Help with Composite Glyphs

[Haldrik]

1.
How do you stop the bearings from moving around while adding a new contour or glyph? Is there a "lock bearings" button anywhere? (It's maddening to deal with the unstable bearings when adding a diacritical mark to a letter.)

2.
Is there any way to copy and "paste special" a composite glyph ... as a new composite glyph (without it converting into contours)?

[Haldrik]

I just downloaded the Font Creator 5.0! Checking it out now. Questions 1 and 2 still seem to apply. Another question:

3.
I can't seem to get the new "Complete Composites" feature to work for the "Greek Extended" range.

I looked at the XML, and it seems the data for the Greek Extended range isn't present?

[Bhikkhu Pesala]

The Greek Extended data for CompositesData.xml has not been added yet. I was hoping someone else with knowledge of Greek would do it.

I didn't give it much thought to be honest, assuming that few if any users would need it. I needed some of the Latin Extended Additional characters myself, so I thought I might as well complete the character set, but only half expecting that a few Vietnamese users might need them.

[Bhikkhu Pesala]

Erwin has posted my updated version of CompositeData.xml for those who need to complete composites for the Extended Greek character set. Make a backup copy of the existing CompositeData.xml file, and save this update to your <install directory\Composites\ folder.

Your font will need to contain the basic Greek characters in the Greek and Coptic character set. The diacritics for extended Greek, which are required for the compose composites feature to work, can themselves be composed from spacing modifier accents, diaeresis, grave, acute, dot accent, and right single quotation mark. Accent Prosgegrammeni must be created as a simple glyph.

It is recommended to adjust the position of diacritics for lowercase alpha, before manually adjusting the position for other characters using snap to guidelines. As always with this feature, the precise position is only correct for the font used while creating the data file. All other fonts will require manual adjustment of diacritics, and entry of some font metrics.

Extended Greek Character Set
Copy this data to the Insert Characters dialogue to add the entire Extended Greek Character set.

7936-7957,7960-7965,7968-8005,8008-8013,8016-8023,8025,8027,8029,8031,8032-8061,8064-8116,8118-8132,8134-8147,8150-8155,8157-8175,8178-8180,8182-8190

Please reply to this thread if the feature is not working as expected.

[Joop Jagers]

>>snip<<
Accent Prosgegrammeni must be created as a simple glyph.
>>snip<<
Extended Greek Character Set
Copy this data to the Insert Characters dialogue to add the entire Extended Greek Character set.

7936-7957,7960-7965,7968-8005,8008-8013,8016-8023,8025,8027,8029,8031,8032-8061,8064-8116,8118-8132,8134-8147,8150,8157-8175,8178-8180,8182-8190

Please reply to this thread if the feature is not working as expected.

Thank you for adding the Greek Extended set to the CompositeData.xml. I'd like to make three remarks:
- Apparently you omitted the characters 8151-8155 in the Greek Extended set you listed in your message; fortunately they are present in CompositeData.xml.
- I prefer to use the diacritics which are part of the Greek Extended set, especially the combined ones, because these are easier to position than separate components.
- Actually, there's no need to create a simple prosgegrammeni. In combinations with accented Capitals a regular sized lower case iota can be used, in combinations with unaccented Capitals I prefer a regular Upper case Iota.

[Bhikkhu Pesala]

Thanks for the feedback. With a little more effort, we can get this right.

I edited my post to include the missing characters: 8151-8155

I agree it is better to use the combined accents where they exist. It would make positioning easier. I thought I had used only Greek extended accents, except for dialytika (diaeresis) which doesn't exist as a separate accent. I will look at that again. Edit: I couldn't find any place where I used two single accents where one double accent could have been used. Can you find any?

In the Gentium font that I used for the data, prosgegrammeni was indeed just a simple iota. I was misled by looking at Titus Cyberbit Basic, in which prosgegrammeni is a subscript, which looks like an accent. Which form is better, or are they both used? Either could be composed from iota. Edit The Unicode chart also shows prosgegrammeni as a very small iota, but positioned after the letter, not subscripted like Titus.

ἧἨᾗᾘ

[Joop Jagers]

[quote="Bhikkhu Pesala"]
>>snip<<
I agree it is better to use the combined accents where they exist. It would make positioning easier. I thought I had used only Greek extended accents, except for dialytika (diaeresis) which doesn't exist as a separate accent. I will look at that again. Edit: I couldn't find any place where I used two single accents where one double accent could have been used. Can you find any?

In the Gentium font that I used for the data, prosgegrammeni was indeed just a simple iota. I was misled by looking at Titus Cyberbit Basic, in which prosgegrammeni is a subscript, which looks like an accent. Which form is better, or are they both used? Either could be composed from iota.
>>snip<<
[quote]

You're right, you *do* use Greek extended accents; I didn't realise you composed the combinations as well. Personally I don't like to use multiple composed glyphs, because they are changed to simple glyphs in FontLab.


On the proper way to print prosgegrammeni you might have a look at http://omega.enstb.org/yannis/pdf/boston99.pdf page 13.

Thanks for your interest in these rather obscure matters

[Bhikkhu Pesala]

The article is too much for me to take on board. We do not need to concern ourselves with how Greeks or classical scholars want to write Greek. All we need to decide is how best to compose the composite glyphs - whether with small iota (953) following like Gentium, or with prosgegrammeni accent (890) underneath like Titus. The font designer is still at liberty to design the glyphs any way he or she likes, but won't have the benefit of automation.

If there is any problem with the way it is, please let me know, and it can be fixed for the benefit of all potential users of Font Creator.

I will add the following definition to compose prosgegrammeni accent as a small, subscripted iota, as it might be useful. As with all these composed accents, it can be made simple, then tweaked to get it just the right shape, or left as a composite and adjusted only for size and position.

<Composite> <!--Prosgegrammeni-->
<GlyphMapping>890</GlyphMapping>
<Member>
<GlyphMapping>953</GlyphMapping>
<XPos>-614</XPos>
<YPos>-469</YPos>
<XScale>0.45</XScale>
<YScale>0.37</YScale>
</Member>
</Composite>

The composite double accents can be composed, then converted to simple glyphs. If I understand you correctly, that should solve the problem with Fontlab, since you won't have any composites of composites, only regular composites of two simple glyphs.