On Thu, 12 May 2011, yael.aha...@nokia.com wrote: > > I'd like to propose a few changes to the definition of the dropzone > attribute. > > 1. In the same sentence, the spec says that it is not valid to add the > feedback value more than once, but it also says that the first > occurrence is to be used. > > Can the spec change to simply say that the first occurrence of the > feedback value will be used (and it is valid) ?
I don't understand the question. Can you elaborate? > 2. Using "f:" and "s:" for distinguishing strings from files seems > needlessly confusing. Why not use "string:" and "file:" ? My concern was that it would be needlessly verbose, but I guess it's pretty rare to use this so it would be ok. I've changed the spec to use the longer prefixes. > 3. It seems very useful to allow wildcards in the data type, so instead > of defining multiple types for images, "image/*" should be allowed. Well you'll have to pick a specific type when you actually receive the drop, so it seems simpler to just list the supported types. Otherwise you might find that actually you don't support any of the image/* types that were dragged in. -- Ian Hickson U+1047E )\._.,--....,'``. fL http://ln.hixie.ch/ U+263A /, _.. \ _\ ;`._ ,. Things that are impossible just take longer. `._.-(,_..'--(,_..'`-.;.'