Hello,

Thanks for the PR but we cannot merge code without understanding a problem.
Please provide a repro that triggers the assert and please explain the problem you are trying to solve.

Some fonts crash with pack_rect.was_packed set to 0.

Instead, in the case of pack_rect.was_packed 0, we fixed this problem by adding only advanceX of the glyph data without image data.

Some fonts crash with pack_rect.was_packed set to 0.

Please be specific and provide fonts/size/ranges information.

Some fonts crash with pack_rect.was_packed set to 0.

Please be specific and provide fonts/size/ranges information.

NanumSquareB/18/GlyphRangesKorean

Same information.
#5788 (comment)

Hello,
I scratched my head about this for a while, as there was something fishy.
Stepping through the packing procedure and comparing both variants (stb and freetype packers) I found the fix: 9150c23
It's a surprisingly innocent-looking fix but has very large effects on packing. I think the sentinel logic of rectpack may have been affected accidentally by this, but I am not sure. Anyhow, which this minor fix it seems like both #5788 and #5829 are correctly fixed.
Thanks for reporting this strange issue!

--stbrp_init_target(&pack_context, atlas->TexWidth, TEX_HEIGHT_MAX, pack_nodes.Data, pack_nodes.Size);
++stbrp_init_target(&pack_context, atlas->TexWidth - atlas->TexGlyphPadding, TEX_HEIGHT_MAX - atlas->TexGlyphPadding, pack_nodes.Data, pack_nodes.Size);