227 comments

[ 3.1 ms ] story [ 247 ms ] thread
Not bad. Can’t see if they have a fixed width version though.
I made a custom build[1] of Iosevka[2] that chooses glyph variants based on Atkison Hyperlegible. I like it a lot and over time prefer it to the other Iosevka stylistic sets.

[1]: https://github.com/jdknezek/Iosevka/blob/jdk/scripts/hyperse...

[2]: https://typeof.net/Iosevka

Since I discovered Iosevka a few years back it has been my daily driver for 80% of the time. I use other fonts from time to time but I always go back to my custom build of Iosevka.

Yours actually looks a bit more readable than mine, gonna check it out your build plan.

Thanks!

It reminds me of IBM Plex, if you're just looking for similar looks.
> Please enter your email address to begin downloading.
This sentence is indeed written on the linked page. What exactly is it that you want to express?
It's available as a package on all the linux and BSD distros I'm running.
They don't tell you you can put anything with an @ and a . in that field.
Yea email fields without required verification get my maximum character length pseudo email.
Why not your SQL injection attack email?
That's a hard nope, for me.

It's only a "nice-to-have" font.

I have an OG mac.com email, and some woman used it to sign a donation to ASPCA, a number of years ago.

They then sold it to some liberal group, who then sold it to some ultra-liberal group.

I get dozens of extreme-left emails, every day. It's crazy.

I'm not extreme anything, and want to puke, just reading the subject lines (for the record, the extreme-right is an even more effective emetic).

I would be interested (but not really that interested) in finding out who these folks would sell my email to.

Pretty awesome font, I've found it pairs really well with other readable fonts like Lexend too.
While the characters are certainly distinctive, I find paragraphs to actually be less legible than, say, Times New Roman.
> It improves legibility and readability

Agreed.

It seems to improve legibility. I'm not so sure it improve readability.

Readability is largely influenced by the x-height (height of lowercase letter), because your brain is trying to discern the "shape" of a word - which allows you to read faster. Hence why ALL CAPS is difficult to read because the word shape is the same.

Atkinson appears to have a high x-height, which reduces readability.

I think SIL's Andika font [1] is designed with similar goals, although I think informed more by teaching reading. I didn't see any large samples, but the font seems a little subtler. Presumably, given the name, the Braille Institute's readers are more visually impaired than SIL's, so for those us with less impairment, maybe SIL's would be more readable?

OpenDyslexic [2] is another interesting font. I've heard that asymmetry helps dyslexic readers, and (given a cursory glance) they seem to have done a good job having asymmetry and style. I've also heard that SIL's font is helpful for dyslexics, but that was a while back and now they have multiple font styles, so I'm not sure if the one above is the one or not.

I didn't realize this until I worked on some software for dyslexic readers, but the normal computer font makes bdpq look pretty much identical except for rotation. It has a minimalist elegance, but for new/foreign readers, it's pretty subtle. I remember learning Chinese and had to spend quite some time with two characters, which differed only with one have a dot and one a short vertical line at top (a dot looks pretty similar to a vertical line at a cursory glance, as it is more inverted teardrop shape than a circle). After that the western alphabet looked a lot more uniform than I'd originally thought.

[1] https://software.sil.org/andika/andika-and-the-visually-impa...

[2] https://opendyslexic.org/

Andika is beautiful, thanks :)
There's a font apparently based on Andika with some disambiguating improvements, called Cadman.
I'm not a font or readabiliity expert, but it sounds like that is the tradeoff they had to make; paragraph legibility is not as important as being able to make out the words in the first place, and people with visual issues will struggle to read even if the font is better suited for them.
See also the font designed for readability* in Airbus cockpits

https://b612-font.com/

The header image on that site seems to show two font variants (slashed-zero and non-slashed) in a single display. This puzzles me!
I see the slashed zero on the right, but I cannot find a way to get that out of the Google font display. Perhaps the slashed zero is actually a different font?
Found a git issue that indicates the slashed 0 is in private use area of the font

https://github.com/polarsys/b612/issues/20#issuecomment-5448...

To answer the grandparent, unslashed 0s are used in numerical figures (speed, pressure) while slashed 0s are reserved for the alphanumeric figures (headings and other idenitifers)

I saw that usage pattern, but then noticed that zeroes in times are slashed.

So, "slashed in strings even if non-alpha, but non-slashed in numerics, and times are strings"? Seems viable enough, but I'd choose consistency instead. :)

I guess we'd have to get into whether clock time is ordinal or cardinal :^)
It's a good one to add as a custom font on Kindle/ebook devices
Is there a monospaced version, or an inspired monospaced version based on this available?
I need that too for hyper-OCR-able monospace labels.
Not that I've found, but I use Commit Mono alongside it with good results

https://commitmono.com/

"Safari download not working, try Firefox or Chrome"

You gotta be joking.

I know right? Who uses Safari nowadays?
A website where a download link requires FF/Chrome. The whole website I thought was silly, but that takes the cake.
I looked at the code to see why it would be doing that. It seems to handle the font customization stuff by basically downloading all of the variants and then combining them into a zip file in client-side code - even when you do no customization at all. Apparently that code which makes the zip (which I'm guessing is an external library) creates a corrupted one when run in Safari according to a comment buried in the JS.

Maybe it really is a bug on Safari's part but creating custom zip archives is something which would be far saner to do on the server side in the first place.

that'd include paying for computation time vs just static content on a CDN
yeah it's obviously a low priority for them but a monospace version of this font would be very interesting.
I made a custom Iosevka build by selecting glyph variants based on Atkison Hyperlegible.

Iosevka: https://typeof.net/Iosevka

"Hypersevka" build plans: https://github.com/jdknezek/Iosevka/blob/jdk/scripts/hyperse...

Screenshots: https://imgur.com/7BZS3Pp https://imgur.com/sudNqWM

Iosevka is absolutely wonderful!

I have my own builds -- not based on AH's glyph choices, but also chosen to minimize glyph ambiguity.

I'll mention another great legible monospace project: 0xProto

---

Build: https://github.com/AndydeCleyre/archbuilder_iosevka/releases...

Screenshot Mono: https://cdn.imgchest.com/files/k46acrxl297.png

Screenshot Mono with syntax highlighting: https://cdn.imgchest.com/files/l4nec9d8lm4.png

Screenshot Proportional: https://cdn.imgchest.com/files/6yxkce8p6a7.png

0xProto: https://github.com/0xType/0xProto

I really like the aesthetics of Iosevka, but the glyphs are really narrow - resulting in severe readability problems for me (I'm diagnosed dyslexic).
You can configure your build to be wider than default. In order of increasing width:

- semi-extended

- extended

- extra-extended

- ultra-extended

I think by default the extended variant is included in most builds anyway, if you want to try it.

FWIW here's a sample of my usual build's extended variant: https://cdn.imgchest.com/files/b49zcjd53oy.png

I'll add that https://github.com/0xType/0xProto is worth checking out, if you haven't. What mono (and proportional) fonts do you find work best for you?
* MonoLisa was pretty good, minus the cursive. As a late/16yo diagnosis under a British curriculum, I cannot express in words just how hostile cursive is. Its continued use in society is an embarrassment /rant

* Intel Commit.

* I am currently using Maple Mono.

I recently replaced mine and ruled out a bunch of others based on just two rules:

- the zero must have a mark - the base of the lower case L must go to the right and not the left, for better distinction from the number one

Related. Others?

Atkinson Hyperlegible Font - https://news.ycombinator.com/item?id=32799872 - Sept 2022 (234 comments)

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)

You know, this feature could be a little side bar on the top right, or just as a header for every post. (I assume its semi automated, but I would be happy to see it all the time)
Isn't that the point of the "past" button though? I've always assumed the purpose of Dang's "related" comments were to help bridge the discussion.
It seems as if the past button allows you to view a snapshot of the site on a given day, whereas Dang's bundles are previous instances of the topic/link being discussed on HN.
There is also a past button for submissions
Thanks for sharing! I'd never thought to click through and see
Me either! Maybe if dang included a "past" link in the comment, more of us would learn about it.
Before I post something on HN I always do a search to see if it has been posted recently. If it's more than 2-3 years old, I post it again in order for new users (and others who forget it) to notice it.

There have been other similar suggestions i.e. when filling the information for a new submission to have an automatic duplicate search and produce a warning, but I've come to understand that adding features to the simple/lean HN page is not desired. Same goes for native dark mode (with having to load scripts/plugins/whatever).

I can understand Feature Creep (https://en.wikipedia.org/wiki/Feature_creep) fears, but it seems that here we're doing the exact opposite :-)

Agreed. The 3rd party app I'm using to read HN on mobile (Glider) has this feature and it's both useful and unobtrusive.

Screenshot: https://www.imgpaste.net/image/SDoBg6

I use Glider too and this is the first time I noticed this feature. Interesting!
Clicking the URL next to the post title provides this, at least for fairly unique pages. It’s less useful when the link is from a popular news site.
I wonder if making old comments too-easily available backfire by reducing new takes and fresh discussion. Probably not in this particular case, it seems like a potential risk when designing a social news site.

P.S.: I choose to believe the neurons that instinctively throw up "but what if it's actually bad" represents a useful software design skill, rather than just a personality flaw. :p

Atkinson Hyperlegible appears to slash their zeroes in the same direction as backslash, unlike all other fonts I have used where the slash in slashed zeroes have the same direction as a forward slash. Not sure if this was a deliberate design choice.

https://fonts.google.com/specimen/Atkinson+Hyperlegible?prev...

https://fonts.google.com/specimen/Inconsolata?preview.text=0...

(comment deleted)
Thanks for pointing this out. It looked weird to me, but I couldn't quite put my finger on why.
Probably since that it's intended for legibilty, it's to differentiate 0 from Ø.
>it's to differentiate 0 from Ø [a letter used in Danish, Norwegian]

most importantly it's to differentiate it from ∅, mathematical null set.

"computer people" feel this understandable need to differentiate their zeroes, but "CS-is-math people" are like "but don't use null set!" <— ᵇᵘᵗ ʸᵒᵘ ʲᵘˢᵗ ᵘˢᵉᵈ ᶠᵃᶜᵗᵒʳᶦᵃˡ

Is the exclamation mark character for factorial different from ASCII x21?
Hmm I think there are probably more Danish and Norwegian people than mathematicians writing about null sets.
No way, DK and NO aren't that big, and you learn about empty sets in HS in most of the world. I'd wager there are currently more Indian high schoolers than people in DK/NO.
In Germany (>20 years ago) we learned to use {}.
> you learn about empty sets in HS in most of the world

lol no.

You're (probably) correct, considering that the specimen used Ø instead of ∅ (see https://braileinstitute.app.box.com/s/gw2s8k5kfc6t1uabkiyqlc...), and the font do not have the null symbol.

(also like the German commenter the null set is an empty set = {}, I only encountered ⌀ as diameter symbol, actually)

Wow, what ever that website is doing manages to completely and utterly break my gpu rendering in a way I haven't seen a site do in over a decade. It just paints the text over whatever was there last, including other windows above my browser.
I use DejaVu Sans Mono for programming, it places a dot inside 0.
Possibly idiotic question: I'm fairly new to actually writing code myself in any thing more than a tinkering way. What do you use for programming where you can choose the typeface?
You mean a text editor where you can change your font? I am pretty sure most of them support that feature, VS Code just to mention one. Or do you mean other kind of tool?
I just tested it (have tested it before, always looking for mono fonts)

DejaVu Sans Mono does not look anywhere near as good as Lucida Sans which is LibreOffice Calc's default for me, comparing just the numerals. The digits in this variable pitch font are fixed pitch so it's suitable for spreadsheets.

(I mentioned the extra marking on zeros in the spirit of the thread, but personally I don't care about it, I'm not mixing up letters and digits much irl or especially in a spreadsheet.)

but in any case, the DejaVu Sans Mono glyphs rise up above the baseline, floating in the cells, and they are larger for a given pointsize. I don't know whose fault this is, but my complaint is that every time I try to look at different fonts, the futz-around-and-find-out factor is so great it's just a waste of time. It's a problem across systems (windows, linux, mac), across generations (Motif, 3.1) it's a problem across apps, it's just a problem that's completely unnecessary. I think font people spend too much time looking at their font on a tabula rasa wedding invite, and not enough time doing comparative font selection in ordinary apps.

My guess is that the other fonts do it to differentiate the slash zero and the Norwegian letter Ø.

If that's true then AH's choice seems counterintuitive to me.

Theoretically, a slashed zero should not have the slash extending beyond the main body of the character, but in practice, especially in sloppy writing (when the slashed zero is more common) it often does. Also when designing a font for someone who has difficulties distinguishing character shapes, small differences like that may not be immediately obvious. Using a reverse slash immediately shows that it’s a zero without any clashes.

I think it’s a fairly elegant solution at the slight cost of being a little unfamiliar (although the slashed zero itself is quite unfamiliar to most people).

I understand if this is supposed to be better for dyslexics, but the fact that the distinctions of the letters are so heavily emphasized makes it harder to read, since I generally read entire words at a time, not letters, and the individual letters are less important than how they look together in a word or a sentence.
I really dislike them having lower L just be a straight bar - And I'm saying that as someone that hates serif fonts with a passion.

It doesn't really help you while reading that capital letter i looks different. You have to (1) know which font you're currently reading and (2) know that a straight bar isn't a capital i.

This is kind aggregated by the fact that I'd naturally assume if only one letter looked like that it would be capital letter i. An assumption that is backed by this very website: I, l

Personally I find Comic Sans easiest to read, as dyslexic. I even use a monospace version of the font for coding and in the terminal.
Would that by chance be the Pointfree font?
Dunno. It's just called Comic Mono.

https://dtinth.github.io/comic-mono-font/

You might be interested to give Comic Code [0] a try instead.

Unlike Comic Mono it is a paid font, but I find it a lot nicer to use/read.

[0] https://tosche.net/fonts/comic-code

I use Comic Code in my editors and console windows and really like it a lot. Adds some practical frivolity to the process!
Oof that k looks way too much like r to me

EDIT: the font the website uses, not comic-code

That’s fascinating do you have any insights as to why?
I think it's related to how irregular the shapes of the letters are. The circles are all slightly bent out of shape in different ways, and the line strokes all have slightly varying slants.

Most attempts at making a font try to make all that stuff as regular as possible, even the one in the article we're discussing does this.

I am not dyslexic and I use this font too, because it's easier on my eyes. I find myself increasingly thinking "Courier" when I look at it and have to remind myself its not, and I chose it.
Likewise I've used Fantasque Sans Mono (related to Comic Sans) for a long time as a coding font.
I'm not sure this font takes into consideration dyslexia as it seems, from what I read, focused on people with very bad eyesight.
Yes, but it does include some features that seem to be included with dyslexia in mind. Letters like d and b have small differences so they're not mirror images of each other.
(comment deleted)
I like this idea, but for most companies, they've already decided on their fonts and many corporations already have strict guidelines on what fonts to use and many are either variants of popular fonts, or changed in some way that aligns with their branding.

I still think its a good step forward.

My go-to is Verdana based on previous empirical tests. It's bigger at the same pt size compared to other fonts, but if you size it down about 1/12th it is just as legible but at a higher characters per inch, which is good for trying to squeeze text into an interface, especially when that interface may be viewed by a group on a projector or screen share.

I just tried to half-ass a similar test by editing the Google Font explorer UI to put them next to each other. Atkinson beats out Verdana for width, by about 4% (eyeballed and envelope math), however it's also 1 pixel taller per line at that size. So it's a more rectangular font. I'd have to think about how I'd want to use it, if I care more about lines per page or characters per column.

If you’re a Verdana fan, you might appreciate this write up I did on it some years back: https://neosmart.net/blog/tahoma-vs-verdana/

It backs what you say!

I don’t recall why Tahoma didn’t win. The letters may have been too close together for comfortable reading, resulting in the need for a larger point for the same comprehension. I mentioned above in another reply that we had one dissenter in the taste test, I think now that they may have voted for Tahoma.
Tahoma may be the HN font but for me Verdana is and forever will be “the Encarta font.”
> My go-to is Verdana based on previous empirical tests.

Do you have any handy links around this? I've a passing interest in information-dense UIs and font choice is a vital but often overlooked aspect of such interfaces. Would love to see how others evaluate such things.

The empirical test was the entire team standing at the back of a conference room in shitty lighting basically taking a vision test. One person I believe voted for a different font, everyone else thought Verdana showed up best.
Is there any peer-reviewed studies about this or are those claims unsubstantiated?
No studies at all, all the claims are anecdotal at best.
There is some research which suggests that the human brain “learns” to read an unfamiliar font reasonably fluently after about 30 minutes of use, and that recognition continues to improve with continued use. The well-known saying “we read best what we read most” suggests that a font itself contributes only partly towards readability - your personal habits and experience are also significant.
I found it amusing that the end user license agreement is available only as a PDF on Box. That's not exactly the most accessible format but I got it downloaded. It appears to be a relatively straightforward license, prohibiting commercial resale of the font and its derivatives, but allowing it to be bundled with commercially sold software.

https://braileinstitute.app.box.com/s/rin3vzegmcy7sil28yfqsl...

It's actually just the SIL font license, which is what basically every open source font uses. They just removed the preamble of it for some reason.
Yeah, I wish they would just keep the boilerplate OFL intact. I spent way too much time trying to figure out what the "real" license was, and if I could use it in my open source projects.
It's obnoxious really, and borders on amateurism.

If they want the OFL, then use the OFL. Don't make me try to understand yet another licence; I'm not a lawyer.

Infuriatingly, they didn't actually use the same text but they introduced multiple tiny changes so that you're forced to check word-by-word whether it's still an open/free license.

Was it really necessary to replace "Original or Modified Versions" with "The Original Version or Modified Version", or "Reserved Font Name(s)" with "Reserved Typeface Name"? So necessary that you're willing to sacrifice potential reuse?

I found it amusing that the license PDF does not use the font. (At least, the slashed 0 in the license pdf is not the backslashed 0 on the font web page.)
What screen reader did you use? NVDA didn't really choke on it. The dialog for downloading is weird though - it just announces a dialog and then sign up/login links (which are double labeled). The weird spacing that a lot of PDFs have makes the speech synthesizer slightly mispronounce words right before line breaks though. I would say the PDF itself is more accessible than the process provided to download it.
That’s one ugly Å, without any separation. The font looks nice in general though.
In my opinion as a native Swedish person (where that letter is part of our language), not separating the ring from the A is justified only when the font size is very small.
(comment deleted)
Weirdly, I got eye strain from that page. I think it's not the font itself to blame, but the combination of its default size in my browser with being #000000 on #FFFFFF - too much contrast.
Same here. That page is not very easy on the eyes, and that's with my brightness and contrast tuned for comfort.
[flagged]
It's BrailleInstitute, an org catering to those with visual impairments. Understandably, their site is accommodating to their core audience.
It's the internet, a venue catering to those capable of detecting a simple joke.
So, um, this website expects a nontrivial amount of their traffic to be sight impaired. It probably isn't designed this way for your benefit if this is your complaint.
I appreciate that an expert institution designed a font for low vision readers with actual research. Every time I run an app that has OpenDyslexic as a font choice I die a little inside. It's the worst sort of "pretend to be helping" option. It actually is worse for reading than ordinary fonts. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5629233/
> “OpenDyslexic was compared to Arial and Times New Roman in three reading tasks: (a) letter naming, (b) word reading, and (c) nonsense word reading.”

Yikes, not a strong design IMO. I know very little about dyslexia, but reading individual words as a small child doesn’t seem like a comprehensive way to measure “how dyslexic people read.”

Presumably the primary envisioned use-case for a specialized typeface like this is paragraphs of text, read by people who have the language skills to ”think in paragraphs,” even if the visual processing is a challenge.

This is all armchair, I’d welcome correction or nuance.

I guess I’m the rare duckling who would report preferring OpenDyslexic! IMO it scans better, and is excellent when combined with beeline reader.

When I was at the ages of the sample, my lexile scores were through the roof, compared to the sample here which has poor to middling reading ability (typical for dyslexia) — i do not have dyslexia but still feels like OpenDyslexic makes it easier to read.

The data in the study does not support your conclusion that it “actually is worse”, unless i missed something in my interpretation? It seems to be a wash.

This smaller N=3 study supports the efficacy of OpenDyslexic, using _actual reading tasks_ instead of the more artificial task in the study you provided: https://dergipark.org.tr/en/pub/tayjournal/issue/76493/11524...

One of the testimonials implies that they've forced all websites to use this font, and it improved their experience significantly.

I suspect that would be true of Arial, even! This has been my goal for some time, but no browser has such an option. (You can change the default font, but almost no website leaves it set to default, so it does nothing.)

In Firefox, you can the advanced font settings and uncheck “Allow pages to choose their own fonts, instead of your selections above” to force websites to use the default font.
Does this break icon fonts, or will icon fonts still display since the font you choose wouldn't have those ligatures defined?
An icon font ligature is just an entry in the font file that says to display a particular sequence of characters as a single glyph. In the absence of the font the browser doesn't know that the page designer (or the font designer) intended a ligature. If you force an alternative font the browser will just display the sequence of characters that made up the ligature.

A few ligatures do have their own unicode code points, e.g. ffi, so the browser knows what is intended independent of the font. The ligatures used by icon fonts do not have their own code points.

> An icon font ligature is just an entry in the font file that says to display a particular sequence of characters as a single glyph.

Most icon fonts use private-use codepoints, not ligatures. These codepoints have no standard glyph, so disabling site-specific fonts will make those characters fail to render at all or display placeholder characters (e.g. �).

Thankfully, icon fonts are declining rapidly in usage - SVG icons are superior for most use cases.

> A few ligatures do have their own unicode code points, e.g. ffi, so the browser knows what is intended.

The precomposed ligature codepoints (e.g. U+FB00-FB06) only exist for compatibility with legacy encodings which included similar characters. They shouldn't be used in new documents. If you want text on a web page to use ligatures, use the CSS font-variant-ligatures property to control which ones are used. (And make sure to disable them on monospaced text!)

Its interesting that Firefox would offer a feature that disables all custom fonts rather than simply prioritizing the user's preferred font.

It does add a bit of performance gain not loading the fonts, and it may be a small security improvement, but if the goal was usability/readability that seems like a huge miss.

A preference based system would be almost impossible to maintain considering the literal infinite number of type faces out there (thousands plus more built every week).

Add to that, this feature also needs to be understandable for non-technical people who might never have view a page source in their lives.

Much as a granular system would be more powerful and preferable for the HN community, having a dumb toggle makes a lot more sense from Firefox’s perspective.

I wasn't thinking granular control. It could work exactly like the existing preference with users picking a preferred font. That font is always set as the highest priority font and used for any characters it supports, but the page's custom fonts are still fallbacks and kept in the font stack for any characters missing from the user font.

This would be helpful for icon fonts, but also for user preferred fonts that don't cover all language character sets.

Oh I see. Is that not how it works already?

Apologies if I’ve completely misunderstood your comments

No problem at all! I haven't been by my desktop the last couple days but I do want to test it. I hope that's already how it works, that would make more sense to me than blocking all custom font loading (though that does have its purposes).
No, using this setting in Firefox blocks all fonts site by a website, which means if a site uses an icon font the website UI will look very broken. I am not a fan of icon fonts.
> Thankfully, icon fonts are declining rapidly in usage

I wouldn’t be so sure about that. Resources like FontAwesome are still used heavily in VuePress and similar documentation generators.

I have one project using them, mainly because SVGs can be a bit of a pain to use in XSLT.

edit: it may not be SVGs themselves. The icons on that site are repeated and I would reach for an SVG sprite that can be reused, I ran into issues with that working cross browser in XSLT - Firefox was actually the problem if I'm not mistaken.

Yes, it does break icon fonts.
I know that safari lets (or at least used to let) you apply a custom CSS file on top of pages’ own styles.
Chrome has that in the preferences. Don’t even need a feature flag or user styles.
(comment deleted)
(comment deleted)
I would like to see HN use this.