W3C home > Mailing lists > Public > public-webapps@w3.org > October to December 2008

Re: [Widgets] URI Scheme revisited.... again

From: timeless <timeless@gmail.com>
Date: Mon, 13 Oct 2008 18:01:24 +0100
Message-ID: <26b395e60810131001q502c4177j4fca76d9695db143@mail.gmail.com>
To: "Marcos Caceres" <marcosscaceres@gmail.com>
Cc: "Mark Baker" <distobj@acm.org>, public-webapps <public-webapps@w3.org>

On Mon, Oct 13, 2008 at 3:59 PM, Mark Baker <distobj@acm.org> wrote:
> file:, despite the name, doesn't have to be mapped to the file system.
>  Its scope could be limited in exactly the same way you've limited
> widget: there.  Similarly, ftp or http - even part of the space -
> *could* be mapped to the file system.  So the issue you're worried
> about has little to do with the URI scheme.

On Mon, Oct 13, 2008 at 4:21 PM, Marcos Caceres
<marcosscaceres@gmail.com> wrote:
> That's absolutely true.

actually, it isn't.

file has a host field and a well defined if utterly ignored meaning.

it basically means that if i am not the host listed, then i'm supposed to error.

file://foopy/tmp should only work if i'm foopy, or something strange like that.

you can't simply hijack the host field from file, it's both reserved
and has hacks in the most likely user agents to handle cases where
people improperly used it (resulting in either file:/// or
file://///).

> It could be that, for instance, we recommend
> "file://widgetEngine/widget.wgt/path/to/file" or just
> "file:///widget.wgt/path/to/file". But we are still stuck on the fact
> that file: hasn't been formally standardized anywhere. Does that
> matter?

and the fact that file has behaviors which are not going to be
friendly to such abuse.
Received on Monday, 13 October 2008 17:02:05 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 18:49:28 GMT