Overprinting two glyphs

[elipsett]

I want to do something and don't know if it's possible in FontCreator or not.
I'm hoping someone can explain how to do it, or state that it's not possible.

For purposes of explanation, say that I have an underbar (_) mapped to keyboard character A and a plus mark (+) mapped to keyboard character B as two separate glyphs in a font.
Is there any way for the user to type A and then B on the keyboard to overprint the two glyphs, producing something that looks like a plus-minus symbol (±)? This is merely an example, I understand that there already is a plus-minus glyph, and I understand I can make a new plus-minus glyph inside the font.

My question is whether it is possible for the user to type A and B to overprint these two glyphs.
It makes no difference HOW this is handled internally — pair adjustment, kerning, diacriticals, whatever — but the point is that the glyph ± does not exist in the font, and is created on-the-fly by the user from two components that are in the font.

I'd appreciate any suggestions you may have to offer.

[Bhikkhu Pesala]

I already answered this question in your previous thread.

Design B like a diacritical mark with zero advance width and a negative left side-bearing.

[elipsett]

Yes, you did reply to my question.
However, as I pointed out, either your solution does not solve my problem, or I am doing something wrong.
I asked for clarification and didn't get any, so I rephrased the question and tried again.

If you can't help me understand what the problem is, please stop responding.

[Erwin Denissen]

It heavily depends on the software you use. If you make a font with OpenType layout features to solve your problem, but the software does not support these OpenType layout features, it won't work.

Based on your information I still think mark-to-base is the way to go, but maybe the software doesn't support it, and then you might be able to solve the problem with ligatures.

[elipsett]

Thank you, Erwin.
So apparently it is not impossible to do what I want.

In which case there are several more questions, if you have the time:
1. Would Word support this function for a mark-to-base implementation?
2. Would Word support this function for a ligature implementation?
3-4. Same questions but with Adobe Illustrator.

And lastly, since I made a test font with the mark-to-base implementation (assuming I made it correctly, of course),

5. Does the "Test current font as TTF/OTF" function in FontCreator support this sequential-key input method for these characters?
I ask because it displayed the two glyphs (base and mark) sequentially without overprint, even though the image displayed in FontCreator when making the font is overprinting correctly.

Edward

[Erwin Denissen]

Yes, I think this should all work even in the Font Test dialog.

It should also work in your web browser.

[elipsett]

In which case it seems I simply haven't made it correctly.

Would you (or someone) be willing to look at a test font and get it working for two characters?
If I can see what you've done, I should be able to make the needed font myself.

I'm attaching the test font here in hopes that someone can spare the time.
The base glyphs are numerals 0-9, and the marks are upper-case C and D.
In theory, they are all already defined properly, but since it doesn't work pretty clearly I've screwed up...

Edward

[Erwin Denissen]

The font you provided did not contain any anchor for the glyphs. I have added a top anchor to several glyphs; see the attached font project.

However, Word will treat common Latin characters as base character, so this will solution will not work in Word. It does seem to work in a web browser though.

In your case a special trick might work: keep C as a base character, but replace it with a glyph, C.mark which is basically a copy, but it a mark. This substitution can be added to the liga feature, but you will need to enable Standard Ligatures in Word.

testmark.png (134.88 KiB) Viewed 2144 times

[elipsett]

Thank you, Erwin.

I see several things I need to learn more about before I can get anywhere.
I appreciate the time you spent.

Edward

[elipsett]

Thank you again.

After a large number of failed experiments I've finally understood what you did, and made a new font that works.
Now that the mechanics are working, building the actual font will be pretty simple.
Long and boring, but simple.

Edward

[elipsett]

Well, making the font was not complicated.
Thank you.

It can input the needed glyphs into Word correctly, once ligatures are enabled.

Two new problems, however:
The Word document cannot be saved as a PDF without printing the base and ligature separately (adjacent to each other on the line).
And they cannot be used in Illustrator. The AI issue may be due to the fact that we're using the Japanese version of AI, which doesn't handle ligatures the same way as the European language versions.

Do you have any ideas on how to approach these problems?
Or are these issues only resolvable on the app side now?

Best,

Edward