The end of Type 1 fonts | Hacker News Reader

The end of Type 1 fonts

typenetwork.com

177 pointspavlov133 comments

"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).

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.

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.

I think we can safely view the 1980s/1990s as the testing grounds for finding out what works and what doesn't.

Naturally, the things that worked were acquired and consolidated by larger and larger companies, which allowed tech to be cheap and commonplace.

The #1 necessary factor for something being accessible later on is the sheer amount of popularity it enjoyed in the past.

E.g. I think some common interfaces, like SATA / USB Type-A will never die, willing to take a $10k long bet that they will still be easily accessible in 15-20 years. Maybe not as popular as some newer ones like M.2 or USB-C.

> Lots of files from the last two decades are inaccessible to me

Yes, but not by volume. The amount of files ever stored on all 3.5 floppies in the world could very easily fit into a single server rack these days.

Are these 0.01% of files more valuable in proportion than others? If not, then it makes 0 sense to focus on preserving those interfaces.

Is the data produced in the past month more valuable than, say, all the writings of the 1st century just because there's more of it? Would the data produced in the first week of 2099 be worth more than all of history prior to 1900 just because there's more?

I think the point is that small data can be continuously carried over to new mediums if it's valuable enough.

How many pre 1st century writings will fit on a 20TB hard drive? I'd venture to say - all of them.

> Would the data produced in the first week of 2099 be worth more than all of history prior to 1900 just because there's more?

No, and I think that's why we're mostly in agreement. Valuable and small data will always have accessible interfaces, because it is propagated throughout many mediums.

If there's an apocalypse scenario looming, you bet your butt that someone will be printing the most valuable information to good old paper, in case all machine-readable interfaces vanish overnight (unlikely scenario that I'm not going to lose sleep over).

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...

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.

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.

If there's anything I've learned in my career, it's that almost no one even acknowledges the existence of font licensing, let alone abides by it. And in general it's usually the kind of thing that the "font police" has limited ability to discover from outside the business.

In practice, people will do what they need to do to use their software, and tbh, there may be legal exceptions for interoperability, such as the DMCA has. You're allowed to circumvent DRM, under DMCA, if you're doing so solely to enable interoperability with different software/hardware (rather than, in order to distribute it). Not a lawyer, don't @ me.

almost no one even acknowledges the existence of font licensing

Ha, that seems to imply that people even know there’s such a thing as font licensing.

I’ve been a graphic designer for a long time now, and I barely even remember font licensing exists. Even though I know better, my gut instinct is “well we paid for it, we can use it however we want internally, send it along with the InDesign file to the printer, why not?”

The super fun thing about fonts that isn't quite intuitive to most, is that once you rasterize them you're freed from all shackles, meaning you can use a $12 font license and make an iconic piece of art, print it out and sell 50,000,000 copies of it and make a billion dollars, and you're completely in the clear. Or even produce a JPEG and sell that the same way.

But if you make it into a webfont and render that same font in a browser as text, lawsuit city.

Depends on country. Font makers scanning magazines to check which ads use unlicensed font is a thing that exists. Also for TV. And often a TV license is different from a print license.

>> "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.

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.

> 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.

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

In some cases, but software engineers tend to decide to make that environment change more, so even if your environment isn't otherwise changing, you still have deal with a bunch of change (e.g. you don't get to have software unless you can pay someone to regularly track framework updates, because some software engineers would rather not do the work to maintain backwards compatibility or support old versions).

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.

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.

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.

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?

Concrete requires a lot of energy to produce and the ongoing practice of making it would stop if there was ever a collapse of civilization.

Some of it, but its just there. It comes up sometimes but isn't omnipresent.

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

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

I studied graphic design and designed (by hand, and later "digitized" some using Fontmaker and Freehand) about a dozen of fonts.

So, about that "public goods" part... Are works of art a public good? I think, well... there isn't a very good definition of what constitutes public good. Usually, this is used in the context of economics, describing a good or a service managed by the government and provided to most of the population governed by the said government.

Some works of art do provide a service, in a way (well, entertainment is a kind of service, right?) and some are sponsored and otherwise managed by the government. Not every work of art earns the same attitude. The governments are very selective about what works of art they put on display in museums that are free for the general public to attend. Museums are public good though, right?

So, back to the fonts. The ones I developed, well... they are barely more than just an art academy student projects. I showed them to couple of my friends and my dad... and that was it. I doubt the government of whatever country I end up in will want to invest much efforts into preserving those. However... the teacher of my teacher (when it comes to fonts), Bazhanov created this font: https://meganorm.ru/Index/42/42375.htm . The "ГОСТ" in that document means that this font received a government-managed identification number. The font is still in use today, and the government released the IP of that font to the general public to use. I don't know if the government has any kind of a digital storage for the modern fonts and how would they be admitted there, but, back in the days when the linked font was accepted into publishing industry, the original films with the letter outlines were placed in an archive managed by the state's standards committee (that's what "ГОСТ" stands for).

Bottom line: it's complicated, and fonts aren't automatically public goods, but some of them are promoted into being public goods, based on many criteria, so, it's not easy to tell which are public goods and which aren't.

The historical record is a public good.

The historical record isn't saved as Adobe Illustrator and InDesign files.

Final products are saved in things like PDF's which continue to work just fine.

At some point, does not the software itself become part of the historical record? Otherwise we wouldn't care about things like the Internet Archive hosting its DOS emulator.

Not with that attitude

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".

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.

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.

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.

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.

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...

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

But it’s THE producer of PDF software.

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

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?

Correction: It's the producer of enterprise-adjacent PDF software, with some offering for smaller users.

Adobe is focused on capturing the media producing pipeline, and there is where it puts most of it efforts.

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!"

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.

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.

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.

> are safe—as long as

This does not sound like "safe" to me.

PDF's have virtually always embedded fonts except for a small collection of core fonts like Times New Roman that are part of the PDF standard. You've got to really go out of your way not to embed, and I'm not sure which software even gives you that option at all.

Otherwise it's kind of defeating the whole purpose of a PDF which is that everybody sees the same thing. Font embedding has been with PDF from the start.

It's safe.

Not sure if this is the kind of not-embedding-fonts you're referring to, but in layout and design software you often see PDF saving options to either rasterize fonts or embed the fonts as vector curves.

I believe both Gimp and Inkscape have been able to do this for a while, for example.

This is why PDFs intended for long term use or wide distribution should always (and often do) embed nom-standard fonts (eg as in PDF/A).

My usecase is to create pdfs through the "print to pdf" menu. I kinda assumed the fonts were embeded.

I don't seem to have any options when creating pdfs to embed or not embed fonts. Is this some feature of PDF creation software?

Yes. From the top of my head the usual print to PDF behaviour is to embed the relevant subsets of any font other than the fourteen core ones given in the PDF standard.

The default is to mostly embed, except for a number of standard fonts.

You can list the embedded fonts using standard tools. The pdffonts binary is pretty universal, as part of the Poppler set of tools.

And a fair number of old Adobe, Apple, and BeOS technical docs lack embedded fonts for some reason.

Can you send a link to an example? Which fonts were they using that weren't embedded?

Never in my life have I come across a PDF intended for public distribution missing an embedded font. (The sole exception being the PDF standard fonts like Times New Roman that are never embedded.)

I remember an Adobe presentation[1] tried to walk back the "14 standard fonts" thing as being an Acrobat "convenience" feature rather than a PDF standard feature, probably because it doesn't look good to attempt to standardize a (format unimplementable without a) set of fonts you're unwilling or unable to let people freely redistribute. (But then that presentation also promised the nascent ISO PDF standard would always stay accessible free of charge, and we know how that turned out.) Certainly PDF/A does not permit you to omit them, though that's only one specific kind of PDF.

In any case, sure, here you go:

- The Be Book[2] for DR8 uses but does not embed AvantGarde-* fonts.

- Inside Macintosh: Interapplication Communication[3] uses but does not embed Palatino-* fonts (for this one I could be convinced it's because the uploader merged the original per-chapter PDFs[4] incorrectly, though).

- The Mac OS 8 Human Interface Guidelines[5] also have the Palatino problem (and look legit, even though other Apple HIGs from that era do embed their fonts).

- Even the bloody spec[6] for PDF 1.3 uses and does not embed Caecilia-Heavy and MyriadMM_565_600_, whatever those are.

[1] https://www.youtube.com/watch?v=poc9PVmFzpc

[2] http://bebits.irixnet.org/be/docs/DR8/BeBook/acrobat/05_Medi... and others in that directory (1996 metadata)

[3] https://vintageapple.org/inside_r/pdf/Interapplication_Comm_... (1993 copyright, 2014 metadata)

[4] https://thrysoee.dk/InsideMacintoshInterapplicationCommunica...

[5] http://interface.free.fr/Archives/Apple_HIGOS8_Guidelines.pd... (1997 metadata)

[6] https://web.archive.org/web/20101214132912/http://partners.a... (2000 metadata)

Thanks!

I took a look at a few of these, and some of them are maybe just strange bugs.

E.g. the Adobe PDF spec embeds almost everything including most versions of Myriad, just not those two you mention. Similarly the Apple Guidelines embed Palatino Roman and Standard (as TrueType), just not Bold (as Type 1).

The Media Kit one does just not embed anything though (no Avant Garde), so that's clearly intentional.

It does make me wonder if technical documentation intended for a specific platform would sometimes try to save space by not embedding fonts standard not to PDF but to the platform. E.g. Avant Garde has shipped with Macs for a very long time. Still, what a terrible idea.

But fascinating to see documents with these problems in the wild, first time I've ever come across it. Thanks for taking the time!

As another data point, I've run into a variety of old DEC VMS documentation prominently featuring non-embedded New Century Schoolbook, e.g.,

http://www.vaxhaven.com/images/a/a7/AA-PS6KA-TE.pdf

It's probably not a coincidence that New Century Schoolbook was one of the fonts included with all PostScript Level 2 printers, but not included with older PostScript printers or Acrobat.

Note that the documentation in question was originally distributed as PostScript .ps files, not PDF, and it indeed makes no sense to embed New Century Schoolbook in a Level 2 or higher PostScript file, for both practical and likely legal reasons.

Per a handy diagram in MacWorld's review[1] of the original release of Myriad as one of the first multiple-master[2] Type 1 fonts, "MyriadMM_565_600_" is a semibold, regular-width instance.

Myriad Pro Semibold and Myriad Variable Concept Semibold should both be very close matches.

Myriad Variable Concept is bundled with the current version of Adobe Illustrator (and possibly with other Adobe apps not currently installed on my system); Myriad Pro was bundled with most (all?) versions of the pre-cloud Adobe Creative Suite, and has the benefit of being a traditional, non-variable OpenType font that should work pretty much everywhere.

This suggests an interesting question: are tools available to faithfully convert multiple-master Type 1 fonts to variable OpenType format?

"Caecilia-Heavy" is PMN Caecilia 85 Heavy; Caecilia LT Std 85 Heavy, a currently-available OpenType font, is presumably a close match.

Both Myriad Pro and Caecilia LT Std 85 Heavy are included in Adobe Font Folio 11 and, um, maybe available for activation via Creative Cloud, but this is not immediately clear from Adobe's Web site, and I'm having problems launching the damn CC app to check.

Of the many Adobe products I happened to have licensed before discontinuation in favor of a subscription-only replacement, Font Folio is probably my favorite. And, unlike older Adobe applications, it's still 100% compatible with every modern OS, and likely to remain so in perpetuity.

Or at least until OpenType is deprecated in favor of some dystopian online-only replacement…

[1] https://archive.org/details/MacWorld_9207_July_1992/page/n19...

[2] https://en.wikipedia.org/wiki/Multiple_master_fonts

Back then it could have been due to file size?

Coupled with the lack of builtin compression in contemporary PDF versions (and a strange but widespread aversion to gzipped pdf), perhaps, but then there shouldn’t be anything special about those particular companies. Maybe there isn’t, though, maybe that’s just where I come across mid-to-late-90s PDFs.

If the document was intended for distribution, it is likely that any non-embedded fonts are going to be common system fonts that will either have direct replacements or compatible replacements. Of course all bets are off, if someone created a document for personal use, to share with family/friends, or simply didn't know what they were doing.

Unless I'm missing something, if the fonts aren't embedded in the first place then it shouldn't make any difference. The text should be encoded properly so that it can render now without the fonts (in which case we know which glyphs to use), the same will be true later.

From what I understand, currently, if the font (actually: the parts of it needed to render the PDF) isn’t embedded, but you have the font installed, the renderer will use that font, an you’ll see the document as intended.

When support is removed, the renderer will look for/guess at the best replacement you have installed. As the article says, that may have subtle or not so subtle differences.

I don’t think it will be that bad, though. If you care, I think you’ll already have embedded fonts in your PDFs for decades, even for fonts that ‘everybody’ already has, because the probability that ‘my Font’ is exactly the same font as ‘your Font’ is fairly slim.

As you say, I feel like it probably doesn't really matter that much. If you're relying on a PDF rendering correctly based on an unusual local font then you're in trouble as soon as you send the document to another machine. If you have the Type1 font locally, you could always convert it.

You could have a situation where your font has really weird cmaps and without that specific font the text becomes garbled. More likely, you can substitute it for another font and it's mostly fine.

Keep in mind that the PDF standard has 14 base fonts that are used for a lot of documents that people send about. https://appligent.com/what-are-the-base-14-fonts/

I don't have a dog in this fight, but this sounds to me like an optimistic view of people and their behaviour.

Your typical “save as PDF” will embed fonts nowadays.

I think only professional (typically commercial, costing serious money) PDF writers have had a flag to NOT embed fonts in PDFs for decades.

> 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.

> 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.

> It does not burden a system to have it.

Any and all software is a burden.

There’s no code better than no code.

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

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)

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?

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.

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?

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

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

> 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.

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.

> 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).

Type 1 fonts are a subset of Postscript sufficient to describe glyph outlines, plus hinting and font specific metadata.

PDFs are a larger subset of Postscript that keeps more features but still strips it down from a programming language to a document description language while adding a stricter structure, embedding of files, and other features relevant to its use as digital paper.

The Venn diagram of PostScript, PDF, and Type 1 fonts has a LOT of overlap. Any useful Postscript or PDF viewer inherently has to support everything about Type 1 fonts other than hinting, and hinting only matters at sizes where subpixels matter.

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?

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

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

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

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

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

Oh, I thought it was Type 1 Fun:

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

I also expected sth. about diabetes.

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

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

I thought it was type 1 personalities.

(cc @dang)

There I was thinking they cured diabetes :-(

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.

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

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.

Hyperinsulinemia is reversible. You reverse insulin resistance by stopping the constant, excessive stimulation of insulin. That's the disease, that's what does the most damage over the longest period of time years and years before an a1c exceeds the thresholds on a test. Once people restore their sensitivity to insulin and they change their diet, eating habits, they don't have type 2 diabetes. They have normal a1c, they have normal Kraft tests. They don't need medical intervention. They don't need drugs anymore. They will have normal a1c and healthy glucose metabolism proved by advanced lipid panels. We've known since the Kraft tests in the 1970s that type 2 diabetes can be predicted atleast 5-10 years before an elevated a1c would allow a doctor to diagnose a patient as type 2 diabetes. Insulin resistance is accurately measurable for decades now. Type 2 diabetes is curable.

All of the things you've stated can also be explained as it being mitigated / in remission. It's no longer effecting the person, but that doesn't mean they no longer have the "condition".

To quote diabetic.org [1]

> So, can type 2 diabetes be cured? The answer is no. Type 2 diabetes is a chronic medical condition with no cure.

> Instead, it’s best to look at it as a manageable condition. For some people, remission is possible.

[1] https://www.diabetic.org/type-2-diabetes-can-it-be-cured/

The American Diabetes Assocation is not an organization I would trust very much. The advice on those websites is honestly laughable, deliberately misleading. I think the UK's version is even worse basically gaslighting. I could list a bunch of examples. It's actually shocking to read the UK's version there is so much careful misleading wording. Here's the official press release when they added that blurb about "remission" that you quote:

https://diabetes.org/newsroom/press-releases/2021/internatio...

2021? That's them being forced to play catch up LOL. That's a joke.

"Type 2 Diabetes is a chronic disease that can only be managed" was the mantra they stuck to and wouldn't budge, refusing to acknowledge growing science and even evidence from decades ago that was buried by bias. They couldn't ignore it any longer. They basically were forced to finally change their statements, they could no longer hide behind "Type 2 diabetes is chronic and only manageable."

"It's no longer effecting the person, but that doesn't mean they no longer have the "condition".

Again, if the person has normal a1c, normal cpeptide test, normal response in a Kraft test, normal lipid panel even if LDL is high, completely healthy NMR on LDL particles, normal healthy advanced lipid panel, biopsy shows healthy liver function, BMI normal. They have HEALTHY NORMAL glucose metabolism by all measurable standards.

What other "condition" do they still have that evidences type 2 diabetes? What test or evidence shows any evidence of "the condition". They may have damage leftover from all the harm that T2D causes, yes. But those are not T2D

It's the same story as the failed diet heart lipid hypothesis - incredible bias and ignorance, beaurocratic dsyfunction over many years. Who wrote the American Heart Assocation's first check? Procter & Gamble - right when Crisco was giving everyone heart attacks. Who puts their stamp of approval on sugary childrens cereals? The american heart assocation! Who's a funder of the American Heart Assocation? THe makers of breakfast cereals!

That said, Why would the American Diabetes Assocation accept funding from Pepsi and Coke? LOL.

The American Diabetes Assocation and the UK's version of that are the LAST organizations I would trust for the latest science on metabolic syndrome. They're funded by food industry and drug makers, atleast in US. I've lost most of my trust for them.

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

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"...

"type" is a very common and overloaded word. Consider its meaning for designers, programmers, taxonomy, keyboard enthusiasts, diabetics, Pokemon fans, and so on.

The website being called "typenetwork" doesn't clarify much to me.

If I had to pick a topic to associate "type 1" with, it would be diabetes, personally.

You're absolutely right that it's not clickbait, but it is confusing, considering we're not on a website focused on typography or graphic design.

>"type" is a very common and overloaded word

Yes. Just not in the audience of that publication.

You're just not their type of reader (pun intended)

The request was not to change the title on the linked website, but have a HN title reflect the clarification that is important for many HN readers.

The point is that the HN title should be changed in this case, even though the normal advice is to keep the original title.

My thoughts went; diabetes.. network provider. I have worked extensively with PS Type 1 fonts. The title is unnecessary short, even if this was a PostScript working group.

I clicked the comments here just to find out what type 1 means, I saw that the site was called "typenetwork", and that didn't mean anything to me.

It might not be clickbait as in it is not intentionally misleading, but it still is a very confusing title.

Not everyone that visits this site is a typography expert...

Which made me think it was going to be about Haskell.

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

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

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.

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.

Just append the word "fonts" to the title.

I was expecting the end of Type 1 diabetes.

I like this one better

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

Type 1 without extra qualifiers refers to diabetes.

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

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)

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.

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"

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

> Type 1 without extra qualifiers refers to diabetes.

That may very well be accurate.

In your world, though.

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

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]

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

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.

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

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

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.

> (remember that "tech" does not mean "computer science")

If you're going to be that nitpicky, this article isn't about anything that could reasonable be called "computer science" either

> I'm not sure what "this context" refers to.

HN, see https://news.ycombinator.com/newsfaq.html

Did you read that? Because it doesn't contradict what I said.

You probably meant this: https://news.ycombinator.com/newsguidelines.html

And that doesn't mention computer science either. Instead it says this:

> On-Topic: Anything that good hackers would find interesting. That includes more than hacking and startups. If you had to reduce it to a sentence, the answer might be: anything that gratifies one's intellectual curiosity.

Note that this doesn't even mention technology, just "startups" and "hacking" (which to be clear isn't specifically about technology either).

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

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

To you it does.

Reply on news.ycombinator.com