- From: Philippe Le Hegaret <plh@w3.org>
- Date: Fri, 25 Apr 2014 10:20:57 -0400
- To: Anne van Kesteren <annevk@annevk.nl>
- Cc: Arvind Jain <arvind@google.com>, Jonas Sicking <jonas@sicking.cc>, "public-web-perf@w3.org" <public-web-perf@w3.org>
On Fri, 2014-04-25 at 15:49 +0200, Anne van Kesteren wrote: > On Fri, Apr 25, 2014 at 3:35 PM, Philippe Le Hegaret <plh@w3.org> wrote: > > On Thu, 2014-02-13 at 18:25 +0000, Anne van Kesteren wrote: > >> 2f) You don't deal with closed blobs (unclear yet how that should be done). > > > > The size attribute for the Blob would be 0 [1] after a close so I'd > > expect the raw data to be empty for a closed Blob. Looking at the send > > method for XHR or the Blob constructor itself, they don't seem to do > > anything special in those cases, so they'd expect the same thing as well > > I presume. A note could be added to step 6 in Beacon if needed but it's > > not clear to me why we would need to do so in Beacon, while this is not > > done in XHR or the Blob constructor definition. > > > > [1] http://www.w3.org/TR/FileAPI/#dfn-close > > As I said, it's not clear yet how it should be done, but this is most > definitely wrong in XMLHttpRequest. Isn't the problem in the File API itself then? It's missing a definition for "raw data", ie if size = 0, raw data is empty (or undefined?). Looking further, I note that Blob.slice doesn't return an error when invoked on a closed Blob either. It returns an empty Blob instead. So, I still believe that sending a closed Blob on Beacon should succeed with no data being sent instead of returning false. Philippe
Received on Friday, 25 April 2014 14:21:09 UTC