W3C home > Mailing lists > Public > public-annotation@w3.org > April 2016

Re: [web-annotation] Model should allow multiple Selectors per SpecificResource

From: Ivan Herman via GitHub <sysbot+gh@w3.org>
Date: Sat, 30 Apr 2016 14:03:28 +0000
To: public-annotation@w3.org
Message-ID: <issue_comment.created-215965681-1462024996-sysbot+gh@w3.org>
@azaroth42:

>>But... doesn't that leads to interoperability issues. You pick, 
randomly, one and I pick the other? What is the use case for this?
>
> The idea is that the more you specify, the more likely it is that 
one will continue to work and that clever clients can use the set to 
be even more precise. For example, if you use just TextPosition, you 
have no way to tell that the document changed, but it's as precise a 
way as possible when the document is static. So if there's also a 
TextQuote selector, then you can use that to either find the selection
 (even though it's maybe slower or harder), or to verify that the 
content at the specified position is the same.
> This is towards robustness of anchors, but clearly not the entire 
solution.
>
> Second use case is if one system implements only TextQuote, then you
 can use the annotation, whereas if you supplied just a complex 
RangeSelector of XPathSelectors, that might be more accurate and 
faster ... but worthless to the simpler implementation of just the 
Quote pattern. Related to this is if there are new, super awesome 
selectors that are developed in different communities, it would be 
good to at least provide one of the selectors from the model for 
systems outside that community.

Ok, these are all good issues. I believe the underlying thought is 
that if there are two selectors (or a state and a selector, reflecting
 to issue #205? Or is that a different issue after all?), they 
___SHOULD (MUST?) select the same content___.

My problem is how to spec this. Is this a testable statement for an 
implementation? How would an implementation be sure about this? How to
 ensure interoperability? Or should we just stop by saying 'MUST' and 
leave the rest to implementations?


-- 
GitHub Notification of comment by iherman
Please view or discuss this issue at 
https://github.com/w3c/web-annotation/issues/207#issuecomment-215965681
 using your GitHub account
Received on Saturday, 30 April 2016 14:03:30 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 18:54:45 UTC