lengths strictly to 80 characters isn’t really necessary in these days of wide aspect ratios for screens. Breaking lines at points that make semantic sense in the sentence is really helpful for editing, especially in the context of version control, because it isolates changes to the clause in which they occur rather than just the nearest 80-character block.

I also like it because it makes my crappy prose feel just a little bit more like poetry. ☺


Webmentions

You can respond to this post, "Semantic linefeeds: one clause per line", by: liking, boosting or replying to a tweet or toot that mentions it; or sending a webmention from your own site to https://erambler.co.uk/blog/semantic-linefeeds/

Comments & reactions haven't loaded yet. You might have JavaScript disabled but that's cool 😎.

Comments

Powered by Cactus Comments 🌵