- From: r12a via GitHub <sysbot+gh@w3.org>
- Date: Thu, 12 Nov 2015 17:18:12 +0000
- To: www-international@w3.org
ok, I'll try to summarise what we just discussed in the i18n telecon. the issue here is that some ideographic characters may look foreign to the reader unless a language-sensitive font is applied. (btw, this can be the case for other scripts too, including arabic and latin) we are only talking about minor differences between characters, and they are more likely to affect readability rather than meaning. When significant differences arise, such as for 门 and 門, which have a different number of strokes, then Unicode uses separate code points, including (as for that example) to reflect the differences between Simplified and Traditional forms of the same character. Nevertheless, it *is* a real issue, and it is important that it be addressed. there needs to be some information associated with the content in question that describes what language it is in. If information is taken, say, from an HTML page, the `lang` attribute should provide that information. in order for the appropriate font to be applied by the receiver at display time, it will be necessary to pass information with JSON strings to indicate the language of the content. There needs to be a mechanism to allow this. It should use BCP47 language tags. on the receiving end, the application that handles the display of the strings needs to be able to associate a language-appropriate font with the string so that Chinese text uses Chinese font forms, and Japanese text uses Japanese font forms, etc. All the major desktop browsers already do that (fwiw see the test results at http://www.w3.org/International/tests/repo/results/default-font – click on the text in the 'Test link' column to try the test). HTTP's accept-language addresses a different problem (language negotiation), and so is not likely to be useful here. Basically, it is used to request a specific language variant – eg. i want the content in Chinese or in Japanese, etc. If the content is available in that language, it can be sent to the receiver, but the problem of choosing the correct font for display still requires a knowledge of which language the content is in (which is the problem we are concerned with). does that help? -- GitHub Notif of comment by r12a See https://github.com/w3c/presentation-api/issues/218#issuecomment-156171699
Received on Thursday, 12 November 2015 17:18:14 UTC