Shervin Afshar

1. Adding some basic definition about these term would help us in other parts of the document. 2. Explain the recommended behaviour based on Unicode specifications.

I reviewed the draft for this section and have few comments: - Naskh being "the standard script for the Arabic and Muslim world" is highly debatable; - Kufi was not...

@ntounsi, while this is a valid case with the hyperlink breaking, but the question here is more about which method should be used to correct the gap caused by either...

@asmusf, was a mistake. Corrected that. ~~could you please be more specific? I don't follow.~~

ZWJ appears immediately after character from the script subject to joining behavior to force initial/medial form from an standalone character.

Ideally we should have a paragraph introducing the gist of math layout topics for Arabic and refer to Arabic MathML document.

We can briefly talk about the issue in the section on numerals, but the discussion of how this issue can be circumvented in the real world is, as Mostafa mentioned,...

If those features mentioned above have common wider usage (different colors in one joining run vs. horizontal parallel translation) they can be considered general use-cases.

See [this comment](https://github.com/w3c/alreq/issues/128#issuecomment-363494322) on #128.

As mentioned in [Issue#101](https://github.com/w3c/alreq/issues/101#issuecomment-1028216938), CLDR shows the distinction between Arabic and Persian: - `ar`: https://st.unicode.org/cldr-apps/v#/ar/Weather/33c368f9174f3ad  - `fa`: https://st.unicode.org/cldr-apps/v#/fa/Weather/2bb26ab5dcc064f6