27326 2014-11-14 17:33:45 +0000 Supporting text/plain fragment identifiers 2014-11-15 09:56:42 +0000 1 1 1 Unclassified TextTracks CG WebVTT unspecified PC All RESOLVED DUPLICATE 22562 P2 normal --- 1 dret dave.null cyril.concolato philipj silviapfeiffer1 public-texttracks oldest_to_newest 114939 0 dret 2014-11-14 17:33:45 +0000 What about supporting text/plain fragment identifiers? Then WebVTT users could point into a WebVTT document and say "there's a typo on line 42": http://tools.ietf.org/html/rfc5147 114954 1 philipj 2014-11-14 23:30:47 +0000 Are there mainstream editors/browsers/something that will generate such a URL from the cursor position? If not, it seems like far less work to say "there's a typo itendifier" than generating a URL, especially since the receiver would also need software that understands the scheme, or be able to read it manually. 114960 2 silviapfeiffer1 2014-11-15 03:54:11 +0000 The fragment identifier scheme would need to be defined for text/vtt specifically. Positions, characters and lines aren't so relevant to VTT. Since VTT has a structure, having references to cues via their identifiers and to time ranges (i.e. subsets of cues) makes a lot more sense. Something more in line with media fragment identifiers: http://www.w3.org/TR/media-frags/ . As for use cases: there's also a discussion in https://www.w3.org/Bugs/Public/show_bug.cgi?id=22562 - I think we should take it back there. *** This bug has been marked as a duplicate of bug 22562 *** 114964 3 dret 2014-11-15 09:56:42 +0000 good point about VTT being structured and thus using that structure making more sense than just using text/plain fragment semantics. thanks for taking the time to respond to this!