Hello everyone,
could you please help me? I’ve designed a font, and when it was used on a website, I found out that some of diacritical marks are far away from where they’re supposed to be. So it looks like y ´instead of ý and i´(with the tittle) instead of í. The problem appears randomly, sometimes it’s alright, sometimes it’s not in the same letter. In some articles it’s bad across the whole text, in others it’s totally fine. Why is this happening? I’m not familiar with web technologies, but it works fine when I test it myself. Seems like it has something to do with anchors, or maybe it depends on how the copywriter types the copy…?
0
Comments
What you see will then depend upon the rendering environment.
Some environments may attempt to redirect to a precomposed codepoint, all on their own initiative, following Unicode composition/normalization rules, if the codepoint is present in the font’s cmap. InDesign does this, for instance. But most web browsers do not.
If the renderer is not directed to a precomposed codepoint and it does not find a {mark} feature to handle the positioning of the combining accent, then it may handle the sequence in a couple different ways.
In some situations, you’ll find that the renderer acknowledges that the accent is intended to be a combining accent (based upon its codepoint) and will simply center the combining accent over the preceding glyph, in the absence of any other direction.
In other cases, however, the renderer will just present the zero-width combining accent at the right sidebearing of the preceding glyph, just as you would expect from a simple sequence of two codepoints.
If your encoded combining accents are not zero-width (like they’re supposed to be), then the effect of this latter handling will be exacerbated accordingly.
A bit off-topic, but is there some kind of list of must-have features that every good functional font supposed to have?
The preferred solution would be a {mark} feature.
That, or whoever is composing the text for the web should probably use precomposed codepoints (except for those combinations that do not have Unicode codepoints, of course). IMHO.
Do I understand it correctly that in this case a font should include both {mark} feature and precomposed glyphs for different environments? Or any other reasons to use precomposed glyphs at all?
Here’s one for you: http://typedrawers.com/discussion/2282/does-anybody-use-combining-accents