@enkiv2 What surprised me about that comment was how old it was. I'd made the first version of it at least six years ago.

I've been looking at typography, layout, readability issues, and more, since, and ... it's interesting.

One of the more interesting (and beautiful) concepts I came across were the canons of page construction, which are graphical, rather than measured (so they can be reproduced with either a simple straightedge or straightedge and compass), and achieve quite beautiful and useful results.
https://en.wikipedia.org/wiki/Canons_of_page_construction

The key with information is that form should serve function. Generally that is getting out of the way of the reader (something far too much Web design fails utterly at), and, as a recent HN submission noted, "conserving the most valuable resource you have as a writer --- reader attention."

https://practicaltypography.com/why-does-typography-matter.html

1/

@enkiv2 The author / typesetter (frequently one and the same, or one invoking automated tools for the other) is faced with constraints, of course, most notably limitations of the medium, media, and/or formats, and for online information, a huge problem is that there are so many different types and varieties of devices for accessing written material.

I'm sitting in front of a 5K Retina iMac right now. My new mobile device is a 13.3" ebook reader with a 200 DP 2200x1650 16-shade greyscale display and a rather slow refresh rate.

I use both for much the same material, but what they can do differs fairly strongly, and there are a wide range of other devices, densities, colour renditions, refresh rates, portability, etc.

2/

@enkiv2 I find myself struggling with the capabilities and limitations of each, as both a reader and as a publisher / republisher of content. (Markdown -> LaTeX -> PDF does quite well on the Boox.)

And I'm noticing the issues of lots of Web content (and some intended for print). Colour does horribly on the Boox. There's something about bold-faced fonts that looks horrible (possibly a mode setting), and there are tweaks to refresh rates, font-sizes, and contrast which can be made, but need to be individually applied to each website (or often: pages within them).

I'm wondering if e-Ink will become widespread, in which case the issues might have to be addressed, or if it will remain a niche. The battery life, display resolution, and daylight-readability of the Boox are strongly appealing.

And pretty much all highly-tuned Web content translates exceedingly poorly to the Boox.

Line-art features beautifully (there's a reason the standbye / off screens are of this form: https://www.boox.com/wp-content/uploads/2020/09/Max-Lumi-10-1.png). Design for e-ink would tend to emphasize fine lines for subtle colour gradations.

Similarly, animation does horribly on e-Ink. Rather than a smooth transition, all one gets is a chaotic region on the display. If that speeds the demise of gratuitous animations, so much the better.

3/end/

@dredmorbius

HTML/CSS doesn't have straightforward support for some of the most powerful information-presentation tools of the codex form -- tools that have been in use for centuries, and that were the basis for the concept of hypertext in the first place: marginal notes, indexes, parallel columns, and random access.

And, of course, there's no end user annotation either (the whole point of hypertext in its original conceptions!)

@dredmorbius
For this reason, I consider it essentially broken.

It's like a shiny new sports car with no engine in it: it looks nice in all the ways that nerds who like that kind of thing look for, but it can't actually get you anywhere, and is therefore useless.

@dredmorbius
Web tech is popular because we handed the reigns of our entire industry to the kind of people who (to extend the metaphor) obsess over sports car bodies to the point that they'd rather have a sports car with no engine than a 20 year old Subaru that works reliably. And as a result, these people spend all day every day polishing a completely stationary vehicle.

Of course, it's TBL's fault that we would need to retrofit an engine! Even in 1992, it clearly should have shipped w/ one!