The end of Type 1 fonts

https://typenetwork.com/articles/the-end-of-type-1#

by pavlov • 3 years ago

177 133 3 years ago

typenetwork.com

numtel2• 3 years ago

That's a misleading headline if I've ever seen one.

Type 1 without extra qualifiers refers to diabetes.

admash• 3 years ago

Oddly enough, I was not misled and assumed that it referred to Adobe Type I, given that is HN and end of Type I diabetes would be making headline news in most all other news outlets. Which it is not.

numtel2• 3 years ago

An article about therapies under development before they reach the market would absolutely fit on HN while being ignored by news outlets.

I guess that article would be titled, "The coming end of Type 1"

nyrikki• 3 years ago

It was type 1 errors or “false positives” that popped into my head especially as a post just a few lines down had “false positives” in the title

skissane• 3 years ago

I immediately thought it was about fonts, because I saw the website was “typenetwork.com”-never heard of them before, but their name rather obviously suggests typography over anything else.

Personally, I wouldn’t associate “type 1” without context with diabetes over its other possible meanings (type 1 fonts, type 1 errors, type 1 hypervisors, HIV type 1, herpes type 1, etc). Diabetes may well be the most famous example, but numbered types are very popular in medical science, and many other fields besides

pastage2• 3 years ago

That is an interesting fact, considering you seldom see type 1 refering to diabetes without the qualifier. I guess medical society is used to naming things and the problems with that. (Even though T1D is really too broad, often used to describe other types of diabetes)

russdill• 3 years ago

And see here I was hoping it was a discussion of post Kardashev type 1 civilizations.

ur-whale• 3 years ago

Type 1 without extra qualifiers refers to diabetes.

That may very well be accurate.

In your world, though.

jhoechtl2• 3 years ago

We are not writing here for AI bots but for humans. In this (CS) context, the first association clearly isn't about diabetes.

hnbad2• 3 years ago

I'm not sure what "this context" refers to. The site (typenetwork) is probably more "design" than "computer science". The HN user base includes a lot of people from computer science backgrounds but not exclusively so, nor does the majority of the articles featured on here (remember that "tech" does not mean "computer science"). HN itself is an outlet of a venture capital startup accelerator. I'd say the expectation for HN users would be to at least be "tech-adjacent" but computer science is neither a necessary nor sufficient part of that and the topics in submissions far exceed direct professional interests (or even "hacking" in any sense of the word, despite what the guidelines suggest).

An article about research that could mean the end of Type 1 diabetes would hardly be out of place here, though I agree that assuming that "Type 1" without qualifiers automatically refers to diabetes is not a safe bet here, if anywhere.

Dalrymple• 3 years ago

This headline is also initially alarming for those interested in the most performant type of hypervisor.

thom• 3 years ago

I was worried that Wizards of the Coast had announced the end of the Vintage format.

lynx23• 3 years ago

I think you might be biased. I didnt think about diabetes in the slightest when seeing the headline.

rejectfinite• 3 years ago

To you it does.

cs02rm0• 3 years ago

It's where my head went too and I'd be in favour of the headline being updated, but I don't think diabetes owns the association between a letter and number.

bawolff2• 3 years ago

Personally i thought it was clear. This is a tech website, and we're obviously not going to solve type 1 diabeties overnight out of nowhere. That's two context clues for what it is about.

[Which is not to say that the headline shouldn't be disambiguity]

gpvos3• 3 years ago

Also, Type 1 fonts have been on the decline for a long time, so announcing their end is a likely thing to do.

elevaet2• 3 years ago

I got nerd-baited, expecting this to be about Type I Civilizations.

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

holistio• 3 years ago

I was expecting the end of Type 1 diabetes.

elevaet2• 3 years ago

I like this one better

mtrycz2• 3 years ago

It's a common problem on HN to have too little context. Even in the tech space, there is so many subdomains, that it's easy to get confused. It also doesn't help that often (not in this case, tho) writers assume you already know what they are talking about.

A couple of words of context (or maybe a "category tag"?) would be so much helpful, most of the time.

msla• 3 years ago

That, and the re-use of abbreviations.

Like VM for example: Is it Virtual Memory, Virtual Machine, or Virtual Machine? Distinguishing Virtual Memory from the other two is easy, but is a Virtual Machine a bytecode interpreter like the JVM or a hypervisor like IBM's VM or Xen? You might well have to think hard about the VM settings you use for a VM that runs on a VM.

Sharlin2• 3 years ago

Just append the word "fonts" to the title.

pantulis• 3 years ago

I remember installing Adobe Type Manager from 3.5'' diskettes on Windows 3.1, with a special supplementary disk that included the Lucida fonts. "Goodbye Times New Roman!"

zmmmmm• 3 years ago

There I was thinking they cured diabetes :-(

romeros• 3 years ago

yeah me too. :-( dang should change the clickbaity title.

coldtea2• 3 years ago

No one involved in typography, typesetting, DTP, and graphic design (the target audience) would even conflate this title with diabetes...

The website is called "typenetwork" even.

Not everything is "clickbait"...

zimpenfish• 3 years ago

I mean, I'm pretty sure you'd have heard about that from about a million other outlets before seeing it here, no? It'd be Nobel Prize winning level stuff.

spirit5573• 3 years ago

Well, type 2 is definitely curable and reversible. type 1 diabetes..... yea no.

RHSeeger2• 3 years ago

Type 2 can be managed, mitigated, and treated; but it cannot (currently) be cured. The actions taken may prevent all the symptoms, but the underlying condition is still there.

Sunspark• 3 years ago

This is the same as saying "we are removing your ability to look at jpg, use heic instead".

T1 is a perfectly fine file format with many quality fonts. It does not burden a system to have it. Support is not lost on Linux systems which use the freetype renderer.

pilif• 3 years ago

It does not burden a system to have it.

any file format you support poses a significant attack surface, especially an old and creaky one whose parser you've written in the 90ies and ever touched.

hulitu2• 3 years ago

any file format you support poses a significant attack surface,

and yet the trend is to use more and more libraries, sometimes from dubious sources and sometimes (hello npm) with malware.

scrollaway• 3 years ago

Actually it’s the same as “removing support for XBM because nobody uses it anymore”.

Would you not make completely out of scale comparisons just to be contrarian on HN? Please?

hedora• 3 years ago

How many xbm files were used to store archival copies of documents?

Adobe Reader has always been terrible; I’ll happily continue to avoid Adobe software whenever possible.

However, I hope this doesn’t lead to other PDF software ending support for Type 1 fonts.

maxnoe• 3 years ago

This is about removing support from applications creating PDF files (InDesign, Acrobat, ...).

From the article:

Keep in mind that older PDFs created with Type 1 fonts are safe—as long as their font data was embedded in the PDF when it was made. PDF readers, whether from Adobe or elsewhere, will continue to render these documents as they always have.

hulitu2• 3 years ago

are safe—as long as

This does not sound like "safe" to me.

layer8• 3 years ago

This is the part I don’t understand. They retain the ability to render Type 1 glyphs, but they remove the ability to load those from external font files, and/or don’t ship any “built-in” Type 1 fonts anymore?

My question is just about rendering/viewing. I understand that support is completely removed for authoring.

madeofpalk• 3 years ago

It does not burden a system to have it.

Any and all software is a burden.

legulere• 3 years ago

There’s no code better than no code.

eternityforest2• 3 years ago

Some is a negligible burden, or a net unburdening because it replaces something else that was a bigger problem

shp0ngle• 3 years ago

Any old code is burden by definition.

You can still use old versions of Adobe software. (I… guess? I am not sure how do the Adobe Creative Cloud licensing shenanigans work nowadays, I haven’t used Adobe suite for ages)

zorked2• 3 years ago

Moderator: title request change to "The end of Type 1 fonts" to disambiguate from Type 1 errors, Type 1 diabetes, etc.

comradesmith• 3 years ago

I thought this would be about the transition to a Type 2 Civilisation…

dmak• 3 years ago

Me too! I was so disappointed when I saw the article.

lucakiebel• 3 years ago

We would have to be a Type I civilization first, we’re only at about 0.73

sealeck• 3 years ago

(cc @dang)

Doctor_Fegg• 3 years ago

Oh, I thought it was Type 1 Fun:

https://www.rei.com/blog/climb/fun-scale

ho_schi• 3 years ago

I also expected sth. about diabetes.

jcutrell• 3 years ago

Was going to mention this, and also the Bezos classification of Type 1 and Type 2 decisions - definitely would benefit from a better title.

ant6n• 3 years ago

I thought it was type 1 personalities.

charlieyu1• 3 years ago

Yea I thought it was about Type 1 decks in Magic the Gathering

gpvos3• 3 years ago

Good to see PDF viewers will at least keep supporting it.

wolrah2• 3 years ago

Good to see PDF viewers will at least keep supporting it.

No, the reason PDFs created with these fonts will remain viewable is the same reason printed documents will remain readable. The glyphs used are part of the document. You don't need to have the font installed or even have the ability to render fonts to display a PDF, that's the whole point (*Portable* Document Format).

You may have noticed that some PDFs, especially older ones, either don't allow text to be selected or have weird behavior when selecting where the selection might go in a line across columns instead of following the text. This is because they don't include the metadata required to explain the document as text, it's just a bunch of glyphs.

gpvos3• 3 years ago

Well, they are embedded in the PDF as a Type 1 font. That support can in principle be removed (e.g., if your font rendering depends on a library shared with other software that drops Type 1 support), although indeed that would not be in the spirit of PDF (and go against what is now an ISO standard).

Reasons for text selection going across columns can be very varied, although indeed documents with metadata will also otherwise usually be structured enough to have proper text order.

ncphil• 3 years ago

"No longer support writing on papyrus, in favor of more modern vellum based media."

And then, the Dark Ages.

Of course, backwards compatibilty will be maintained: at least to the end of the century?

Digital culture is already so ethereal, fragile, that you have to wonder if millennia from now our descendents may posit this period as another "dark age" because none of what we've generated since the 90s will remain (or be decypherable).

klodolph• 3 years ago

I am somewhat more hopeful. There are conversion tools. Fidelity is reportedly not perfect—I haven’t seen problems, but the metrics may not match exactly. Tools like FreeType, FontForge, and GhostScript will remain in their current form, even if development / maintenance is abandoned.

Archivists have all sorts of guidelines for how to preserve digital media. We won’t save everything, but we ARE churning out plenty of people with master’s degrees in library sciences, who often go to work (in some capacity) preserving the world’s digital data.

kps• 3 years ago

The article is vague, but OpenType CFF fonts remain supported. These are Type 1 outlines in an OpenType container. Type 1 fonts should be 1:1 convertible to OpenType CFF, techincally. The hitch for commerical users is that fonts often have ridiculously restrictive licensing forbidding this rewrapping.

nerdponx3• 3 years ago

It's a lesson in what happens when you give private for-profit entities responsibility for public goods.

crazygringo5• 3 years ago

Fonts aren't public goods. Same as computer programs aren't public goods either.

Swizec• 3 years ago

Digital culture is already so ethereal, fragile, that you have to wonder if millennia from now our descendents may posit this period as another "dark age" because none of what we've generated since the 90s will remain (or be decypherable).

Mark Lawrence kinda explores this idea in his Prince of Thorns universe. Former AI researcher, now fantasy author.

Without spoiling the books too much: There’s a lot of stuff left over from The Ancients who had these weird rectangular things that sometimes pop to life and do magic, but are otherwise inert, useless, and have no moving parts you can analyze. Very strange black rocks that seem to do things when you hold them just right but nobody knows what makes the assembly work, or how, or when it might suddenly pop into life. You kinda just fiddle with it until something happens and hope for the best.

Oh and The Ancients had a lot of “liquid stone”, which is very strong and would come in handy for building forts, if anyone knew how it works or how to make more.

Filligree• 3 years ago

Concrete is an old, old recipe though. It's hardly some complex modern novelty. I wonder how it got lost.

Does the series involve recovering this lost knowledge? Or is it just the backstory for a fantasy novel?

SideQuark• 3 years ago

The Dark Ages had nothing to do with changing writing media. Also modern scholars don't use the term as they're learned there was a lot of stuff discovered and done during the "Dark Ages".

tablespoon2• 3 years ago

> "No longer support writing on papyrus, in favor of more modern vellum based media."
And then, the Dark Ages.

Nitpick, but IIRC vellum is far more durable and long-lived than papyrus.

Of course, backwards compatibilty will be maintained: at least to the end of the century?
Digital culture is already so ethereal, fragile, that you have to wonder if millennia from now our descendents may posit this period as another "dark age" because none of what we've generated since the 90s will remain (or be decypherable).

One of the ways software engineers (and software engineering organizations) tend to be terrible is they prioritize their myopic desires over others' long term needs. Part of that is probably driven by software engineering's tendency to create systems that require constant active upkeep (due to its deficiencies), which leads software engineers to think it's right for everyone should be forced to perform constant active upkeep (or they "get what they deserve").

IMHO, the right thing to do is to carrying around some legacy baggage to prevent old documents from becoming unusable.

enriquto• 3 years ago

Maybe that's the point? Breaking backwards compatibility, even if for a noble intention towards a better future, results in practice in annihilation of the past. No good deed goes unpunished.

lucumo• 3 years ago

software engineering's tendency to create systems that require constant active upkeep

Nearly everything requires active upkeep, especially if it interacts with a changing environment. Software is no exception there.

xp843• 3 years ago

In a way, I think I'm comfortable with no one 200 years from now being able to read every tweet, because nobody bothered to keep an archive of all of Twitter and move it from one form of media to another.

I'm sure there was a great deal of stuff of little value being spoken but not chiseled in stone in 100 BC as well. Only important stuff survives.

If we have anything of great value to say, then let us put it intentionally in an archive for posterity. There is a limit, after all, in how much time future historians will have to read our stuff. Nobody can read all of Twitter today, nor in 200 or 2000 years.

eternityforest2• 3 years ago

I still think that at least part of the issue is that many devs hate tech. They don't want to make tech more reliable or durable, they want to do their job, get paid, go home and work on toy side projects, and continue using paper and low tech stuff, even to the point of wood stoves and hand drills for anything they care about.

Tech is a tool for exploring ideas to them more than a method of developing replacements for old fashioned goods, and I suspect they see the concept of trying to make it as durable as vellum as just shoehorning in the wrong solution when there's already vellum.

OliverJones• 3 years ago

My brother the academic librarian / archivist has, for many years, been warning about this kind of obsolescence. It's necessary for we software folk to ask ourselves "two centuries from now, how will people make sense of our digital files?"

One of the challenges is, of course, that it isn't obvious to any generation of creators what will be of lasting value.

Academic archivists and MLS degree holders can't do it all, any more than they could control the 19th-century paper factories that churned out all that junk paper that turned brittle in a few decades.

thfuran2• 3 years ago

Two centuries? I don't have an optical drive or zip drive or any flavor of floppy drive anymore. Lots of files from the last two decades are inaccessible to me even before considering whether they're still parseable / executable.

int_19h• 3 years ago

I ask this myself all the time.

One thing that I think we need to adopt as standard practice is self-describing storage formats. That is, for any file (or, more broadly, a binary blob - say, hard drive image), I should be able to look for a well-known marker in the bitstream, and, having found it, be able to easily extract a detailed human-readable spec for the format with no special tools. Better yet if it comes with a simple reference implementation optimized for simplicity and ease of understanding, even if it's very slow.

This doesn't solve the problem for physical representations of data, but so long as we can keep re-archiving the bits...

upofadown• 3 years ago

Adobe doesn't support bitmap fonts either, but that doesn't stop me from using them in my terminals. This is only limited to a single producer of software. It will be interesting to see if any other generators of PDFs will pay any attention...

detourdog• 3 years ago

Yes, at this point type 1 fonts are bigger than Adobe. Type 1 fonts just exist. Adobe needs subscriptions to exist.

simooooo• 3 years ago

But it’s THE producer of PDF software.

Several of my clients are already moving their entire catalogs to new fonts

DaiPlusPlus• 3 years ago

Correction: it’s the producer of PDF software with up-sells baked into it, even if you paid for the full-fat Adobe Acrobat with a perpetual license (which is still a thing, fortunately) you still get steered towards using their Subscription SaaS features like Adobe Sign, Document Cloyd, etc.

Fortunately Adobe’s PDF specification is an open, public standard (a surprisingly altruistic move for a company seemingly ran with the same level of malevolence as if Lucifer got himself an MBA) so now, in 2023, there’s no shortage of high-fidelity PDF viewers and virtual printer drivers. Browser JS-based PDF viewers are the best thing to happen in this area in decades - remember all those Acrobat Reader splash screens in the early-2000s?

manv12• 3 years ago

Some of the fonts (Beowolf) will never be duplicated because the font re-renders itself every time, which lead to some neat effects when ripped for print

There's an OpenType version, but since TT fonts aren't programs the way Postscript fonts are it's not really the same.

https://www.fontshop.com/families/ff-beowolf

I believe it's still the only RandomFont ever made.

nereye• 3 years ago

For other examples, see http://luc.devroye.org/randomizedfonts.html.

Amongst other things, assuming what that page says is true, only Type 3 fonts support that (randomizing glyphs at render time), not Type 1.

barefootliam• 3 years ago

Beowulf was a Type 3 PostScript font, not Type 1, so that it could bypass the font cache and could access random numbers on the LaserWriter PostScript interpreter.

Type 3 fonts don't have hinting in the renderer, and require a full PostScript interpreter, so they have never been widely supported.

manv12• 3 years ago

I saw this and went back to the black book to read up, and yeah, type 1s are limited to a certain subset of PS because the font is pre-tokenized.

Interesting I never noticed that before now.

FYI Type 3s were totally widely supported, and supported by everything that supported Type 1s. Type 3s may not have worked with ATM, but I had a huge collection of fonts and never remember seeing a Type 3 font that didn't render correctly. I did have commercial fonts from real foundries, so maybe the ones made by people didn't render.

WirelessGigabit• 3 years ago

How does it work when they are embedded in a PDF?

Does the renderer have explicit support for these fonts or are they embedded in a way that transcends the format itself?

compiler-guy• 3 years ago

They are embedded as postscript subroutines so all is well that way.

schrijver• 3 years ago

You should be able to convert them to OTF files with Fontforge. OTF supports both Postscript outlines (cubic bézier) and TrueType outlines (quadratic bézier). That should be lossless for the curves, however does someone know if it’s the case for the kerning data as well?

elmimmo• 3 years ago

You are but, as of today, reliably, only via the GUI, and therefore, one by one, not in batches via scripting due to a bug[1]. That is the case at least for Type 1 Mac fonts.

[1] https://github.com/fontforge/fontforge/issues/4668

The end of Type 1 fonts

Related Stories