Atkinson Hyperlegible Font | Hacker News Reader

Atkinson Hyperlegible Font

brailleinstitute.org

799 pointszdw234 comments

With a few tweaks (monospaced variant), it could make a great coding font.

Coding fonts have essentially the same goal: make similar looking characters different: 1liI|, 0O, etc... Code naturally strains our visual acuity: we want to see as much as possible, it means that we may use a smaller font size than we would find comfortable reading text with. Single letter, punctuation, etc... can be really important, in regular text, you can understand even if you can't read all the characters. So a hyperlegible font can help coders, even those with good vision, and many coding fonts already have "hyperlegible" features.

Indeed, that's what I thought too when I read about the letter distinction. The reverse slash on the 0 seems a little odd, but otherwise this looks like a proportional version of Consolas.

Pardon me, where do I find the monospace variant?

What OP was saying was a few tweaks could make a monospace variant, which would be a nice programming font. I don't think a monospace variant is available right now.

Ah yes, sorry. I think MonaLisa actually fits the bill: https://www.monolisa.dev/

For the life of me I just can't get over how different the g is from the rest of the letters. I find it incredibly distracting, so I can't use this font.

Increasing differentiability (within reason and identifiability) is the point of fonts that are trying to increase legibility. g has the potential of being confused with 9, y, 0, etc. for folks with non-optimal sight.

Their paid (120 USD) Monolisa Plus variant does have the regular g, though.

For some reason I can't stop giggling at this.

Check out DejaVu Sans Mono[1], if you haven't already. That's what I use for programming. Has good versions of 1LiI0O (etc).

[1] https://dejavu-fonts.github.io/

(examples in PDF: http://dejavu.sourceforge.net/samples/DejaVuSansMono.pdf)

Unless you're on a Mac, in which case just use Menlo, which is already installed. Menlo and DejaVu Sans Mono are both descendants of Bitstream Vera Sans Mono, and are nearly identical. (And where they differ, Menlo is superior IMHO.)

No font should lack crossbars on the capital "i." That's just straight-up dumb. Nice to see that the font in question doesn't suffer from that mistake.

Failure to distinguish O from 0 ranks pretty high on the "don't" list as well.

I made my best attempt at this. I made custom Iosevka[1] builds that use the letter shapes from this font. I call it Hypersevka, and the build plans are available at [2].

[1] https://typeof.net/Iosevka/ [2] https://github.com/jdknezek/Iosevka/blob/jdk/private-build-p...

Nice. Any chance you can post a screenshot?

Sure! Here are some screenshots (Iosevka customizer then VS Code):

https://imgur.com/a/iLejIfc

Any chance you'd make a derivative font - monospaced or not - that uses different letter heights (i forget the technical term) for dyslexic users?

What are the PTL files in the repository? GitHub doesn't identify the language, and I don't recognize it, either.

That’s the "Patel" programming language used to define Iosevka. See https://github.com/be5invis/PatEL.

I also find the website confusing, the font is available here however https://fonts.google.com/specimen/Atkinson+Hyperlegible

Thank you; couldn't get the download link to work.

The Braille Institute site makes the file available by http instead of https, which a lot of browsers will reject for security reasons.

I wonder which nation-state will exploit Windows font rendering and MITM the Braille Institute with a malicious payload. Iran, surely?

Also here, if one would like to avoid Google somewhat https://www.fontsquirrel.com/fonts/atkinson-hyperlegible

GitHub link: https://github.com/ThomasJockin/readexpro

That is not a link to this font, unless Readex Pro is a rebranded Atkinson Hyperlegible. :) (From the Readme file, I'm pretty sure it isn't.)

However: https://github.com/googlefonts/atkinson-hyperlegible

Oops you're right, Readex is the rename of Lexend. Sorry!

Reminder that fonts are subjective.

This one is good for specific people, not everyone. I can't use it because I find it overly rounded with tics.

An example of a font that is also meant to be legible, but is on the squarish side, is Tiresias Screenfont.

Increased legibility does NOT mean "flow". Just because each character is more "distinct" does not mean that the whole word is able to be processed with ease. Important to evaluate them in body text.

Also, one's cultural background absolutely affects which fonts are preferred. If you grew up with German blackletter writing, you're not necessarily going to like or appreciate a sans-serif font. In Germany's case, the change from blackletter to latin shapes happened fairly recently (WW2). It's happening now again, Kazakhstan has decided to stop using cyrillic and change to latin characters.

> An example of a font that is also meant to be legible, but is on the squarish side, is Tiresias Screenfont.

Here's a quick example that a DuckDuckGo search turned up: https://catalog.monotype.com/font/bitstream/tiresias/screenf...

Personally, I rather liked the Atkinson Hyperlegible as well, though I found the "bgrpq" characters too similar in shape regardless, but maybe that's a criticism of the alphabet itself.

Regardless, a lot of the time I find myself using whatever is popular (e.g. Open Sans for web development) or something like Liberation Mono for writing code (though the rest of the Liberation fonts are great alternatives to Microsoft fonts, for example, when using LibreOffice).

Open Sans: https://www.opensans.com/

Liberation fonts: https://en.wikipedia.org/wiki/Liberation_fonts

>Reminder that fonts are subjective.

Yes but you can survey a fraction of a population and make strong statistical claims such as, “most people found so and so font more readable”.

Which population though? There are differences in age groups. People who grew up only with print won't lean toward a sans. People who grew up only with screens will lean toward sans, etc.

If you go with that metric of doing a survey, then the answer is Times New Roman and Arial, not because they are superior (though they have excellent hinting) but due to long-term familiarity and exposure.

I don't have a problem with providing what people are comfortable with, but comfort does not necessarily translate to better. It's subjective.

Companies commissioning their own fonts, is not due to a desire to get improved quality, but simply to not have to pay a licensing fee for usage.

>Which population though?

The best population to sample would probably be the readers of your content.

>If you go with that metric of doing a survey, then the answer is Times New Roman and Arial, not because they are superior (though they have excellent hinting) but due to long-term familiarity and exposure.

I don't know the answer to this because I didn't do any studies. I'm probably reasonable to assume you didn't either, and thus should be careful about making such assertions.

>I don't have a problem with providing what people are comfortable with, but comfort does not necessarily translate to better.

I don't know what you define as "better", or what you're trying to get at.

>Companies commissioning their own fonts, is not due to a desire to get improved quality, but simply to not have to pay a licensing fee for usage.

Could be. This is a pessimistic take, and I'm sure it's right sometimes and wrong other times.

Fonts can be a core part of a company's brand identity, the present-day fad for incredibly bland Helvetica in a solid color notwithstanding.

Preference has nothing to do with legibility.

There's something that feels... off about this font. The horizontal rhythm is out of whack here. I don't know what it is, I'm not a font designer, and the last time I looked into details of that was 20 years ago. Bad hinting? I don't know.

I think the technical term is `kerning` (though also not a typography person).

I feel as though there should be a standard set of tests that any or all fonts run through to demonstrate the various properties they want to claim.

Don't tell me "better", tell me "Scored an 8.6 on legibility in the standard font assessment test".

That would be valuable in comparing and searching for fonts. But some of these criteria are aesthetic.

Why do content creators need to do this? Particularly if they are experts and live off the trust in their experience. I am very happy if someone publishes something and states there goals an claims, so someone else can verify them independently. If it is not working they would risk reputation.

If a few publisher's really check accessibility against relevant end users and verify that this works it is much better, than any kind of scores that often overlook important usability/accessibility issues.

a11y Checkers are OK like any linter but I think we should not overdo this. That something has a higher score does not translate to better overall a11y.

It's from The Braille Institute. They ARE the standard when it comes to dealing with vision challenges / vision loss.

Ironically those who need Braille are also unlikely to care how readable a font is.

The Institute is named after the same person as the writing system for the blind, rather than being named after that writing system.

The Braille Institute provides all kinds of assistance to partial- and non-sighted individuals. Not limited to the kinesthetic alphabet known as “Braille”. For example, teaching echolocation (yes, for humans) and how to use a cane to navigate safely.

This font is an example of how the institute helps the partially-sighted, like for example, people with glaucoma or macular degeneration.

Yep, and I truly believe that they've made something great.

But the onus is on them to provide evidence that they did so.

Otherwise, I can stand up and say: "this font makes reading slower and causes more mistakes" and it's a pissing contest between me and them rather than a matter than can be settled.

Atkinson Hyperlegible – a font by the Braille Institute designed for legibility - https://news.ycombinator.com/item?id=28010540 - July 2021 (1 comment)

Atkinson Hyperlegible Font - https://news.ycombinator.com/item?id=26011945 - Feb 2021 (86 comments)

Atkinson Hyperlegible Font - https://news.ycombinator.com/item?id=25154417 - Nov 2020 (10 comments)

Atkinson Hyperlegible Font - https://news.ycombinator.com/item?id=24853550 - Oct 2020 (3 comments)

A Free Hyperlegible Typeface from the Braille Institute - https://news.ycombinator.com/item?id=24791358 - Oct 2020 (4 comments)

Although the font did really try to nail its goal, I think Iosevka really nailed it.

On top of that, somebody took some of what Atkinson did with their font, and applied it to Iosevka, using its insanely powerful ability to be customized wildly, to produce http://thedarnedestthing.com/iosevka%20hyperlegible

FWIW, the font on that blog post is not Iosevka,

You can find Iosevka here -> https://typeof.net/Iosevka/

I find the 0.5em character width reduces readability somewhat. YMMV.

It is an Iosevka.

Off-topic, but I think this might be the first time I've loaded a url including `%20` and had it actually render in my address bar as a space, rather than the percent-code. Is this something new in Firefox?

It's been like this in Firefox for quite some time

the irony of having "hyperlegible" in the title when the title itself is barely legible.

using a tiny 11px grey colour font, the body text is borderline while the navigation text is almost transparent.

Meybe they did that for their target audience to see that the font is legible.

Umm that website is barely legible, let alone hyper. Small, thin, spindly, faint, what the hell? (On Safari Ventura)

I think that was done on purpose.

It's rather awful contrast, with the color choice making it worse, yet I can read it at 100% zoom when I would otherwise be unable to with any other font choice.

An aside: Why is it that most posts about a new product or project some people have put together is usually greeted with skepticism and negativity in HN comments? What is that all about? It is a pattern Ive continuously seen and it seems like only the true home runs receive any kind of praise.

Most replies are either about lack of evidence or someone’s alternative preference. I see that latter one a good deal.

probably the same reason new JS frameworks get a lot of criticism. adding another one expands an already daunting array of choices to make, most often without a clear benefit, as seems to be the case here.

Because you are the easiest person to lie to, yourself. Just because you work in a field doesn't mean your assumptions are valid. The more people (or you yourself) consider you an expert in something, the more you should test your assumptions through studies etc. If you don't, then claiming you made a thing "for X" without any proof that X benefits from that means you've almost certainly lied to yourself, convinced yourself that lie was true, and are now perpetuating that lie by putting it on the internet. Especially when you do things based on your own experience, it's incredibly easy to forget that there are almost infinitely more people who are different and have different needs and experiences.

And of course, your things can be just fine, but without proof of that, claiming that they are just tells people you have no idea whether what you did was actually worth anything, and that undermines your effort.

The thing you made MIGHT actually be great! But it might also only be great for a super select few hyper-focussed people that you happened to ask for help (or not even, you might have purely relied on "your own past experiences"!), and be terrible in general.

Make tiny claims centered around tiny groups (e.g. "I/my customers needed something better so I made this"), not a problem. Make big claims involving "everyone" (e.g. "we made a typeface that improves legibility")? Back them up with proof. Show the studies where you've pitched it against a wide set of other typefaces, and have people of all visual impairment levels perform (also scientificially justified for this purpose) tasks that hinge on legibility.

(Because winning awards may be nice, but doesn't tell us anything. It just says "other people already liked this". It doesn't say anything about whether the thing you made is actually good)

>Why is it that most posts about a new product or project some people have put together is usually greeted with skepticism and negativity

Because most posts about a new product or project tend to make extravagant claims without anything to back it up. "The best", "the fastest", "the smallest", "the most legible", etc.

Because most of the time, the original post has enumerated all of the good things. There’s no reason to reiterate the points made in the post in the comments, and so the comments add an opposite view.

There is a healthy amount of skepticism that is required to produce good for the society. It can be adjusted based on supply/demand of such commodity or service. But, by large, skepticism is good. As a creator myself, it is important to equip yourself with the right mentality. I would never blame everyone else for being skeptical. After all, most people mean good and there is often a solid reason they're spending time writing about it.

Highly, highly recommend reading Adam Savage (Mythbuster fame)'s book about creativity and skepticism, it is filled with good wisdom: https://www.goodreads.com/book/show/43319933-every-tool-s-a-...

If there is a complete lack of criticism/skepticism, you get a regressive world: https://fs.blog/chestertons-fence/

- The comment structure favors the 'middlebrow contrarian' style of response. This phenomenon is particularly acute on Reddit but on here it's a more understated process that's proportional to HN's design differences to the former.

- Insecurity. Criticism is the most acute and well researched on here with child prodigies, more so than any other topic. Since people here get their sense of identity from their self-perception of intelligence, knowledge, and capacity to build certain things, there are many vulnerability points.

- Wide, industry ranging experience with bullshit technology claims for most users.

- Class and ethnographic differences. Criticism will vary depending on whether the product in question is inherently appealing to adult high-income nerds, no matter its overall utility

It does feel like sometimes we go beyond "healthy skepticism", and assume bad intent. I think there is good intent here. Perhaps their work is coming from more of a liberal arts angle, than an engineering angle.

It is not just when there is a post about a new product or project. More broadly for ANY post here, the top comment and most comments will generally claim the opposite of what the original post presents is true (or some variation of that).

I would say mainly it is just the need for some people to feel good about themselves that they know better, or to try to show that they are smarter than the author of the post (so that again, they can feel better about themselves).

Don't take the comments too seriously. A lot of people think they are smarter than they actually are.

And next time notice, whatever the post is, when you click to view the comments expect the top one to have a contrarian view.

EDIT: As an example, as of right now, this post is the top post on HN. The second top post on HN?

https://news.ycombinator.com/item?id=32800869

Title: "Build Your Career on Dirty Work"

Top Comment: "Really bad advice! Hard work does not pay"

Upon further thought, looking at the bright side (less cynical), I guess this does help give a more balanced view of the topic as you get to quickly see both sides of a story. But it's still kind of funny that this consistently happens.

>Why is it that most posts about a new product or project some people have put together is usually greeted with skepticism and negativity in HN comments? What is that all about?

It makes perfect sense why these things are greeted with skepticism. Because Most things in the world don't succeed. Most things fail. Thus when you see things treated with negativity and skepticism that is MORE likely to be inline with the actual reality. Thus if HN is actually more negative then normal, then that means HN users have a more realistic view of reality.

>Most replies are either about lack of evidence or someone’s alternative preference. I see that latter one a good deal.

Why is the latter a good deal if it isn't true? If 99.999% of the world doesn't have this alternative preference it's biased to even bring it up. There's no need to be personally insulting but if you think something is genuinely bad then i think it's perfectly ok to just say what you think.

When someone asks for praise, or when someone asks for positivity in a way they are by probability more likely to be asking for fake opinions and white lies. I don't come to HN for that kind of thing. But I will say that Dang (the moderator) loves this kind thing; and the moderation culture he promotes is more inline with your attitude. If you see a negative post, you can flag it, and he will take a side.

I'm uncomfortable with people publishing work like this without studies to back up whether they work for their intended audience. I'm sure the Braille Institute is expert in needs of low vision readers. And the design certainly looks promising. But AFAICT no one has studied whether this font is actually more readable. https://www.maxkohler.com/notes/2021-02-16-atkinson-hyperrea...

The mess with the popular-but-not-effective OpenDyslexic isn't good for anyone except publishers wanting to tick off an "accessibility" checkbox. (Thinking particularly of the library e-reader Axis360 which includes only two fonts; a bad regular font and a "dyslexia font". Neither are particularly readable IMHO.) https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5629233/

This implies there is a study showing the open dyslexic font is actually not good? I can't say that would shock me, but it is surprising that they wouldn't have done some studies to justify the claims.

Edit: the second link wasn't loading for me, so now I see the study. Bad phone internet... :( Again, still surprised they didn't have the counter study.

That's the NIH study that was linked.

> Results from this alternating treatment experiment show no improvement in reading rate or accuracy for individual students with dyslexia, as well as the group as a whole.

That's the key finding. And while "no effect" may sound harmless the Discussion section of the paper highlights all the ways having a popular ineffective solution is actively harmful to people with dyslexia.

Yeah, apologies for my post before the extra links loaded for me. Definitely should have either worded it with the expectation that the link was that study, or waited.

My intent was to express that surprise that they didn't have studies showing benefit. That is, the implication that really surprised me was the reverse, that there are not studies showing it works.

To be fair, most studies on fonts, at least as they relate to reading speed, have shown that there's much more cultural variation than we realize. Different line spacing, serifs vs sans serifs, font weights, etc can all affect reading speed positively or negatively based on age of the reader or cultural background.

It turns out that the fastest fonts to read in are the ones people have practiced using the most.

Given this, perhaps a more useful metric would be the maximum reading speeds for each font. This could give us something of an idea of what is possible given sufficient proficiency.

Fonts like Atkinson Hyperlegible are very well designed around some simple but well accepted principles yet, to most people without a designer's eye, it looks like any other sans serif font

---

PS if you wanna keep up with the latest in font readability research, you should check out the Readability Consortium:

https://www.thereadabilityconsortium.org/

As a dyslexic the mechanism for OpenDyslexic always seemed a little off for me. How does a differently shaped font help adjust for a difference in cognition?

However in this case it would seem like focusing on ensuring that the letters can be distingushed with poor vision has a more direct mechanism? - though it may be that other fonts are better

>"How does a differently shaped font help adjust for a difference in cognition?"

AIUI for some people with cognitive difficulties when using text they find orientation of glyphs (which form letter characters) to be difficult to discern, and similarities across glyphs to be confusing. Thus, if glyphs are more differentiated from one another, and if they have a non- rotationally-symmetrical shape, then letters can be easier to comprehend.

I'm curious whether fonts like Dyslexie mighty bed better for those learning to read. Children learning to read often confuse letters, b/d/p/q for example. I can see ways it could both help and hinder.

This is a tension in the field of dyslexia/reading. Most academics in the US believe that dyslexia is phonological, not visual. If this is true, then visual supports like OD would not help dyslexic readers.

Many practitioners (and researchers outside the US) have a different view, which is that there are different strains of dyslexia, and some strains are more visual than others. I've talked with SPED teachers who laughed when they heard that researchers think dyslexia is phonological, not visual.

My own belief, based on years of working in the field, is that there is a significant portion of the dyslexia population who can benefit from visual changes to text presentation. This may be a direct symptom of their dyslexia, or it may be an indirect effect of (1) having dyslexia and struggling with reading, which leads to (2) not reading as much and having less-developed pathways related to the visual aspects of reading.

But given how dyslexia is defined as a residual category (roughly: a person who has a low reading level, not caused by visual impairment or deficits in general intellectual ability), it seems highly unlikely that no people with dyslexia have any visual aspects to their condition. There may be some, or even most, for whom the condition is phonological. But reading is visual in nature, so it would be very surprising if the group of people who struggle with it didn't happen to include anyone whose difficulties are visual in nature.

Of course, one can define dyslexia more narrowly (and some do), but schools typically don't. So if a broad range of kids are diagnosed as "dyslexic" in school, then it doesn't make sense for experts to proclaim "dyslexia is never visual, and people who say otherwise are wrong!".

My experience is based on launching a speed-reading tool (on HN, of course! [1]) that ended up becoming popular in the dyslexia and ADHD communities. It is a tool that is visual in nature, and I have gotten tons of emails from people with dyslexia who describe it as life changing. Some experts believe in what we're doing, but others are completely opposed to it. The dogmatism among certain experts conflicts with what I hear from people IRL, some of whom I have literally seen brought to tears by how effective our tools are. Even if what these people have is not "dyslexia" as defined by some people, they struggle with reading and are told they are dyslexic.

1: https://news.ycombinator.com/item?id=6335784

The answer to the question "Was there any scientific data or studies used in the design?", which begins "The design comes from the tradition of type design. It’s not really rocket science... " does little to inspire confidence.

I work in this area and it's true that OD hasn't fared well in various studies, but there are some people who swear by it. When it comes to matters of cognition and perception, it's hard to say what does or does not actually make a difference for people, and what evidence should be accepted.

After enough people asked for an OD option in my browser extension (which is used heavily in the dyslexia, ADHD, and vision impaired communities, as well as by other readers), we decided to offer it. I know the science behind it is not stunning, but who am I to tell people that the thing they think helps them read does not actually help them read? Even if they read more slowly with OD than without (something I'm not sure is true for people who choose to use OD), it's possible that one might enjoy reading more with OD even if it doesn't improve reading speed.

I wholeheartedly agree with your point about companies wanting to check the box on accessibility by offering OD, and it's unfortunate more companies don't do more.

Note: the page itself is set in Atkinson, which looks more than good enough to stand on its own, even if you would not be aware of its hyperlegible properties.

Are there any monospaced fonts like this, with a focus on hyperlegibility?

There are a number of very nice monospace typefaces that have lots of character variants so you can essentially pick and choose different letter styles (dotted, crossed or regular zeros, etc).

I personally love Fira Code with its ligatures.

I love Fira Code. I suspect Iosevka is the winner when it comes to customization though:

https://typeof.net/Iosevka/

You can basically mimic features of almost any other big name among monospace fonts.

The only downside I can think of is that it updates really often and the website doesn't have a way to save your customizations, so upgrading is a drag.

Given that it's a free font I'm not complaining though.

Yes, that's another great one!

I recently discovered Fira code through https://www.codingfont.com/. I played with font names off, and Fira code won two times in a row :) I'm still amazed at how pleasant it looks on my screen.

Oh, I didn't know of https://www.codingfont.com/, the tournament style competition is pretty cool. I also like that you can turn off the labels.

I've used Fira Code for years and tried to find an alternative on https://www.programmingfonts.org/ but it's very difficult comparing so many different options at once.

Font preference is very subjective.

My personal fave is bitstream vers sans mono.

But the one that ships with pychcharm is really good, too.

well legibility is an objective, measurable characteristic

Python developer here seconding Bitstream Vera Sans Mono. I like having that tiny space between successive underlines, i.e. in things like

  def __init__(self, a, b, c)

I use Atkinson for the browser things, and I personally recommend DejaVu Sans Mono. The rest of that font family is excellent as well.

I'm going to assume you want to use it for development. For screen fonts, you typically have different design goals, because otherwise anti-aliasing and ClearText will reduce contrast and void your legibility improvements by spreading things out over subpixels.

I tried to make my own highly legible high-contrast coding font using TensorFlow to fit things onto a high-resolution pixel grid based on a low-resolution draft. That way, characters align with the pixel grid (at the right scaling) which makes them appear much clearer.

https://hajo.me/images/HajoCode16px_hr.png

https://hajo.me/blog/2021/07/24/making-a-font-that-doesnt-su...

Smaller bitmap fonts tend to be easy to distinguish. ... once you learn the patterns.

I am a big fan of Berkeley Mono. It has (for me at least) this really hard to describe fusion of retro and modern that just feels perfect for coding and console to me. Note: Unfortunately not free, but I did manage to figure out how to make it a static definition as part of my NixOS declarative configuration, lol.

https://berkeleygraphics.com/typefaces/berkeley-mono

Note that I just noticed that you can easily add Atkinson Hyperlegible Font to your Nix config, the package is called (unsurprisingly) `atkinson-hyperlegible`

Oh wow — I love Univers and Eurostile, makes Berkeley Mono a contender!

The other cool thing about it (not sure if this is common) is that you can customize your variety of the zero character and a few other characters that have debatable or preferential forms

I ended up forcing my browser to use Atkinson Hyperlegible for sans-serif text and Berkeley Mono for monospaced text and I may never change this config lol (although I wish I could force sites to use just the fonts, but not the fontsize settings, in Firefox...)

Glorious!

Browsing around, the popular monospaced fonts for coding already have high distinction between glyphs like 8Bi1l.

Also 0/O and '` often matter. There are probably others that don't come to mind immediately.

Everyone has a favorite for clarity (mine is Andale Mono) but see the hyper-configurable https://typeof.net/Iosevka/ to make one your own.

Perhaps surprisingly, the ancient Monaco is still a good choice for development work. It has the key features of distinguishable 0/O and 1|Il. (Which you may not be able to distinguish here, depending on font choices).

The designer/YouTuber Linus Boman has a video [1] about this font and his involvement in the project, it's a pretty interesting watch IMO.

[1] https://www.youtube.com/watch?v=wjE5eHLICzc

Linus' channel is severely underappreciated. I especially enjoyed his dive into the typography of comedy club logos: https://www.youtube.com/watch?v=XPmlv7cDk6Q.

Also worth checking out is the Lexend family: https://www.lexend.com/

I love seeing this push into legibility!

Yes! I auditioned dozens of fonts for the UI of https://figure.game and chose Lexend for its legibility, even at small sizes. It’s also got lots of understated charm and character (no pun intended) in my opinion. After reading about the project, I was very impressed with the reading fluency improvements reported, especially considering that (to me) it just looks like a very classy contemporary geometric sans.

It is more legible than your average "font-family: sans-serif” (aka Arial, Roboto, Helvetica) but it’s not that much of an “original idea”. For example Frutiger by Adrian Frutiger, a type family that a lot of airports use as it’s designed with legibility at long distances, or even the Verdana by Matthew Carter, ubiquitous web font that powers this and a lot of other websites’ typography that has specifically designed for legibility on low resolution screens, aren’t that less legible than this font. Still, I like the idea that some people on Braille Institute has decided to commission a typeface with a “free” license.

It looks almost exactly like Helvetica.

I was talking to my opticians recently and he said "Everyone gets cataracts, PVDs[1] and macular degeneration, if they live long enough". So we should all be glad that people are working in this area.

[1]Posterior Vitreous Detachment.

Over the past 2-3 years I went from needing reading glasses for small print to needing reading glasses for nearly everything in "normal"-sized print. I even keep HN at +125% (thanks, Chrome) because the default is too small. I still need my reading glasses at this magnification, but I don't want to magnify it too much more because of limited screen real-estate. This is just from normal age-related presbyopia.

I'm hoping that as the tech-savvy population ages, more thought will be put into design for vision challenges. Honestly I very low expectations, though.

Tech is mostly designed by young white English-speaking males and they tend to design with young white English-speaking males in mind. On the plus, this means that there are quite a few unexploited opportunities.

> Tech is mostly designed by young white English-speaking males

Indeed, and readability is one of the less pressing problems that result from that. It's also why I have marginal hope, at best.

How effective is the German FE-Schrift license plate typeface for this purpose?

https://en.wikipedia.org/wiki/FE-Schrift

It was designed to make different characters visually distinct from one another, but also especially difficult to alter into something visually similar. The latter part is a slightly different goal.

Maybe it would work, but I can't see myself ever using a font without lowercase letters.

If anyone else is having trouble loading the page: https://archive.ph/ok1ro.

Interesting that the slash on the zero character goes in the opposite slant direction than you normally see?

Also, in the middle of that webpage, they show how the shapes in "ER79jr" are slightly different from conventional (they didn't notate it but I assume the yellow background alternate shapes are what are usually seen), but could have used some better explanation why those alternate shapes are confusing.

I imagine the "backslash" in the zero is intended to make it more distinct from Ø and ø.

I assume this is to differentiate with 'Ø'

As an "Ø" language user, the 0-with-backslash is very wrong, it makes my brain say "syntax error". It does definitely not help reading speed.

It would be much better using a 0-with-dot.

Feels odd to use such a weird variation when there is AFAIK only a single language using Ø and we never have problem with 0 being mistaken for Ø.

Just FYI, I downloaded the font (after some time) and put it on my Kobo. When I tried to use it my Kobo just rebooted. Bad news, everyone.

But good news, everyone. It's available from Google at https://fonts.google.com/specimen/Atkinson+Hyperlegible and this one works with my Kobo.

I don't know why, I'm no font expert, but it does make for nice reading.

Great idea! One comment though - on Firefox when I try downloading the font, it disallows the download stating that it is a security risk. That is because the font is being downloaded of a HTTP url rather than an HTTPS url.

I had to manually turn off that restriction in the firefox download to get the font downloaded. May be a good idea to have the download initiated from a HTTPS url

I made a 3px height font by giving the letters a color but couldn't figure out if the experiment was a success or not.

http://synesthesia.go-here.nl

The link is down for me. Here is a cached copy so you can at least see what all the hype is about: https://archive.ph/ok1ro

I am happy to announce;

i wrote a GDOv7 module to ship this font and will use it in one of my projects.

Of course, this projekt uses only Latin characters. (The fonts are like 20kb, i doubt they feature something asian.)

Wow, this is extremely well-designed and legible. Almost perfect, but I found one small issue in only one example on the screen. I read "B8 1Iil" as "B8 1TiL".

I'm glad someone finally took a stab at ‘8’, ‘9’ and ‘6’—I regularly have to lean or whip out the screen magnifier in to tell which one it is.

Do they have data backing up their claims? Also, how does it compare to other high legibility fonts such as Highway Gothic?

It seems like a promising font, but it needs testing.

> download and change the world

> 335 glyphs per font

Sounds like much of the world was left out.

Note: I wouldn't have brought this up if it wasn't for the hyperbole of "change the world"

https://news.ycombinator.com/item?id=26011945

Curious: was this tested on users with various kinds of vision deficiencies?

That study would be super interesting !

"Please enable JavaScript in your browser to complete this form."

Any noscript/basic (x)html download link?

Here it is https://brailleinstitute.org/wp-content/uploads/atkinson-hyp...

thx, did I miss the link?

I use this for my IDE, I dont have any vision issues or anything, its just really nice to look at :)

Maybe opticians should use this as their eyesight testing font. If they don't already.

https://github.com/denispelli/Eye-Chart-Fonts says "The Sloan font file was created by Denis Pelli based on Louise Sloan’s specifications and used for the Pelli-Robson contrast sensitivity chart (Pelli, Robson, & Wilkins, 1988).

Louise Sloan’s design has been designated the US standard for acuity testing by the National Academy of Sciences, National Research Council, Committee on Vision (NAS-NRC, 1980)."

One of the goals of eyesight testing fonts is to make it difficult to distinguish between certain letter shapes unless your vision is good enough to see them.

Their EULA looks similar to the OFL. I wonder why they didn’t just use that.

This seems like a great use case for those long zoom meeting codes.

Did we hug it to death? Not loading over here

EDIT: it loads now

Really is quite a nice font

I like this idea, but when I go to the web page...

> Download the Font… and change the world! By downloading, installing and/or using the font software, you confirm that you have read and agree to be bound by the terms of this End-User License Agreement...

Sigh. I guess I need to read that EULA.

Anyway, clicking on that then downloads a pdf rather than a nice link to some web page text. Sigh.

So then anyway, trying to open that pdf then fails in the first two readers I tried. Sigh.

Why must everything in the world be so hard? Fuck it, never mind. I'll let someone else change the world and wait for widespread adoption for this font to reach me in whatever other ways.

Not sure why it wouldn't open for you, but it opens fine in Chrome on Mac. I thought it would be some ridiculous 50 page legalese but it's actually only 1.5 pages long. it's basically a license similar to what you'd see on a Github repo.

But why PDF at all? Is there a reason to put the online license in a PDF? Or any other places, like the restaurant menus? is there anybody who wants/needs to print the license or the restaurant menu?

My guess? They use an agency to service their website and made the license in PDF so the org can change it without incurring additional cost of an update.

To OPs point, a .TXT would work just as well.

Orgs place a lot of value on control. A TXT, even if better, would be a hard sell. Especially in a 501(c), where optics matter and they are targeting a demographic that usually prefers PDFs.

I think you probably meant to write "can't change it without".

I was alluding that the Org has the ability to re-upload the pdf in a shared resource without the need of a developer.

Vs making the license file a static page that would incur billable hours to update.

I'm sure you can upload HTML files in shared resources as well. Or not?

I kind of understand it for restaurant menus. Both "real" restaurant and takeway places need to print actual menus and/or flyers, and it's just easier to create a PDF document and upload it to your website than it is to reformat the document to be useful as web page in many cases if you are not well versed in this. Since PDF readers in browsers got decent, this isn't too much of an issue anymore anyway, IMHO.

Not an issue indeed if I go to the restaurant table with a computer. Which happens very rarely for me, if ever.

So organizations that publish documents can store it for legal purposes. A pdf has useful metadata that a text document does not, and can be easily authenticated. If there's a dispute everyone in the legal process understands what a pdf file is.

I just wonder how do all other 99% of organizations where the license is a MD file only handle their court cases...

> Or any other places, like the restaurant menus? is there anybody who wants/needs to print the license or the restaurant menu?

Web devs are expensive, sadly, and not everyone has the patience to figure out the DIY platforms (Square, SquareSpace, Wix, etc.). It got a bit better during the pandemic when takeout took off. But generally restaurants need to print menus anyway, so it's just easier to upload that design as a PDF than to redo it for the web with labor they don't have, or paying for it with margins they don't have...

Or rather, they just take their Word menu and "save as PDF". Although they could also "save as HTML" and replace the same file, right? But then the Word formatting would get messy because it wasn't thought for a web page, so here we are.

I really hate those restaurants that only have QR codes on the table for menus. A restaurant is for eating and talking, not for messing around with my phone.

I avoid those now. During Covid there was a slight point (though it was pretty soon proven that surface transmission was so minor a factor not to warrant such actions). But now we should just go back to normal menus.

Heh. I went to a sushi restaurant recently and sat down to an iPad menu. After ordering a few rolls and confirming the order, it was delivered by a robot[0]. After finishing the plates, I returned them to the robot which carried them away. I scanned a QR code on the table which allowed me to pay for everything from my phone.

Curiously, I was presented with options to tip 18%, 20%, 30%, or other. I thought about it and tipped accordingly to the effort the waitress put into my experience.

Shortly after, the manager came to me (essentially the first point of human interaction) and asked if there was anything wrong that caused me to tip 0%.

So yes, I too really hate the direction restaurants are heading in.

[0]: https://www.pudurobotics.com/product/detail/pudubot2

I'm curious who would get the tip in this scenario. The sushi chef? Ok, I guess. The restaurant owner? I'm going with 0%. The robot? Maybe if it can pass a Turing test.

> The sushi chef?

Yes, and most likely split with other cooks, host/hostess, and waitstaff for people that refuse to use the iPad.

> The restaurant owner?

Illegal in most states.

What an iconoclast, refusing to tip hourly workers...

The social contract is currently approximately 'when I interact with human wait staff I am expected to add some cash to what the bill says on it' This is evident from, for example, the drive-thru example. There is no waiter / waitress, you just get your stuff and keep rolling. A tip jar is sometimes available but is rarely expected. If I never interact with a human and the restaurant expects a tip I think they should tweak their presentation to more closely adhere to that social contract.

In actual fact I wish we could stop dancing around the fluffy idea of when to tip and when not to tip and just expect restaurant owners to pay their staff and include that in the price charged for the goods. It's super annoying having to think about this stuff all the time.

Except that in most areas, tips are legally required to go end workers (not management or ownership) and in jobs where tips are allowed, end workers make more than in service jobs where tips are not allowed on average. We're talking about America, I don't know why you would expect restaurant owners to pay workers a dime over the market clearing price. The chances service employees get employer provided health insurance is very low.

I don't know your financial situation and noone is forcing you to tip. But if you make, say, more than 3 times what the people servicing you make, and you choose not to tip, I think that's quite selfish.

...Wages are also legally required to go to end workers.

I'm suggesting employers pay their employees. This isn't a radical concept and works just fine everywhere else. They already do this everywhere other than restaurants. If they aren't paying employees enough, that's a problem. You don't solve that problem by adding an optional tax on customers that is kinda-sorta-not totally required. All that does is ensure that in a certain percentage of cases employees get paid zero percent of the invisible tax. I'm just saying make the tax visible and required and stop this weird dance we do in america.

The comment you're replying to describes how they wish the world would be, after recognizing the reality of the world as it is.

Tipping should be illegal. Tipping is necessary currently most places. Both can be true.

Only in the past few years has it been remotely expected that you'd have to tip someone who didn't actually take your order at a table and then walk your food out from the kitchen.

We've all been tricked into tipping in situations where, honestly, the management should be paying better. And yeah, I know that means that the cost of things will be higher on say the menu, but the real cost of a $20 dinner should be $24, why not just pay the $24 up front and raise the base wages to what they should be.

I'm pretty sure this is only an American lunacy too.

Yeah but if you are in America where higher minimum wages or public healthcare are not happening anytime soon, maybe you should tip the hourly workers if you can afford it?

I think people just don't want to be confronted with the idea that the people servicing them make a pittance. It brings up uncomfortable feelings to have to decide how much a service employee "ought" to be paid, each time you eat at a restaurant.

Those hourly workers are getting screwed by an employer who won't pay them what they're worth. It isn't my responsibility to feel shame and make up for their circumstances with an arbitrary tithe.

I heard he also didn't tip the hourly cashier at the grocery store, what a jerk (and they even had an in-person interaction).

> What an iconoclast, refusing to tip hourly workers...

If everyone refused to tip, wages and prices would adjust.

(Tipping isn't even that old in the American culture. Perhaps a hundred years or so. And its introduction was seen as thoroughly unAmerican, and more in-line with those supposedly obedient Europeans groveling for handouts from their social superiors, instead of a proper American transaction amongst equals.)

> What an iconoclast, refusing to tip hourly workers...

He did tip the hourly worker who served him: he gave a tip that was a percentage of that servers wage.

What's unfair about that?

I agree with tipping the robots (for whom service is always the same) N% of their wage; which is still 0.

Tips should be _not-legal_ and proper wages for the workers should be _required_.

Lol. What’s next, they’ll want tips for the dishwashing machine too?

ill put in my robotmancipation notes

> ill put in my robotmancipation notes

Missing episode in The Animatrix: the uprising started when robots didn't get tips.

We have those in Singapore as well, but you are not expected to tip.

(Tipping in a Japanese restaurant is weird anyway. People don't do that in Japan, and they don't do it in any of the authentic Japanese places I've been to.)

IPad ordering is fairly common here. I like it, very efficient.

There's a sushi place around here that does iPad ordering, but they deliver your food on the customary conveyor belt, not via robot.

I don't mind this when it's done well. I'm going to have to mess with something to find out what food the restaurant sells and my phone is about as good an option as any. It can allow easier menu changes, inclusion of specials, and the like.

In practice, however most restaurants seem to go for a PDF of their previous menu design formatted for printing on a paper many times the size of my phone. That's a bad experience and restaurants that do it deserve to lose business over it. I'm reminded of about a decade long period not long ago when a restaurant website was more likely to have animations and music than the hours and menu.

Menus on phones blow. It's a pain in the ass to fussily scroll around on a dinky screen and try to remember what has scrolled away that you were interested in, instead of being able to glance across a page.

I sympathize with the PITA of changing printed menus, but that's part of what you pay for. If I go out, I'm not paying for a lunch-counter experience.

E-paper menus would be a pretty good compromise, if they could be made practical.

Sometimes now this is because of supply problems that make a single, consistent menu embarrassing to the business: nobody wants to have to say, "I'm sorry; we're out of salmon and eggplant, and we haven't had coconut milk for a week. Can I offer the three of you something else?" An electronic menu can be changed immediately and at practically no cost.

An electronic menu can also have prices that change every hour based on how busy the place is or even have prices based on demographic information gleaned from the user's phone.

What restaurants do this?

International House of Hypotheticals, Strawman Cafe, and the Red Herring Bar and Grill.

+1 for you, sir

Great now i want to go to a fish grilling bar.. Thanks a lot.

None that I'm aware of, but it WILL happen.

This sounds like over estimating a lot of restaurants ability to handle electronic menus.

I don't think I've seen that functionality in many restaurants at all.

There is always a cost in updating the menu.

A case in point:

https://www.sixteen-nine.net/2019/08/15/paper-and-tape-are-n...

I see your point but I've been told so many times "sorry sir we're out of salmon and eggplant" and nobody seemed to care except me regretting the salmon and eggplant. So, not sure it's a real world argument.

i actually prefer it. I've never been to one here in Austin that doesn't have a paper menu if you ask tho. I much prefer handling my filthy, but personally contaminated phone, to menus that a ton of other people have handled after rubbing their eyes and noses.

PDF works better when downloaded, sent over e-mail, printed to be used in a court case or hard archives, and has stood the test of time.

99% of the browsers in use have a built in pdf reader these days.

Which is exactly why I'm not complaining that I cannot open it. I'm complaining that the experience sucks - if you have a mobile phone you know what I mean. Unless you go with a computer in the restaurants, of course.

> But why PDF at all? Is there a reason to put the online license in a PDF?

Yes. It's so that you have a dated copy. An web-page with the license on it can be retroactively changed.

Are you distinguishing between a PDF that downloads by default (rather than using an in-browser preview)? Because I’m that case, you could download the license page just the same.

> Are you distinguishing between a PDF that downloads by default (rather than using an in-browser preview)? Because I’m that case, you could download the license page just the same.

The metadata on the license page does not offer the same guarantees as the metadata in the PDF.

The PDF can even be signed; the license page cannot.

You can save web pages as easily as you can save PDFs. You can also edit PDF files with common office tools. The only thing PDF works for is ensuring the rendered copy looks exactly the same everywhere and to add a signature but I don't see the value in those two options for this use case; if anything, this institution should let the user agrnt decide how the text is remdered best and we don't know their public key to validate their signature.

It downloads rather than opening in your browser because that’s how you’ve configured your browser. For me it opens in the browser (albeit in a new tab). There’s nothing wrong with the PDF, I can open it with all the PDF readers on my machine; it’s a basic v.1.6 PDF file. This is, in fact a “nice link to some web page text”: a page of text served over the web.

I think linking a PDF is less friendly to users who need Text-to-Speech technology, something the Braille Institute should be sensitive to...

What font does the PDF use?

Maybe their Atkinson font?

It does indeed :)

Kinda funny that you need to read the EULA before downloading it because once you have the PDF on your computer you have already downloaded the font.

That’s a good point. I have no idea if Text-to-Speech works with PDF.

It does, and in many cases even better than on most web pages.

> By downloading, installing and/or using the font software, you confirm that you have read and agree to be bound by the terms of this End-User License Agreement

... says them. I don't agree and I don't confirm. I can download the font without accepting/consenting to whatever text they put on the web page. And if they're not ok with that - they can limit the availability of the download to only those who accept whatever weird terms they put in that PDF you mentioned. Except they won't do that, since that would harm adoption of their font.

More about this kind of "licenses":

https://en.wikipedia.org/wiki/Shrink_wrap_contract

You can clone a GPL-licensed repository without clicking any checkboxes either, but if you violate the license, you can expect to be sued.

Shrinkwrap/clickwrap is when the license is displayed after you pay for the product. The concept doesn't and cannot apply to freely downloadable content.

> Except they won't do that, since that would harm adoption of their font.

Is it because it would harm adoption or because it's infeasible? You can download copyrighted images and use them for your own profit, copy github code with GPL code for your closed-source project, or sign pretty much any contract (say for a car or a house) without actually reading it. Just because there is no safeguard to ensure you are really read what you "agreed" to, doesn't mean it's unenforcable.

Moreover, the terms are very simple (1.5 pages) - hardly some "weird terms" to define how you are allowed to redistribute intellectual property.

>And if they're not ok with that

Then they can sue. EULAs often state things that don't matter in the jurisdiction of the end user, yes. But copyright is not related to that. If they hold the copyright to the font, then they can set the rules in the license, and if those don't clash with the local laws of your place, then you're liable legally.

And yes, they probably won't do that. But that's, again, a different matter, other copyrighted works are also always stolen.

Even if the downloader was in the US, the results of a lawsuit are at best a toss-up considering case law. Internationally, I'm not even sure such a suit has ever been filed. Has it? A link would be interesting...

> Why must everything in the world be so hard?

linux user?

On FreeBSD with Firefox this PDF opened right away, just saying

Loaded for me too, zero issues.

Firefox and chrome will both open the PDF by default unless you tell it to use your own reader.

Both gnome and KDE come with a capable reader better than your browser.

If you don't like those there are several options.

Reading a pdf under Linux is super easy.

if you want "everything in the world" to not be "so hard" you'd probably avoid Linux.

I had such a different experience. I found the EULA surprisingly readible and easy to understand. Just 5 easy bullet points rather than a multi page document of legalese.

That could have been in-page information rather than a PDF. Or even better, get rid of it. EULA for a font is a wishful move.

The EULA is giving you permission to e.g. copy and modify the font with certain restrictions (e.g. include the license text if you distribute it). In that regard, it's not terribly dissimilar in spirit from an open source software license.

Why is it not an actual open source license? Perhaps CC-By or something?

CC-By would require attributing on any page you use it on, right?

Their license simply says to include the license if you re-distribute the font.

The distinction is more between an EULA vs one of the common open source licenses.

Aren’t you re-distributing the font by using it on a webpage? Well, assuming you self-host or provide a third-party link.

They could have used the SIL Open Font License instead. People in the open source world are often already familiar with it, and resources like tl;dr Legal are available for it.

Their EULA is, in fact almost exactly the SIL Open Font License, but someone decided minor changes in wording were more valuable than standardization.

https://scripts.sil.org/cms/scripts/page.php?item_id=OFL_web

https://tldrlegal.com/license/open-font-license-(ofl)-explai...

If you go to the google fonts version of this font ( https://fonts.google.com/specimen/Atkinson+Hyperlegible )

It DOES use SIL Open Font License. Maybe it's dual licensed?

I was showing this exact problem to my wife, a moment after trying to order something from The Container Store and their website being broken, just after trying to pay a bill and that website not working too.

I swear to God this f**ing industry.

Reminds me of the parking at my local lake were you now have to read a 60 page EULA, a 30 page data "protection" agreement and install an app which requires way too many permissions. I happily paid the quite reasonable parking fee in the past when there still was a human person collecting the money. She was very friendly and always had an interesting story to tell, but now she's out of her job. Anyway, the surveillance cameras have been vandalized and parking is free now.

>Anyway, the surveillance cameras have been vandalized and parking is free now.

This totally reads like you just did the vandalizing. Just saying ;-)

Do you have an adblocker on? It will frequently break sites especially in the checkout process I have found (probably because their crappy code calls some tracking/analytics script without try/catch and can't continue). 90% of the time disabling it for the page will fix these broken sites.

> I swear to God this f*ing industry.

this again. Developers and any publisher should test their stuff with ad blockers enabled.

Why? They test in default configurations. As someone who changes many, many default settings, I don't expect developers to test in every possible combination of non-default settings, browser extensions, atypical devices, text zoom settings, etc.

I'd bet there's a pretty solid correlation between using an ad blocker and propensity to spend money, if only because the good ones only run on desktop browsers. Someone who owns a keyboard is likely to be white collar if they aren't a student.

So if you want to leave money on the table because some dumb script which has no bearing on selling the product breaks, go ahead and damage that funnel.

These analytics scripts aren't critical and the checkout process shouldn't go down just because they don't load. It could be a user on spotty wifi or even the external service being down entirely. I agree that they shouldn't test all the functions of random third party extensions, but this failure point is pretty foreseeable for reasons besides adblock.

Your website breaking while someone is actively trying to spend money on it is very very bad. It means someone who was definitely willing to pay you is not paying you. Ecommerce companies spend a lot of money on optimizations intended to keep users from abandoning the checkout process. Single-digit percentage improvements result in large increases in revenue. Ensuring the checkout actually functions is about the least a site can do to ensure users don't abandon the process.

A quick web search suggests somewhere between 25% and 45% of web users block ads in some way in 2022, depending on who, where, and how you ask.

Thanks for troubleshooting for me but no. Just a stock popular browser.

I'm always coming across airline and hotel sites that don't work for days at a time, or which don't work in Firefox etc. Latest was EVA Air a week ago.

When I think about the money these kind of sites in particular must be losing from shoddy engineering... I honestly don't know how their devs get away with it!

I’ll tell you how, because I had a client in this exact situation. The business just sees fluctuations in sales. They don’t see error rates. In traditional businesses SW engineering keeps the dynatrace logs close to the vest.

I convinced a dev to walk me through their dynatrace console. I dug in and took screenshots then gave them to h to the business. The business demanded access to dynatrace and then was able to correlate error rates with revenue fluctuations (down to the hour of the day, people are habitual). They came up with a cost factor and figured out how much they’re losing. This led to budget being allocated for a project to solve these problems. I bid on the work but was denied. The VP of IT at this point hated me.

PDF also cannot change from underneath you, unlike a web page. Better to have a copy of what the license was at the time you agreed to it, imo.

Pages can be saved in modern browsers about as easily, if they aren’t actually js apps in disguise. Never had a problem doing just that for plain html pages like this could have been.

PDF can change the same as any other format. Yes, line length in an html file if not specified is left to the browser, but css exist and also there are other formats, like txt.

I guess it depends on what wrote the PDF as to how much can be edited later, but I'm constantly opening/modifying/resaving PDFs.

ff on linux w/ various ad/script blockers can't ever get that far. a never loading page that proclaims itself to be "privileged" on my machine.

nope.

I don't know dude I clicked on it and it just opened in a new tab? I don't think it gets much easier.

Reply on news.ycombinator.com