- From: r12a <notifications@github.com>
- Date: Fri, 24 Jul 2020 02:53:53 -0700
- To: w3c/manifest <manifest@noreply.github.com>
- Cc: Subscribed <subscribed@noreply.github.com>
- Message-ID: <w3c/manifest/pull/926/c663461391@github.com>
hi @marcoscaceres . I think there are a couple of things to discuss. 1. I think you may not need to get to the level of specificity you do wrt the points in the bidi algorithm (UBA), especially since this spec iiuc carries metadata rather than actually implementing bidi (correct me if i'm wrong, i haven't reviewed the spec deeply myself - Fuqiao did that). I think, at the moment, the steps you're pointing to in the bidi spec are not actually the right ones. Maybe it's enough to say simply that if the value is set to `ltr` or `rtl`, the consumer will need to ensure that the base direction is set to one of those, respectively, when the text is displayed for a user. For `auto` it may be enough to say that the base direction will be set according to the first strong directional character. Then mention that all this is as described in the UBA. 2. The text you currently have seems to convey the wrong idea about how `auto` works. It suggested to me that the direction of the string to which it is attached is determined by looking at the direction of the surrounding text (ie. like an inherited value). That's not actually the case. The auto value should force the consuming application to check the first strong directional character in the text to which the value is applied, and determine the base direction from that. So it looks inwards rather than outwards for cues. Does that help? -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/w3c/manifest/pull/926#issuecomment-663461391
Received on Friday, 24 July 2020 09:54:07 UTC