- From: fantasai via GitHub <sysbot+gh@w3.org>
- Date: Thu, 01 Nov 2018 16:56:43 +0000
- To: public-svg-issues@w3.org
1. Humans, and computers, also have trouble with defining "user-perceived characters". Hence why we have the term “typographic character unit” in CSS3 Text: it has to mean different things in different contexts, so pegging to EGC isn't feasible. 2. If normalization is a concern, normalize the string first. The Web platform as a whole has given up on normalization for string matching, but it would probably be fine to do here for counting purposes. 3. Yes, ECGs probably give a better UX here, but stability across time is more important than stability across encodings. 4. Right. Which is why the UA would be allowed (or possibly required) to ignore values that are paired with a combining mark. 5. Right. Because they had some concerns about backwards compatibility. We shouldn't exacerbate the situation and give them more constraints they need to worry about next time they need to fix the grapheme cluster spec. 6. Agree that use counts would be nice to have. 7. The progressive improvements we make don't break pairing. They don't change, e.g. whether a CSS declaration is assigned to a particular div or its next sibling. Making this depend on grapheme clusters would create that kind of instability. I don't think it's a good idea. -- GitHub Notification of comment by fantasai Please view or discuss this issue at https://github.com/w3c/svgwg/issues/537#issuecomment-435108359 using your GitHub account
Received on Thursday, 1 November 2018 16:56:45 UTC