- From: Marcos Caceres <marcosscaceres@gmail.com>
- Date: Fri, 23 May 2008 09:16:01 +1000
- To: "Jon Ferraiolo" <jferrai@us.ibm.com>
- Cc: public-appformats@w3.org
On Fri, May 23, 2008 at 5:19 AM, Jon Ferraiolo <jferrai@us.ibm.com> wrote: > Is the main benefit of the 'widget:' protocol so that widget authors can > have the convenient option of using "/" based relative addressing (e.g., > /images/image1.jpg) instead of having to use ".." relative addressing (e.g., > ../../../images/image1.jpg)? The purpose is to stop implementations from using 'file://', and hence, addressing anything on the file system. In the config doc, authors are still free to use either relative or absolute references (but not full URIs or the widget protocol) to address resources inside a package. Using the widget protocol inside a start file will have no effect (ie. authors cannot address resources inside another widget). > If that's the main benefit, I'm not sure that the benefit/cost ratio is > large enough. Isn't the proposal of a new protocol going to take a lot of > time to work out the technical details and march through the standards > processes, such as coordination with the URL community? For example, perhaps > someone will say the protocol should be called "zip:" instead of "widget:" > to make it general to all ZIP-based formats instead of just the one use of > ZIP packaging by the Widgets spec. (But perhaps I'm being too timid about > the standards difficulties.) I like your "zip://" proposal:) The scheme is very simple, so hopefully it won't be much of a problem getting it standardized. I'll draft up a formal proposal today and CC the TAG to see what they say. Kind regards, Marcos -- Marcos Caceres http://datadriven.com.au
Received on Thursday, 22 May 2008 23:16:45 UTC