B612 Font Family | Hacker News Reader

B612 Font Family

b612-font.com

466 pointsOliverGilan94 comments

Strange, the background image of the site's header shows a very nice cockpit font indeed, but the actual B612 font is quite different - for instance zeros have no line through them. Though the oxygen indicator seems to use this font.

The image shows a 0 without a line as well. Maybe the other one is a Ø? Or maybe there are conventions to use slashed zeroes in some places an non-slashed in other places?

Yes. 0 (not slashed) is ambiguous only in fields that accept also letters.

I think the goal of the designers wasn't to recreate the font in the plane, but to create their own inspired high legibility font.

(for those downvoting I think this, because this is what I remember to have read years ago when this font came out)

It does include the slashed zero, just not at the zero glyph place.

Yeah, that always bugged me! The zero-with-a-slash only seems to be used in that flight plan section of the screen; all of the other zeroes are without the slash. And the "5" and "4" next to the zero-with-a-slash in the waypoint 5240N is clearly the B612 glyphs. So it makes me think that in the Airbus software, in contexts where there might be ambiguity between 0 and O or where there are mixed letters and numbers, they use the zero-with-a-slash, but everywhere else (numeric-only readouts, etc.) they use the non-slash zero.

And I wonder if there's a 0-with-a-slash in the downloadable version of the font, enabled with an OpenType stylistic set number or feature or something, or if that zero-with-a-slash is just something custom the Airbus folks do internally.

Checked the Github repository, there are indeed separate "zero" and "slashedzero" glyphs.

While there is a glyph like that there's no nice way to access it. There is no font feature to enable slashedzero by default and slashedzero is not mapped to U+0030 U+FE00 (Unicode's standard variant for an explicit slashed zero). Instead it's only accessible using the private use codepoint U+E007 and then typically doesn't get copied as a regular zero.

But more generally the font has an empty GSUB table which would be used for such substitutions. I'm wondering if the cockpit display maybe misses support for that and that's why they tried not to use it for anything.

Still a worthy bug to be submitted.

* https://github.com/polarsys/b612/issues/20

The workflow to fix things doesn't seem clear - are people supposed to work on the .glif files?

The hero we need but don't deserve.

does it have to do with "knowing" whether the data being displayed is numeric vs alphanumeric?

When alphanumeric, put the slash, when only numeric leave it out?

Can't wait for a font renderer to get this smart. Any day now...

> And I wonder if there's a 0-with-a-slash in the downloadable version of the font, enabled with an OpenType stylistic set number or feature or something, or if that zero-with-a-slash is just something custom the Airbus folks do internally.

According to the Github issue mentioning in a sibling comment:

> We designed the two zeros having in mind the slash one to be used in alphanumerical sequences, and the normal one to be used in numerical sequences. > > You can find it in the ‘private use area’

I also miss a distinguished 0 (with a line through). With B612 the O and 0 (with a line through) could be mixed up.

It’s very unfortunate that these glyphs are not available via OT features or font-variant in CSS. I assume the cockpit software uses glyphs directly instead of using OpenType.

There are two open tickets about it, the second is almost four years old, maybe someone here who is familiar with the tooling could contribute?

https://github.com/polarsys/b612/issues/24

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

Even the font's designers were unable to contribute due to a lack of merge rights, since they weren't elected by the Eclipse foundation. https://github.com/polarsys/b612/issues/10#issuecomment-4719...

The only user who's demonstrated merge rights on the repo is the repo's Eclipse Foundation project lead, Laurent Spaggiari: https://projects.eclipse.org/content/laurent-spaggiari-proje...

Who's the only listed contributor on the repo: https://github.com/polarsys/b612/blob/master/CONTRIBUTORS.tx...

And who's been publicly inactive on GitHub since the repo's last merge in March 2019: https://github.com/LaurentSpaggiari?tab=overview&from=2019-0...

I don't think any amount of tooling knowledge could facilitate a contribution at this point. It's a dead project. The best anyone could do is apply the steps in issue #24 to a fork and start promoting it.

Good to see the Eclipse Foundation killed something, again, due to shear incompetence.

It's really unfortunate that this glaring issue stops wide adoption. The slash through the zero should have been the default, very silly for the official release to omit that.

Another font that has a similar aim is https://en.wikipedia.org/wiki/Atkinson_Hyperlegible I find that it strikes a balance between aesthetics and legibility. Two differences I noticed immediately are the 0 and I characters which are much more difficult to distinguish with this font.

Thanks you for sharing, it's very interesting.

Huge endorsement for Atkinson Hyperlegible from me. As someone with horrible and declining eyesight in multiple ways, it has improved my computing experience a lot.

There’s been a couple of attempts at making a monospace version, I currently use the one by Edward Shin in my editors and terminal https://github.com/Hylian/atkinson-monolegible

thanks for sharing this monospace version, i've been using atkinson as my browser's sans-serif font

That font does look really nice!

Also noticed the slash in the 0 runs top left to bottom right, rather than the far-more-common [in my experience] top right to bottom left

Any idea why that decision was made?

An obvious explanation would be that this helps distinguish it from øØ (as in Danish and Norwegian), which is slashed from bottom left to top right. The Wikipedia article does demonstrate its legibility when blurred.

> which is slashed from bottom left to top right

interesting way to describe this in that it is described in reverse order of how i've always seen it. my first thought is that's because it is also opposite of how i would draw it when writing by hand.

yeah ... I thought about "top right to bottom left"

because then you are comparing top to top

instead, i kept it left to left :)

I honestly don't know the stroke order Scandinavian kids are taught when learning an alphabet that contains the letter. I also don't know if it's consistent across languages.

Most strokes that start from a lifted position move from left to right (e.g. the crossbar of the lowercase t) and European languages are written left to right, so describing it as left-to-right seems more intuitive to me.

Do you know of other fonts with the same goal as Atkinson Hyperlegible? It's gorgeous.

Thank you. I have been trying to find this font again but I didn't remember it's name. I had it maybe a year installed in my previous computer and I loved it. Really good for coding.

This looks much better than B612. B612 just looks... awkward. It has these weird little cut-out areas at the intersections of some of the lines. I don't get it. Makes it looks like it's awkwardly pasted together from disparate pieces.

Those cut-outs look a bit like ink traps, though whether that actually makes any sense in a font unlikely to be used with actual ink, I do not know...

"Atkinson Hyperlegible is a freely available typeface built around a grotesque sans-serif core,..."

So obviously my understanding of the definition of grotesque is not what's meant as its use in relation to fonts, but the definition in a font's use is just odd on why it is used in this manner.

https://en.wikipedia.org/wiki/Grotesque_(Stephenson_Blake_ty...

If you click through to the next page:

    According to Monotype, the term "grotesque" originates from Italian: grottesco, meaning "belonging to the cave" due to their simple geometric appearance.[14] The term arose because of adverse comparisons that were drawn with the more ornate Modern Serif and Roman typefaces that were the norm at the time.

It looks like two of its design goals were to retain visual consistency at veyr low dot-pich screens, to maximize readability of long passages in uppercase, and to accomplish this all of a full complement of symbols as well.

(https://intactile.com/projets/ameliorer-la-lisibilite-et-la-...)

The kerning looks distractingly bad. It’s easy to see in the word “complete”.

Short words like “in” and “all” end up looking like they’ve got completely different tracking, which in turn makes it harder to see word boundaries.

Probably a high-DPI screen in a well-lit room is simply the wrong environment to judge this font though, since it’s designed for low-DPI screens in a cockpit. But if you’re planning to use it outside of that environment, do test against other options.

Ahhh fonts, where everyone gets an opinion and they're all super important.

If you're designing a font not just for legibility, but primarily for safety, then it seems extremely important that each glyph is uniquely distinguishable from other glyphs. Although this font has different characters for 1/I/l, at a quick glance an uppercase i could still be confused with a pipe (|), and 0 (zero) and O (capital o).[1] I'm sure there are more. So from that standpoint, this font fails for me for legibility/safety.

Also a nitpick, but assuming Chrome is using 60pt B612 font for the title (../fonts/B612-Regular.woff), the "B/6/1" glyphs are hideously formed (that "1" puke) and make me doubt the rest of the character set.

[1] e.g., the FAA has already addressed this with tail numbers: https://www.faa.gov/licenses_certificates/aircraft_certifica...

The font has a slashed 0 for use in alphanumerics and an open 0 for use in numerics. There's discussion of it in other comments here.

I imagine pipes aren't used beside uppercase I in cockpits.

The odd serif on the 1 is to ensure it degrades correctly at low resolution.

When you design for safety, you also have to ask "safety in what context". It's neat that they released the font with an open license, but they didn't design it for anything other than Airbus cockpits.

I couldn't find any evidence for this font having particularly good legibility (such as a Rapid Serial Visual Presentation (RSVP) test), but my hunch is that it's not that great.

It was probably optimized for low-res screens and specifically for monospace usage.

Kerning looks quite off in the Google fonts specimen (https://fonts.google.com/specimen/B612?query=b612)

Posted every so often. A more descriptive post title including something about aircraft cockpits or something would be useful.

Bunch of previous discussion here from 2019:

https://news.ycombinator.com/item?id=18946601

It's very interesting that I picked this font for my live coding language: https://quaverseries.web.app/

The main reason is that I need to use a syntax to represent a music sequencer such as ___65. B612 can separate two or more underscores perfectly.

For a long time, it's always my first choice of coding font although I have no idea on its origin. I just feel it looks great when I browse all Google fonts. But it's great to know the story behind it, and thanks for sharing.

it does look great on ur site!

Thanks! I realised later that it has an issue with parenthesis (). So I had to switch to Fira Code for https://glicol.org/

A question on tracking/letter-spacing.

In my browser, I always use my own font choices (Firefox: Settings → Fonts → Advanced… → untick Allow pages to choose their own fonts, instead of your selections above), which makes me more sensitive to certain sorts of modifications: most significantly, I observe how unreasonably common it is for pages to change the whole document’s letter-spacing, normally to a positive value. (My firm opinion: you should absolutely never do this as a global thing, no exceptions.) I’m not certain if it might make more sense with some fonts than others, but the main thing is that I notice it.

So then, with this font, the main thing that immediately stood out to me is how wide its tracking is, so that I’d almost feel justified in adding `letter-spacing: -0.03em` (except that the shapes and some kerning pairs don’t work that way). Does it feel that way to others? And is there some kind of general trend in the direction of adding more space between letters?

(The font’s bold face, on the other hand, feels a mite cramped when in proximity to the regular weight, since it uses the same metrics. Also certain pairs are kerned much more tightly so that they feel out of place to me, e.g. Vi, DG (especially in bold). And as for the parentheses, I had to check that no, they hadn’t inserted a thin space inside the parentheses, the font really is just that weird.)

Typesetting for labels as is the case with this font is probably more suitable for wide tracking for legibility than fonts for text.

Anyone else just learning about Université de Toulouse III? Did the first one sink into the swamp, followed by the second one burning down, falling over, then sinking into the swamp?

There are in fact 3 universities in Toulouse: Toulouse I-Capitole, Toulouse Jean Jaurès (used to be Toulouse II-Le Mirail), and Toulouse III-Paul Sabatier. And since it would be absurd to have just those three universities they are part of a group called "Université de Toulouse" (formerly Université fédérale de Toulouse Midi-Pyrénées) which also counts a number of schools https://en.wikipedia.org/wiki/Federal_University_of_Toulouse....

Ha! No, that's just the local naming since Toulouse has several of them, they have proper names in addition to the numbering. Toulouse I Capitole is (mostly) law and economics, Toulouse II Jean Jaurès is (mostly) social sciences/humanities/linguistics and Toulouse III Paul Sabatier is (mostly) STEM.

My alma mater is technically Université de Toulouse, the federating entity that formally gives the diploma, but the existence of such entities depends on the city, each university has its own head and administration. My grad school lab was attached to Toulouse III.

As a French I like your description better.

I really like the font aesthetically, and I'd use it as a coding font. However, the brackets and parentheses look too similar at smaller sizes, and arrows (ie. -> and =>) don't quite line up properly.

i used it as a coding font for awhile then realized that i cant distinguish between comma and fullstop on my ide setup

I always put a space after a comma. Never after a fullstop. Coding rules improve readability.

Every time that I pair with someone, they always get confused about it, after months you get adjusted, might be caused by the syntax when coding.

Welp, it's a failure because the capital "i" does not have crossbars... making it indistinguishable from a lower-case L.

Such a tiresome defect in one font after another.

I wouldn't say it's indistinguishable - the lower case L has a curved bottom.

without having a comparison, the uppercase I (as in INDIA) can be confused with a 1 (as in one) or a lowercase l (as in LIMA)

But you don't know that unless they're both being displayed at the same time.

It's not a failure if it meets its stated goals. It's not trying to be a programmer's font, a terminal font, etc. It's specifically only trying to be a font for cockpit displays. On that front, there isn't really a context where I and l would be ambiguous in context. The only place I can think of would be waypoint names, but those are always all caps, so I and L don't present any problem.

On the other hand, putting crossbars on their I would make it much more visually similar to their 1 glyph. Mixing I and 1 (again, in waypoint names, for example), is much more likely and the strong visual difference in I and 1 by not having crossbars on the I seems like a better choice for the stated goals/applications of this font. ("...improve the display of information on the cockpit screens, in particular in terms of legibility and comfort of reading, and to optimize the overall homogeneity of the cockpit")

> It's specifically only trying to be a font for cockpit displays. On that front, there isn't really a context where I and l would be ambiguous in context.

So for example IATA and ICAO airport codes are not within the context of an airplane cockpit?

And the visual difference between "laa" and "Iaa" doesn't matter?

The first one is Lamar Municipal Airport, while the second one is Igarka Airport.

In those cases, they always would be in caps.

So you get LAA and IAA, which are indeed distinct in the typeface.

Nevertheless they (and most other typefaces) should have avoided the ambiguity.

you are having too much of good faith in always-perfect assumptions when it comes to life-critical systems. I hope you are not involved in developing or designing such systems.

I tried out the monospace font for terminal use, but some of the punctuation characters have terrible spacing, including : and ., so things like

    abc::xyz

and

    some_expression
        .method()
        .method();

look awful.

yes i ran into problem with distinguishing between ; and : as well on this monospace font. i'll say it was fun to try it and i used it for awhile but it isnt suitable for coding purposes. im much more happy now with a highly customizable font like Input[0] where i can make it as wide or as narrow as i like it to be and also customize the various characters

[0]: https://input.djr.com/

Maybe that's on purpose to discourage airplane pilots from coding their own JS scripts inside the plane.

but it’s Rust tho

I'm wondering if the image is real. Because there are slashed zeros and non-slashed on the same picture for some reason. Font seems to have non-slashed version.

I would assume in cockpit it's better to always use slashed version to avoid confusion.

Discussed in other comments... but the zero-with-slash is only used in the flight plan section of that display; all of the other zeros are without the slash. It seems they only use the zero-with-slash where there might be a risk of ambiguity between 0 and O... everywhere else where it's just numeric readouts, there's no slash because there's no risk of confusion, and I'm guessing they decided zero-with-slash is more visually ugly/less intuitive for quickly reading numbers than zero-without-slash.

> and I'm guessing they decided zero-with-slash is more visually ugly/less intuitive for quickly reading numbers than zero-without-slash.

Yes, it can be confused with 8.

Or when I was in first grade and my teacher marked me down thinking I had written both a one and a zero. Never will forget that.

0 and O look almost exactly the same in the font, is the distinction of these not important on flight screens?

Another research project needed to uncover such basics

wouldn't you think the better UI choice is to simply avoid situations where you can confuse O and 0 (and l and 1 and I) ?

No, I don't think that's a better choice at all, making two characters look the same is the worse choice, having to design something to avoid situations of two identically looking characters being together sounds much clunkier, than having the characters not look the same in the first place. Plus, airport codes, radio frequencies, airspace zones, GPS coordinates, etc... have numbers and letters.

This font is so spectacular, for some reason unknown yet to me, numerous fonts give me headaches/“tired eyes” when I used it the first time was a blessing I could perceive the difference in a week. A shame that, depending on the application, they look hideous.

Seems really bad to me.

There's a lot of distracting glitches like the K and Q having small gaps where strokes come together, and even glyphs like the N, M, Y and Z have weird little gaps at the joints, the F and B have a bizarre overhang at the upper horizontal stroke, the 1 and 6 have a weird cut in their strokes.

There are lots of weird confusing choices like the parenthesis being a round rect instead of a bow shape, the @ having a "complex a" inside, the 3 and 5 using a "cute" diagonal layout instead of the normal "rounded-and-rect layout".

Seems horrible for something that should be as simple as possible. The default system UI fonts like DejaVu Sans, Segoe UI, Roboto, etc. are far better.

It seems to me that those choices are made with low DPI in mind. Those cuts probably make sure that letters don't become blurry.

I think I'll try this one on my 1080p monitor for a day and see how it goes.

Have to agree. There is a lot of wild stuff going on that is visually tiring and hampers readability

I think these “ink traps” might be effective on low resolution displays. But yes, those parentheses indicate it’s not a coding font.

Yeah, there's a visual example here of how it helps the 'e' stay legible at low res:

https://intactile.com/projets/ameliorer-la-lisibilite-et-la-...

Without that "pixel trap" (?) the intersection (there's probably a fancy typographic term for this) would look too bulky.

For those who wish for a font family to have distinguishable capital "eye" from lowercase "ell", the following is a recommended test: how does the font render "Illegal1=O0"

I was looking into the source and was very confused until I found the UFO glif files and the build.sh script. It looks like the proprietary FontLab format is the "source" and the UFO files are generated from that and TTF files generated from the VFDs are committed into the repo.

BTW, I'm used to FontForge (which is not the most amenable to collaborative work) and would love to know what other font designers use and what would be a good collaborative workflow for open source fonts.

Unfortunately the name has been collided into by several other, more popular and newer, things since its creation. However, 46610 is still available as a font name.

B612 is from The Little Prince.

"46610 Bésixdouze ... is a bright background asteroid .. named after "B-612", home of The Little Prince. ... The title character lived on an asteroid named B-612, which is the number 46610 written in hexadecimal notation." - https://en.wikipedia.org/wiki/46610_B%C3%A9sixdouze

Pretty confusing, the banner image shows zeros with the dash to distinguish them from O's, in the example of the page there aren't any dashes.

previously https://news.ycombinator.com/item?id=18946601

The best legible monotype font for computer screens I know of is M+ 1m regular.

Example: https://www.dafontfree.net/data/41/m/94730/map/0-charmap-m-1...

I’m surprised the zero isn’t dotted or slashed for airplane cockpit applications

I wonder when we'll consider fonts as a 'solved' problem. I keep seeing designers making more of them, I guess it's just a tradition and hobby at this point

not so sure about that, my boss calls me a pedant when i write reports in cuneiform

s/fonts/languages/

I don't know how to say this nicely but your mistake seems to be that you think font design exists to solve one single problem that requires or even can have one specific solution.

You might as well ask when we'll consider fashion as a 'solved' problem. Or art.

Reply on news.ycombinator.com