Text rendering of Unicode text in LC

[Peter Bogdanoff]

To follow up with my question from yesterday…and to continue with Biblical allusions others have begun…

I’ve descended into <font> hell, which I’m trying get out with some light from LiveCode Mark!

To recap, the rendering of Chinese characters in Windows 10 is willy-nilly through different versions of LC. Earlier versions display bitmap glyphs, later versions display drawn characters. Now in LC 9.0 it displays a MIX of apparent bold and plain of characters all set identically.

Examples:
https://spark.adobe.com/page/clCst4mdUu8Jy/ <https://spark.adobe.com/page/clCst4mdUu8Jy/> 

What’s going on here? Bugs?

My personal situation—this is a multi-year project that includes one of the two largest academic book publishers in China who are looking to our project as the vanguard of digital publishing in their country. They are looking at this text display and saying, “we can’t publish this.” Now, they want to show it at a publishing conference...

The Chinese translators say the bitmap text is great, but the anti-aliased rendered text is unusable. And everyone there has Windows. We with roman-text eyes don’t necessarily see a problem; they with Chinese character eyes are very particular.

I’m supposing this is Skia doing the work (with the harfbuzz text shaping library?)?? If true, I see from the Skin/harfbuzz documentation on Font Embedded Bitmaps:

"bool isEmbeddedBitmapText() const
Returns true if Font Engine may return Glyphs from font bitmaps instead of from outlines.”

https://skia.org/user/api/SkPaint_Reference#Subpixel_Text


Can there be a font property option to display just the bitmap? Or are there other fixes that can be made? 

Peter Bogdanoff
ArtsInteractive

> On Aug 12, 2018, at 4:15 PM, Peter Bogdanoff via use-livecode <use-livecode at lists.runrev.com> wrote:
> 
> Songti SC Regular