- From: Markus Lanthaler <markus.lanthaler@gmx.net>
- Date: Tue, 3 Sep 2013 20:33:26 +0200
- To: "'Linked JSON'" <public-linked-json@w3.org>
On Tuesday, September 03, 2013 7:38 PM, Gregg Kellogg wrote: > It would have been useful to voice the opinion more forcefully when we > had the discussion on the last call. Perhaps, yeah. > So, you're suggesting now that we > remove the useDocumentLoader test option and not do any tests > specifically invoking a runtime defined documentLoader? The thing is that we'll have only one compliant implementation for this. It doesn't make much sense to implement another processor in JavaScript at the moment and I don't expect browser vendors to implement it natively anytime soon (at least not before framing and perhaps some other stuff is added). I don't particularly care about the form we use to create those tests (manifest or idltest) but we should keep the consequences in mind. That's the reason why I pushed to include this in the idltests. It's not unreasonable to put it there because it's clearly an API thing that most implementations won't support (in the specified form at least). Does this make sense? -- Markus Lanthaler @markuslanthaler
Received on Tuesday, 3 September 2013 18:34:01 UTC