- From: Ryan Sleevi <sleevi@google.com>
- Date: Fri, 21 Sep 2012 10:35:57 -0700
- To: David Wagner <daw@cs.berkeley.edu>
- Cc: public-webcrypto-comments@w3.org
On Thu, Sep 20, 2012 at 3:22 PM, David Wagner <daw@cs.berkeley.edu> wrote: > When the user selects AES-CBC encryption, the specification should > probably specify what kind of padding scheme will be used to pad out > the data to a multiple of the block length. > > I don't think this necessarily needs to be user-selectable. However, > I do think the specification should say what kind of padding scheme will > be used, so that it is possible to build other implementations that can > interoperate with this JavaScript API (without needing to reverse-engineer > the WebCryptoAPI implementation). Does this sound right to others? > > Agreed, and filed as https://www.w3.org/Bugs/Public/show_bug.cgi?id=18953 Most of the algorithms are under-specified at this point with regards to particular behaviours, let alone including references or conformance tests (both of which are necessary in advancing to CR), as we solicit feedback regarding the overall design choices of the API and the algorithms exposed. However, the bug will help track this particular concern, so thanks for raising it.
Received on Friday, 21 September 2012 17:36:29 UTC