On the layout side rather than the "what spaces are available" side, I really recommend https://gwern.net/doc/design/typography/tex/1981-knuth.pdf , the paper in which the Knuth-Plass algorithm for paragraph layout is defined. (The Knuth-Plass algorithm decides how wide spaces should be on each line and which choices of hyphenation out of some predefined set should be used to lay out a paragraph.) It's super readable and generally quite joyful. Knuth describes TeX as a "labor of love", and it shines through that paper.
For those interested in typst, Laurenz wrote[0] about the differences between the typst and TeX layout algorithms a while ago. The paragraph layout algorithm is the same but the way it interacts with page placement is quite different.
Yeah, one of the things which I always wished for when doing page composition was a way to visualize which paragraphs could be set a line or two longer or shorter while still being set reasonably nicely.
In decades of typesetting, I've had a chapter fall out almost perfectly with nicely pages and appropriately placed figures exactly once (fastest 40 minutes of my life) --- for the rest, it was:
- style the text and place the figures
- check the last page and see if it would be helped by paging tight or loose
- review all the pages and their figure placement to see which was the most problematic/egregious --- fix it
- starting at the beginning, adjust paragraph tightness as necessary, trying to get pages to balance and if need be, figures and references to be placed where the specs call for them --- if need be, adjust figure size/height/placement/style
- if one reaches the end and the selected strategy did not have the desired result, revert back to the initially styled and placed version and try the other strategy
- repeat until everything worked and everything panned out and all pages are balanced and all references/figure placements
> Yeah, one of the things which I always wished for when doing page composition was a way to visualize which paragraphs could be set a line or two longer or shorter while still being set reasonably nicely.
I wrote a LaTeX package [0] that does exactly this. The default settings automatically lengthen paragraphs as necessary, but you can configure it to only tell you which paragraphs can be easily lengthened [1] without actually lengthening any of them.
One thing I find interesting about discussions of typography in Cyrillic is how poor the overall readability of text is in most fonts compared to Latin because of the relative scarcity of risers and descenders (e.g. pqlt etc)
One of my tutors at university claimed that she was able to read 9th century manuscript Cyrillic faster than modern printed books because the orthography was more varied and easier to scan/speed-read.
I remember seeing some studies that experimentally show this to be true for Hebrew (another de/ascender-poor writing system), but can't find them at the moment.
Thanks for the factual explanation! I found the example cyrillic texts unreadable as a set of horizontal lines (serif) and vertical lines (characters themselves) giving the feeling of a grid, but I dimissed it as "I can't read cyrillic anyways".
Now that you wrote it down, it does actually makes sense.
I learned to type in Junior High School in the nineties, and it is extremely difficult to leave a single space after a period. Like that, it took a huge effort for me to break conditioning.
Whenever I type, be it on my phone or on a computer, I always use double spaces after a period. Like you, I'm just used to it and un-learning it is hard!
This was in the US? As someone who didn't learn that rule, I've always found it very strange and, frankly, ugly.
From the article:
> There was just one space width available in the typewriter, so words and sentences were separated by the same distance. The double space was used to differentiate sentences and improve the readability of the text.
I would dispute this. Sentences are separated by a period as well as a single space character, and that's not the same distance as just a single space because the period doesn't have the same visual weight as a word character. A ". " still looks 'wider' than a " ", even if it technically isn't!
The extra space produced a visually "extra" pause.
Just as these blank lines produce an even greater separation. It's about emphasis, and it's going away (IMO) because it's a nicety, not an obligatory part of clarity and communication. Also, because early editing software wasn't complex enough to correctly distinguish between a sentence end and "Dr. Edward Jones". [EDIT: the gd HN editor removed my extra space!!!]
This also makes me think of drumming. There are the sticks that hit the surface and form a pattern of sounds. Lots of different kinds of spaces embedded there!
The space isn't discarded except in monospaced fonts. All the main computer layout engines (web browser, word processor) will add additional spacing there. It's also where line-width and inter-word padding are corrected first so often ends up being at least as big a space as you'd get double-spacing anyway.
Monospace fonts aren't considered generally more readable by people who make or work with fonts. Their particular strength is in reducing character ambiguity and preserving vertical alignment. But "readability" is subjective and depends on particulars of the specific font and of course personal expectation and preference. I find them almost always less readable than a good proportional serif font, except for code.
In monospace the dot is wider than in proportional fonts, thereby adding more space by itself. In addition, the space character itself is wider in monospace than in proportional fonts, relative to the average letter width. In combination, this balances out the difference from proportional typesetting, in my opinion. A dot plus two spaces is jarringly wide in monospace.
I do agree that monospace doesn’t make for readable prose either way.
On the layout side rather than the "what spaces are available" side, I really recommend https://gwern.net/doc/design/typography/tex/1981-knuth.pdf , the paper in which the Knuth-Plass algorithm for paragraph layout is defined. (The Knuth-Plass algorithm decides how wide spaces should be on each line and which choices of hyphenation out of some predefined set should be used to lay out a paragraph.) It's super readable and generally quite joyful. Knuth describes TeX as a "labor of love", and it shines through that paper.
For those interested in typst, Laurenz wrote[0] about the differences between the typst and TeX layout algorithms a while ago. The paragraph layout algorithm is the same but the way it interacts with page placement is quite different.
[0]: https://laurmaedje.github.io/posts/layout-models/
Yeah, one of the things which I always wished for when doing page composition was a way to visualize which paragraphs could be set a line or two longer or shorter while still being set reasonably nicely.
In decades of typesetting, I've had a chapter fall out almost perfectly with nicely pages and appropriately placed figures exactly once (fastest 40 minutes of my life) --- for the rest, it was:
- style the text and place the figures
- check the last page and see if it would be helped by paging tight or loose
- review all the pages and their figure placement to see which was the most problematic/egregious --- fix it
- starting at the beginning, adjust paragraph tightness as necessary, trying to get pages to balance and if need be, figures and references to be placed where the specs call for them --- if need be, adjust figure size/height/placement/style
- if one reaches the end and the selected strategy did not have the desired result, revert back to the initially styled and placed version and try the other strategy
- repeat until everything worked and everything panned out and all pages are balanced and all references/figure placements
> Yeah, one of the things which I always wished for when doing page composition was a way to visualize which paragraphs could be set a line or two longer or shorter while still being set reasonably nicely.
I wrote a LaTeX package [0] that does exactly this. The default settings automatically lengthen paragraphs as necessary, but you can configure it to only tell you which paragraphs can be easily lengthened [1] without actually lengthening any of them.
[0]: https://www.ctan.org/pkg/lua-widow-control
[1]: https://tug.org/TUGboat/tb43-3/tb135chernoff-lwc.pdf
Looking forward to trying this out! (not sure how I missed it...) Thanks!
That's awesome! Thank you very much!
One thing I find interesting about discussions of typography in Cyrillic is how poor the overall readability of text is in most fonts compared to Latin because of the relative scarcity of risers and descenders (e.g. pqlt etc)
One of my tutors at university claimed that she was able to read 9th century manuscript Cyrillic faster than modern printed books because the orthography was more varied and easier to scan/speed-read.
(That wasn't something I found to be true)
I remember seeing some studies that experimentally show this to be true for Hebrew (another de/ascender-poor writing system), but can't find them at the moment.
Thanks for the factual explanation! I found the example cyrillic texts unreadable as a set of horizontal lines (serif) and vertical lines (characters themselves) giving the feeling of a grid, but I dimissed it as "I can't read cyrillic anyways".
Now that you wrote it down, it does actually makes sense.
It is true to some point, yet no so bad
English letters with asc/descenders: b d f h k l g j p q
Cyrillic: б д у ф ц щ
Also ё й could be considered having “stuff” above.
Via https://news.ycombinator.com/item?id=46198040
I learned to type in Junior High School in the nineties, and it is extremely difficult to leave a single space after a period. Like that, it took a huge effort for me to break conditioning.
Whenever I type, be it on my phone or on a computer, I always use double spaces after a period. Like you, I'm just used to it and un-learning it is hard!
This was in the US? As someone who didn't learn that rule, I've always found it very strange and, frankly, ugly.
From the article:
> There was just one space width available in the typewriter, so words and sentences were separated by the same distance. The double space was used to differentiate sentences and improve the readability of the text.
I would dispute this. Sentences are separated by a period as well as a single space character, and that's not the same distance as just a single space because the period doesn't have the same visual weight as a word character. A ". " still looks 'wider' than a " ", even if it technically isn't!
I agree that it oversimplifies.
The extra space produced a visually "extra" pause.
Just as these blank lines produce an even greater separation. It's about emphasis, and it's going away (IMO) because it's a nicety, not an obligatory part of clarity and communication. Also, because early editing software wasn't complex enough to correctly distinguish between a sentence end and "Dr. Edward Jones". [EDIT: the gd HN editor removed my extra space!!!]
> I would dispute this.
I wouldn't. Typewriters don't work like computers. The additional space was objectively beneficial. I personally witnessed that.
This also makes me think of drumming. There are the sticks that hit the surface and form a pattern of sounds. Lots of different kinds of spaces embedded there!
Of all spaces, the space between sentences is discarded because a period is whitespace. However, kerning partly removes this.
Perhaps this is why monospaced fonts are so readable? I like having double-space between sentences.
The space isn't discarded except in monospaced fonts. All the main computer layout engines (web browser, word processor) will add additional spacing there. It's also where line-width and inter-word padding are corrected first so often ends up being at least as big a space as you'd get double-spacing anyway.
Monospace fonts aren't considered generally more readable by people who make or work with fonts. Their particular strength is in reducing character ambiguity and preserving vertical alignment. But "readability" is subjective and depends on particulars of the specific font and of course personal expectation and preference. I find them almost always less readable than a good proportional serif font, except for code.
In monospace the dot is wider than in proportional fonts, thereby adding more space by itself. In addition, the space character itself is wider in monospace than in proportional fonts, relative to the average letter width. In combination, this balances out the difference from proportional typesetting, in my opinion. A dot plus two spaces is jarringly wide in monospace.
I do agree that monospace doesn’t make for readable prose either way.
There is something oddly beautiful in invisible complexity. These little tweaks and minuscule details without which the whole would suffer in quality.
> Standard space, word space, space per se, is the symbol typed using the widest key on the keyboard.
What a strange non-fact to include.
Have you ever had a keyboard where that wasn't true? Even on my cellphone...
Ultimate hacking keyboard