Your favourite font choices

[Bobby Henderson]

Bobby, the actual shapes are protected by law in many countries around the world. Is it so hard to accept that some countries have laws that literally is called "Schriftzeichengesetz" or "Character law" or rather "Typeface law"? I've told you why URW can make as many copies of Helvetica as they want and even sell it in Germany: Helvetica was developed in 1957, the copyright to that typeface has since expired in Germany. Furthermore, URW can sell whatever they want in the U.S., as long as the fonts aren't literal copies of the original.

First of all, Helvetica is NOT a public domain typeface. The Helvetica name is even a registered trademark. Its copyright has been renewed numerous times thru the decades as it has been re-released for newer printing technologies and then changing digital font formats. Copyrights for typefaces can also be renewed when new characters or entire alphabets are added to the character set. Linotype, which is a subsidiary of Monotype, currently sells Helvetica commercially in OpenType format. They even sell Neue Haas Grotesk, which is a "restoration" by Christian Schwartz of Max Miedinger's drawings that eventually turned into Helvetica.

The only thing about Helvetica that could fall into the public domain is its original analog drawings made in the 1950's. But even if another type company accessed those drawings to make their own version of Helvetica they wouldn't be able to call the result "Helvetica."

URW has released imitations of many commercial typefaces that were released far after Helvetica. I mentioned Nimbus Sans earlier; that's really an imitation of the 1980's Helvetica Neue. URW has churned out many knock-offs fonts of 1970's-80's Letraset typefaces, and started doing so in the 1980's and 1990's. Same for a bunch of fonts from ITC and other foundries. URW is not the only company doing this in case it sounds like I'm singling them out.

If the German government was able to enforce design rules of Schriftzeichengesetz as you say they should, URW would not have been able to make any of those knock-off fonts. Many of the sources of their knock-offs were well within 10 years of their original release (plus 15 additional years of protection, if the German government is paid a fee).

But even 70 years of work in that field, wouldn't make you or me an International legal professional.

Yet you're here trying to act like a legal expert and that you know more about type than I do. That's regardless of my education, occupation, decades of experience, purchase history of actually buying type, etc. That would be like me trying to tell Brad Miller how he should really thread a film projector.

No, you're not. Or maybe, you temporarily are... but really, you kind-a love this, don't you?

I love this about as much as someone could love taking a swim in a sewage treatment plant. Basically this thread is pretty much getting killed dead because of your circular trolling. I certainly have zero interest in responding to this shit any further.

[Marcel Birgelen]

What you cited does not get into the specifics of just what is actually copyrighted in a typeface. A company can copyright the font file data and font names. Hell, some companies have even trade-marked font names. But the actual letter shapes are still not protected. That's why you can have a German typeface company like URW churning out imitations of well known typefaces like Helvetica and naming them Nimbus Sans. If the case example you cited applied to the actual letter shapes URW would have never been able to make the hundreds of imitation fonts they have produced.

Bobby, the actual shapes are protected by law in many countries around the world. Is it so hard to accept that some countries have laws that literally is called "Schriftzeichengesetz" or "Character law" or rather "Typeface law"? I've told you why URW can make as many copies of Helvetica as they want and even sell it in Germany: Helvetica was developed in 1957, the copyright to that typeface has since expired in Germany. Furthermore, URW can sell whatever they want in the U.S., as long as the fonts aren't literal copies of the original.

No, you started suggesting this crap via the Adobe Fonts service. You kept insisting that if someone converted font objects to outlines they could still be liable if they discontinued their Adobe CC subscription and/or the type foundry removed its fonts from the Adobe Fonts service. That is WRONG. Once the lettering is converted to outlines it is of zero value to any type company.

And around here, this is still a matter of concern. Given what situation the world is in, there are probably bigger concerns right now, but that doesn't mean that it doesn't exist.

I'm really tired of going round and round with you on this nonsense.

No, you're not. Or maybe, you temporarily are... but really, you kind-a love this, don't you? You don't need to admit it. But hey, why would you otherwise spend so many words on this? :P

I've been doing paid professional graphic work full time for a living for nearly 30 years. I've studied this shit in college. Digital fonts and the legal details surrounding them are a critical element in my work. But you're acting like I don't know what I'm talking about.

But even 70 years of work in that field, wouldn't make you or me an International legal professional. And let me quote yourself, so you have something new to rant about, like how I'm quoting you out of context... maybe a little, but not really though:

The whole thing can turn into a quite a deep rabbit hole.

[Bobby Henderson]

There are actual laws:

What you cited does not get into the specifics of just what is actually copyrighted in a typeface. A company can copyright the font file data and font names. Hell, some companies have even trade-marked font names. But the actual letter shapes are still not protected. That's why you can have a German typeface company like URW churning out imitations of well known typefaces like Helvetica and naming them Nimbus Sans. If the case example you cited applied to the actual letter shapes URW would have never been able to make the hundreds of imitation fonts they have produced.

I guess you're the only one who's suggesting this. I've pointed out why a type foundry wouldn't want to do exactly this. Yet, there are more people to the party than just type foundries who create fonts.

No, you started suggesting this crap via the Adobe Fonts service. You kept insisting that if someone converted font objects to outlines they could still be liable if they discontinued their Adobe CC subscription and/or the type foundry removed its fonts from the Adobe Fonts service. That is WRONG. Once the lettering is converted to outlines it is of zero value to any type company. The only thing that would matter at all is if the person was legally using the fonts at the time the artwork was created. Once he converts the objects to outlines it doesn't matter if he has legal access to the font files anymore.

I'm really tired of going round and round with you on this nonsense. I've been doing paid professional graphic work full time for a living for nearly 30 years. I've studied this shit in college. Digital fonts and the legal details surrounding them are a critical element in my work. But you're acting like I don't know what I'm talking about.

[Martin Brooks]

I wonder if the novelty of having umpty-zillion available fonts has worn off to some extent. Twenty years ago almost every printed document looked like a ransom note. Today, not so much.

Yes, I always called poorly designed documents "ransom notes". But I think we've gone too far in the opposite direction since texts and most phone emails don't permit font selection. I can't stand it when a monospaced document comes into my email, especially if there's a lot of text. It's like time travel back to 1983.

Edit: And I just discovered here that if you choose a different font in the editing window, it doesn't "stick" when you post. But I guess no one ever noticed because no one ever bothered to choose a different font.

[Marcel Birgelen]

No it's not "reality" outside the US. Until there are legal precedents that change the nature of raw letter shapes the topic will remain a philosophical discussion. I have zero to worry about.

There are actual laws:

In 1981, West-Germany passed the Gesetz zum Wiener Abkommen vom 12. Juni 1973 über den Schutz typographischer Schriftzeichen und ihre internationale Hinterlegung ("Law on the 1973 Vienna Agreement for the Protection of Type Faces and their International Deposit", also simply known as Schriftzeichengesetz or "Type Faces Law"), according to which a typeface is initially protected under German copyright law for 10 years from first publication on. After the end of this initial ten-year period, the rights holder may pay a fee to prolong copyright status for additional 15 years only once. According to German law, every typeface thus ends up in the public domain after no more than 25 years from first publication onwards, after which it is also free to be digitized into a computer font, which in itself holds a much higher copyright protection status by German law than analogue typefaces due to being legally classified as a computer program.

And yes, there are legal precedents under that law too...

It's even more ridiculous to suggest a commercial type foundry would try to gouge graphic designers, ad agencies or their clients for an extra paycheck for using their fonts in paid projects after they were already using the fonts legally, either by purchasing outright or thru something like the Adobe Fonts service. That kind of double-dipping would never go over well anywhere in the world.

I guess you're the only one who's suggesting this. I've pointed out why a type foundry wouldn't want to do exactly this. Yet, there are more people to the party than just type foundries who create fonts.

[Bobby Henderson]

It's not a philosophical discussion, it's reality outside of the U.S.

No it's not "reality" outside the US. Until there are legal precedents that change the nature of raw letter shapes the topic will remain a philosophical discussion. I have zero to worry about.

Maybe, many of those companies selling those fonts do have their legal work in order. Maybe they do have the necessary licenses or copyrights to sell those fonts, even if they acquired them from some half-defunct foundry two decades ago. Fonts have been cross-licensed all over the place, it's probably even a bigger mess than licensing on music.

You're just guessing here, trying to conjure up a problem with digital fonts when none exists. Helvetica and many other long established typefaces have had copyrights renewed. New updated versions get drawn every now and then, like the 1957 cut of Helvetica being updated in 1983 with Helvetica Neue and in 2019 with Helvetica Now. Nevertheless type companies around the world can release their "clones" of such faces, like URW did with Nimbus Sans or Bitstream with Swiss 721, and fear any legal consequences.

It's even more ridiculous to suggest a commercial type foundry would try to gouge graphic designers, ad agencies or their clients for an extra paycheck for using their fonts in paid projects after they were already using the fonts legally, either by purchasing outright or thru something like the Adobe Fonts service. That kind of double-dipping would never go over well anywhere in the world.

[Marcel Birgelen]

Philosophical discussions are one thing. Legal precedents are another. When typographical objects are converted to outlines the objects are no longer fonts. Until there is actual specific case law deciding otherwise then that's not going to change.

Good quality commercial fonts are already costly enough to legally buy outright. And it's not the easiest thing to get people to purchase the font files rather than downloading them under more shady methods. The idea of a type foundry demanding extra money from a paying customer for using its type in a logo or other kind of graphic design is all kinds of stupid. If I bought some fonts from a type foundry and they came along wanting a double dip pay day for me using the fonts in paid work I'd never buy anything from them again. I don't think any type companies in existence would be that foolish. Such tactics would generate very bad word of mouth.

It's not a philosophical discussion, it's reality outside of the U.S. Yeah, it's strange, it's usually the U.S. known for draconian copyright measures and patents so broad, they include everything from the wheel to sliced bread. But those discussions and (potential) lawsuits are a thing around here.

Clone digital typefaces have been sold around the world since the 1980's and continue to be sold today. They have been bundled in software all that time as well. If there was any doubt about such fonts' legal status in the rest of the world they wouldn't be sold at all. More than just American companies make and market knock-off clone typefaces. I mentioned URW earlier; they're based in Germany.

Lots of stuff comes into play here...

Maybe, many of those companies selling those fonts do have their legal work in order. Maybe they do have the necessary licenses or copyrights to sell those fonts, even if they acquired them from some half-defunct foundry two decades ago. Fonts have been cross-licensed all over the place, it's probably even a bigger mess than licensing on music...

In Germany, many copyrights on popular typefaces have since officially expired and are in the open domain. Anything like Helvetica won't stick, the same is true for anything else that's 35 years or older or all stuff for which licenses haven't been explicitly renewed 10 years after first publication. So, suing anybody for some font catalog containing a bunch of 40 year old knock-off fonts isn't going anywhere.

Furthermore, copyright and IP cases in general around here usually don't yield the payoffs like they do in the U.S. Stuff like statutory damages are way more limited and if awarded, which usually requires you to prove that the other party acted in bad faith, will mostly just cover the costs for the legal action in itself. Yet, like you can see yourself, there have been cases and those cases have usually been decided in favor of the copyright holder.

Also, I'd say that most Europeans are more litigation-averse. Most legal disputes here get settled before stuff hits a court. We don't know how much of those disputes have never seen a day in court.

Page layout applications, vector drawing programs and other kinds of graphics software default to placing the first letter's origin point on the left margin line rather than trying to align the left edge of the glyph. Not all letters have the same amount of physical space between the left edge of the letter and the origin point. Curved letters like "O" will get closer to the origin point. Some letters may extend left of the origin point. Any attempts to fudge that may result in a left margin of body copy that doesn't look optically proper.

I'd say that's in-line with my observations. I only tested Adobe Illustrator, InDesign and Word. While Adobe's software will honor the origin for the first glyph by default, MS Word doesn't seem to like drawing past the left margin and moves the first glyph to the right for the entirety of the LSB part. In all subsequent glyphs on that row, the origin is preserved as expected.

[Harold Hallikainen]

Thanks! I'm using https://www.freetype.org/freetype2/d.../glyphs-3.html for my ideas on origin locations, bearing, etc.

Harold

[Bobby Henderson]

Look Bobby, by now you may realize that I usually don't tend to just make stuff up while I'm at it. I'm not your legal specialist, but there is enough info out there to conclude that there are quite some debates going on about typefaces and their legal status around the world. Many jurisdictions allow copyrights or design patents on abstract shapes, which include typefaces. There are even advocacy groups "fighting" (last site update back in 2007...) to get typefaces recognized under copyright law in the U.S.

Philosophical discussions are one thing. Legal precedents are another. When typographical objects are converted to outlines the objects are no longer fonts. Until there is actual specific case law deciding otherwise then that's not going to change.

Good quality commercial fonts are already costly enough to legally buy outright. And it's not the easiest thing to get people to purchase the font files rather than downloading them under more shady methods. The idea of a type foundry demanding extra money from a paying customer for using its type in a logo or other kind of graphic design is all kinds of stupid. If I bought some fonts from a type foundry and they came along wanting a double dip pay day for me using the fonts in paid work I'd never buy anything from them again. I don't think any type companies in existence would be that foolish. Such tactics would generate very bad word of mouth.

Yes, there are tons of knock-off fonts around the globe. Their legal status is clear in the U.S., but they can be pretty doubtful in the rest of the world.

Clone digital typefaces have been sold around the world since the 1980's and continue to be sold today. They have been bundled in software all that time as well. If there was any doubt about such fonts' legal status in the rest of the world they wouldn't be sold at all. More than just American companies make and market knock-off clone typefaces. I mentioned URW earlier; they're based in Germany.

And maybe you can ring in on parts of Harold's question of how you think a font should be rendered: Should the first glypth be rendered with the origin at zero, with the left-bearing part of the glyph sticking out to the left, into the left margin (if there is one) or should the first glyph be rendered starting at the left-bearing extreme to be the origin, so that there will be nothing sticking out over the left margin?

Page layout applications, vector drawing programs and other kinds of graphics software default to placing the first letter's origin point on the left margin line rather than trying to align the left edge of the glyph. Not all letters have the same amount of physical space between the left edge of the letter and the origin point. Curved letters like "O" will get closer to the origin point. Some letters may extend left of the origin point. Any attempts to fudge that may result in a left margin of body copy that doesn't look optically proper.

Actually, I think the horizontal origin is usually to the left of the glyph bounding box, so if you place the horizontal origin on the margin, there will be a small gap between the margin and the left side of the bounding box (and this gap would be the left side bearing).

The origin point is on the left edge of the Em Square, not outside of it. Left and right side bearings will go to the left of the origin point and some distance to the right of the letter's right edge. But the value can vary letter by letter.

When I type out a string of text in an application like Adobe Illustrator or CorelDRAW the baseline will be a visible part of that type object. The baseline typically extends slightly past the left edge of the first letter and a little to the right of the last letter in the string. Illustrator and CorelDRAW will both allow users options to align and position type objects based on the edges of the letters. Due in part to my frequent requests Illustrator also now allows sizing and positioning of type based on capital letter height or lowercase x-height.

[Harold Hallikainen]

I

And maybe you can ring in on parts of Harold's question of how you think a font should be rendered: Should the first glypth be rendered with the origin at zero, with the left-bearing part of the glyph sticking out to the left, into the left margin (if there is one) or should the first glyph be rendered starting at the left-bearing extreme to be the origin, so that there will be nothing sticking out over the left margin?

Actually, I think the horizontal origin is usually to the left of the glyph bounding box, so if you place the horizontal origin on the margin, there will be a small gap between the margin and the left side of the bounding box (and this gap would be the left side bearing). Thinking of hand set type, I think the left side of the sort would be pressed up against the left margin leaving the left side bearing of that glyph between the margin and the bounding box of the glyph. As mentioned before, use of the origin as the positioning reference seems to handle leading whitespace well and can be applied to either horizontal or vertical text (using the corresponding origin) positioning.

Harold

[Marcel Birgelen]

I'll bet the guy's legal actions went nowhere. I don't know 100% how it works over in Europe, but here in the US work for hire is something that gives employers a great deal of legal cover. If the former employee was drawing a salary and benefits from a company while designing the fonts it would actually be possible for the company to say they own the rights to those fonts. Unless the guy managed to get a contract specifically stating he retained the rights to anything he created while working for that company he would likely be out of luck in court.

It was settled, out of court. Apparently, the guy still got a nice payday. The guy was only part-time employed by said employer, so he probably had some standing ground to claim that he made those fonts in his own time and not for the employer.

No. There are no debates about it. If you want to insist there are disputes I want some specific examples (like actual legal cases) rather than just vague conjecture. If there was any debate at all about raw vector shapes of letters it would be legally impossible for any company to create visually identical "clones" of popular typefaces, such as Helvetica. Bitstream and URW are a couple companies that have made quite a few typeface clones over the years. A bunch of those clone typefaces have been bundled into a variety of design applications. They're even commercially sold on their own. Those applications and those fonts are sold to customers all over the world.

Look Bobby, by now you may realize that I usually don't tend to just make stuff up while I'm at it. I'm not your legal specialist, but there is enough info out there to conclude that there are quite some debates going on about typefaces and their legal status around the world. Many jurisdictions allow copyrights or design patents on abstract shapes, which include typefaces. There are even advocacy groups "fighting" (last site update back in 2007...) to get typefaces recognized under copyright law in the U.S.

Yes, there are tons of knock-off fonts around the globe. Their legal status is clear in the U.S., but they can be pretty doubtful in the rest of the world. This legal doubtful state doesn't mean there are tons of lawsuits happening right now. Lawsuits don't only bear financial risks, but like you indicated yourself, you may end up looking like the bad guy and this combined with potential reputational damages may simply not be worth it to start any legal disputes. But if you look at e.g. the Wikipedia article, or the TypeRight site, you can see, that there have been some court cases over the years, with some different outcomes. Also, if you check the Wikipedia article, you'll see that e.g. Germany, has a very specific law regulating typefaces and their copyright, even with built-in expiration dates and an explicit need to prolong copyright protection after the first 10 years.

And maybe you can ring in on parts of Harold's question of how you think a font should be rendered: Should the first glypth be rendered with the origin at zero, with the left-bearing part of the glyph sticking out to the left, into the left margin (if there is one) or should the first glyph be rendered starting at the left-bearing extreme to be the origin, so that there will be nothing sticking out over the left margin?

[Bobby Henderson]

The relationship between that agency and him was not meant to last, so their ways parted. I don't know any of the particular details, but he ended up suing or at least threatening his former employer, because they were using some of his typefaces in their clients work. This also included at least one company logo and a few product logos. He apparently also threatened the end-users of those logos with legal action. He claimed they owed him royalties for the use of his typefaces.

I'll bet the guy's legal actions went nowhere. I don't know 100% how it works over in Europe, but here in the US work for hire is something that gives employers a great deal of legal cover. If the former employee was drawing a salary and benefits from a company while designing the fonts it would actually be possible for the company to say they own the rights to those fonts. Unless the guy managed to get a contract specifically stating he retained the rights to anything he created while working for that company he would likely be out of luck in court.

That brings back that example I brought up about Tobias Frere-Jones and his former "partner" Jonathan Hoefler. I put that word in quotes because when Frere-Jones went to work with Hoefler to start the HFJ foundry he thought they were equal partners. Hand-shake agreement basically. Well, Hoefler ended up pocketing most of the earnings and screwing Frere-Jones out of his cut. And Frere-Jones' creations were the real star of the foundry's portfolio. They ended up going to court where Hoefler claimed Frere-Jones was merely an employee and not a partner (regardless of Frere-Jones' initials being in the brand). Frere-Jones ended up getting robbed again. Hoefler retained the rights to Gotham and other typefaces Frere-Jones created while there.

The odd thing is Frere-Jones obviously had a better arrangement when he worked with Font Bureau. When FB pulled its typefaces from the Adobe Fonts service Frere-Jones was able to make his own deal with Adobe to get typefaces like Interstate back on there.

Regarding your conversion of type objects to outlines: Like I said, there really are some debates about this.

No. There are no debates about it. If you want to insist there are disputes I want some specific examples (like actual legal cases) rather than just vague conjecture. If there was any debate at all about raw vector shapes of letters it would be legally impossible for any company to create visually identical "clones" of popular typefaces, such as Helvetica. Bitstream and URW are a couple companies that have made quite a few typeface clones over the years. A bunch of those clone typefaces have been bundled into a variety of design applications. They're even commercially sold on their own. Those applications and those fonts are sold to customers all over the world.

It looks like we have some font experts here! I have been trying to figure out how to determine the X coordinate of the vertical origin in a font file. The Y value is either specifically listed (in the VORG table, as I recall) or in the VMTX table as a top side bearing which can be added to yMax to yield the vertical origin Y value.

All digital fonts have built in dimensions for each glyph. There are multiple horizontal guide lines to establish baseline, x-height, cap height, ascender and descender. Font files will contain metrics data for things like kerning and kerning pairs. Each glyph exists inside a UPM or Em Square. Font editing applications such as FontLab or Glyphs can display those details. And the details can vary from one typeface to the next.

The origin is the zero point on the baseline near the left edge of a glyph. It's rare for the left edge of any glyph to be placed directly on the origin. For something like a capital "E" there will usually be a bit of space. Some letters will over-shoot left of the origin. Then there are left and right side bearings that can adjust depending on letter combinations.

FontLabImage-HelvXBlk-E.png

Digital typeface design is pretty complicated and growing even more so with the arrival of newer type technologies like OpenType Variable fonts and OpenType-SVG.

[Harold Hallikainen]

Thanks for the comments! I've read some of the documentation on Open Type and have not fully figured out a lot of stuff yet. I'm not trying to create my own renderer, but, instead, understand how existing renderers are interpreting font files. This is with regard to subtitles, both horizontal and vertical. Here are a few examples.

1. Horizontal Position - Imagine we have a grid in main picture with a line every 5% of screen width and height. If we have a subtitle that begins with a pipe character (which has a large left side bearing), an Halign of left, and an Hposition of 10%, some renderers place the left side of the pipe bounding box on the 10% grid line, and some place the left side of the bounding box to the right of the grid line. I believe gap is the left side bearing. This would indicate that the horizontal origin is being placed at Hposition. I like this approach instead of placing the bounding box since behavior of leading whitespace characters is then easy to define.
2. Vertical Text Vertical Position - Some renderers place the bounding box top at Vposition. Some place the horizontal origin (horizontal baseline) at Vposition. I think it would be best to place the vertical origin at Vposition, which would result in a small gap (the top side bearing) between Vposition and the top of the bounding box. This positioning would correspond with the use of the horizontal origin for Hposition as described above.
3. Horizontal Position of Glyphs in Vertical Text - In vertical Japanese text, the bounding box of most glyphs is centered on Hposition. However, some, especially punctuation, is not centered. For example, I think a comma is to the left of the center when written in horizontal text and to the right of the center when written in vertical text. I THINK this is similar to our placing the comma close to the previous word in English. In Japanese horizontal text, the same is done. In vertical text, it is still closer to the previous word but, since the columns of text are read right to left, the "closer to the previous word: is offset to the right instead of the left. I THINK the vertical origin should be placed at Hposition with vertical text. I think characters are defined based on the horizontal origin (on the horizontal baseline to the left of the bounding box) with the outline drawing instructions based on the horizontal origin. I think the vertical origin (generally centered above the character) used for vertical writing is specified as an xy offset from the horizontal origin. The Y offset is either in the VORG table or is the yMax plus the top side bearing. But, figuring out the X offset appears to be documented less.

So, that's what I'm trying to figure out!

Thanks!

Harold

[Marcel Birgelen]

I'd not consider myself a font expert... neither in the whole art of designing fonts, nor in the art of building software to render them. Just like you, you're sometimes confronted with those things and then you try to learn what's necessary to get a somewhat clear understanding of the subject.

You're writing your own font rendering engine? Because... yeah... welcome to hell. . Isn't there a particular reason you can't use one of the many libraries out there that can render OpenType for you? I've happily used FreeType for this kind of jobs over the years. It's built in C and compiles on anything that's sufficiently beefy to render OpenType. :P

But back to your question, for the Y value it's one of these:
- It's either in the VORG table.
- If there is no VORG table, then it's the top of the bounding box, which can, in turn be calculated from the stuff in the VMTX table.
Fun stuff...

According to the specs, the X origin is always zero. So, the X origin for your next glyph will simply be the "advance width" + your original X origin. Any "left side bearing" stuff will go beyond the X origin, but that's the entire point of "lsb".

As for determining the origin for the first glyp, it depends on your rendering mode... If you want to assure your entire rendered typefaces are inside a particular bounding box, then, your X origin should be X - lsb. The lsb value is in the HMTX table (I had to google that...). In that case, you can be assured that there will be no stuff sticking out to the left of your X-origin. In all other cases, X simply is 0. I guess it depends on your application domain and on the type of fonts you're using.

There are those fonts I'd call "stupid fonts" that put their "center" on X=0, in that case, there is a ton of lsb, not just some decoration sticking out to the left side, but practically half of the entire glyph... If you render them in "chop-chop-mode", or if there simply is no margin to the left, you're missing about the first half of every first letter in a row, a symptom you may have seen in the wild.

As for what's more common practice? I don't know. There doesn't seem to be a hard answer to this. Most Adobe product seem to honor the X=0 origin, while to me it looks like MS Word is using X - lsb... I don't have ISO 14496-22 at hand, maybe it does say something useful about what's considered to be standard? I'd consider Adobe software to be the more "typographically correct" software though...

[Harold Hallikainen]

It looks like we have some font experts here! I have been trying to figure out how to determine the X coordinate of the vertical origin in a font file. The Y value is either specifically listed (in the VORG table, as I recall) or in the VMTX table as a top side bearing which can be added to yMax to yield the vertical origin Y value.

It SEEMS that the X coordinate (which I interpret as the offset from the horizontal origin to the vertical origin) is the in the BaseCoord table in the VertAxis table in the BASE table, but I have not been able to confirm this. ISO/IEC 14496-22 2019-1 is complicated! I have not yet found anything definitive on the horizontal alignment of glyphs in vertical text.

One final question, in horizontal writing, it you go to place a string at a particular location (such as a left margin), is the horizontal origin of the first glyph placed there? That would leave the left side bearing of the first character between the bounding box of the character and the margin. This makes the most sense to me, but what is common practice?

Harold