- From: Ivan Herman via GitHub <sysbot+gh@w3.org>
- Date: Sun, 03 Apr 2016 10:20:35 +0000
- To: public-annotation@w3.org
iherman has just created a new issue for https://github.com/w3c/web-annotation: == Selector/media table == I have put up an [initial table](http://w3c.github.io/web-annotation/model/medias.html) for media types and selectors (see PR #202). This issue is a spin-off of for issue #165. I also added a second table, where some rows that are unclear, eg, CSS, JSON, and Javascript (obviously, any other programming languages could play the same role). Indeed, the current selector structure is fairly poor in terms of these, although annotating a, say, program, is a clear use case. The text position selector for these is also poorly defined (probably it should be considered in a 'minimized' version only) Some questions/issues * [ ] The reference to official media types may not be always appropriate. What is the media type of a binary data file? Maybe the entry should be 'anything that is not covered above'? * [ ] I do *not* propose to extend the current selectors to encompass CSS/JS/JSON; I do not think there is enough experience for these. However, we should add section in the vocab document, similarly to the extension section on motivation, on the possibility to extend selectors, and also add a reference to that extension possibility in the model document to make things more explicit. We could then use the CSS, JSON, and Javascript as example. A, say, Javascript selector *may* include keys like 'function' or 'variable', meaning that the selector could specify the definition of a function or the creation of a variable; combined with, say, the selector for text position or range it could provide a good way to select a relevant part of a program. Something similar could be done for CSS or for JSON. * [ ] There may be a need to give a more precise specification for some of the selectors for specific media types (if we accept them for that media type). For example, if a text position (or text quote) selector is "valid" for CSV, should the position number include the separator (i.e., ",") character? * [ ] There are some questions marks where I was not sure... * [ ] Some entries are such that there is no formal definition for a fragment, but it is relatively clear what they mean. (E.g., in Turtle, the RDF identification of a term/URI prevails, TSV would use the same as CSV). We may decide to be silent on this, but flagging the issue nevertheless Cc @azaroth42 @paolociccarese @BigBlueHat Please view or discuss this issue at https://github.com/w3c/web-annotation/issues/203 using your GitHub account
Received on Sunday, 3 April 2016 10:20:37 UTC