10 Specific Issues and Their Solutions
Intermediate Persian – H4, Caption, TH Farsi Text Rendering Improperly

Farsi text (specifically in H4, caption, and TH (table header cells) in Intermediate Persian was not being rendered consistently with the rest of the book.


Mary tried to fix this by targeting Farsi tagged text in H4 and forcibly setting the font-family to “Encode sans” and the font-weight to 700 with “!important” to no avail. At the bottom of the two above images, the “Rendered Fonts” are different. This stays the same for the H4 text even when the computed font-family is “Encode Sans”. The font-weight will not change from 900, though, even with “!important”.
It seems the desired font rendering (Segoe UI) is not applicable to 900 weight Farsi text, and this is the root of the issue, which does not only apply to header text, but any text with weight set to 800 or more.
Further investigation into how to fix this:

The two “bolder” properties on top of the 600 (700?) weight makes the final weight 900. This overrides the font-weight setting in custom styles (even when !important is used). So, to my (Mary’s) knowledge, the only way to fix this is by setting the font-weight with inline CSS or by changing the structure of the HTML.

Since the font-weight style cannot be changed globally in external CSS, and this is a relatively sparse issue in this book, I am manually setting the HTML structure to the structure below, which fixes this issue:

This is just un-bolding the Farsi text. This also applies to table header cells where Farsi text is inside <strong> tags (“Bolded” with the “B” button in the tool bar). To fix this, simply un-bold the Farsi text.
TLDR: When using the font “Encode Sans” or similar fonts rendered by “Segoe UI”, do not bold Farsi text inside H4, captions, or table header cells, or it will not render consistently with the rest of the book. That is, if Farsi text is not being rendered properly, try to un-bold it and it will (probably) fix the issue. This is likely also applicable to Arabic/Urdu.