3 ms·
This sounds a little similar to the hubris of the GNU moe [1] manual. You're assuming you could make a "good enough" font that would truly be "good enough". I d
by ashton314 5y ago
This sounds a little similar to the hubris of the GNU moe [1] manual. You're assuming you could make a "good enough" font that would truly be "good enough". I don't think you can even come close.
A font adds so much character to the text. Where one font may be appropriate, a different text would look stilted or silly. Fiction, fantasy, technical, non-technical non-fiction, etc. books all require different fonts, in my opinion, to capture the character of the words.
Noto is a fine font, but I don't want to use it for my blog. There are a bajillion fonts that are perfectly legible and for those of us who care, they make a difference.
What would help would be a little education and maybe some UX nudges: stop making Arial the default on stuff, (looking at you, Google Docs) encourage well-built fonts by making them the only default enabled options in your OS, etc. I don't think there is "one font to rule them all", however.
[1]: https://news.ycombinator.com/item?id=29990149 https://news.ycombinator.com/item?id=29990149
- nomel 5y ago> that would truly be "good enough". I think 99% of the population could agree on a "good enough" font, if reading comprehension is the goal. In fact, I think 99% of the population just reads the words. Our brains necessarily learn to ignore the fonts. Our brains necessarily learn to ignore even the characters themselves [1]. > There are a bajillion fonts that are perfectly legible and for those of us who care, they make a difference. I think this demonstrates that the goal is not reading comprehension, which makes agreeing on some "good enough" difficult/impossible. 1: https://www.livescience.com/18392-reading-jumbled-words.html https://www.livescience.com/18392-reading-jumbled-words.html