Pretty Printing (1979) [pdf] | Hacker News Reader

Pretty Printing (1979) [pdf]

i.stanford.edu

26 pointsazhenley8 comments

The more commonly cited paper is the Wadler paper[1]. It's the backbone of Prettier. Although tbh it's a little hard to decode the semantics of the Wadler paper into a more conventional language because of Haskell's laziness. Stuff like union is implicitly lazy in the paper which can trip you up.

Also while the core semantics are (basically) the same in the Wadler paper and Prettier, there's a thick layer of special cases that allows Prettier to actually print stuff in a human readable way. Turns out there's a lot more than just math when it comes to readability.

[1]: https://homepages.inf.ed.ac.uk/wadler/papers/prettier/pretti...

It’s ironic when academic publications about print formatting and other ways to increase legibility all use full-justified text, which is harder to read than left-justified.

I don’t see the irony. The paper isn’t called “legible printing”. Well typeset full-justified text is considered prettier than the jagged layout of left-justified text.

Justification-wise, this paper doesn’t look bad to me for the time period.

[1] some printers would go as far as tweaking the text to improve the layout.

Is there research showing this? If we're just talking opinions, I'd have to disagree and say I find justified documents are easier to read.

Several years ago there was a study that asserted this, which changed my opinion about justified text. But of course I can’t find that article now. :-(

Amusing that a PDF discussing pretty printing is typeset like this. I've noticed this across many PDFs in the past - anyone know what causes it?

It's a scanned and OCR'd hardcopy (after all, desktop typesetting was just barely a thing in 1979) but nobody has bothered to clean up the OCR results.

typesetting in 1979 was widely practiced, but very different than the last thirty years. Most print of this era is just scanned and reproduced as a raster.

Reply on news.ycombinator.com