Conversation
…n MARC 008 * This works reasonably well with Microsoft Narrator for Latin-script languages * When there is original-script language this seems to work * Reading Romanized script is pretty bad no matter what; not sure what if anything could be done about that. If we decided this is worth adopting elsewhere, I'd want to add the html language tag in Solr so that we don't have to duplicate logic for mapping from the MARC language tag, but this at least gives us some idea.
|
Addresses #46, 1727173 |
|
I had VoiceOver read the whole page of this record, and found some little things. None of these are on the a11y audit report, so we don't really need to do anything about it, but I thought I'd note this one here since it's language-related. VoiceOver did a pretty good job with the languages, as far as I can tell. There's a decent variety on this record via the "similar items" column. The most annoying thing was the title was read in Romanian (I guess? I don't speak Romanian, so I can't verify that), but then it was read again but in the regular VoiceOver English voice trying to pronounce those words in English? Here's the accessibility tree: |
|
I started a list of other screen reader fixes over here |
|
I have some ideas about how to do this "right" for non-Roman writing systems, but for more universal appeal this approach is hard to beat. I mainly tested this with Japanese and VoiceOver. |
|
opened new issue for tracking this work #54 |
|
In discussing with @carylwyatt we should apply the language tag on the title even though it has some boilerplate as well -- based on advice from Deque that it's preferable to read the English content incorrectly than the other language. We could consider changing the order of the parts of the title so the title of the book is first followed by catalog record - hathitrust digital library (to be more like pageturner, and to better reflect the different content in the tab title) |
3 participants
If we decided this is worth adopting elsewhere, I'd want to add the html language tag in Solr so that we don't have to duplicate logic for mapping from the MARC language tag, but this at least gives us some idea.