[mdx] Joe on section 3.2.1

[Cantor, Scott]

On 9/30/13 4:31 PM, "Ian Young" <ian at iay.org.uk> wrote:
>It may actually be version-dependent, see:
>
>http://bit.ly/16aTquI
>
>I'm not sure that means we can assume it's OK, though.

That suggests not to me, which is unfortunate, since embedding a set of
data into a query string parameter ends up creating the same confusion as
using the path.

What concerns about using pathinfo is that the act of URL encoding
characters in an entityID doesn't change their meaning. A slash is still a
slash, so embedding an entityID in most cases is going to result in a
fairly weird pathinfo value given that the embedded slashes are still
being interpreted that way.

Obviously pathinfo itself is not meant to be interpreted by anything but
the resource the web server hands it to, but I'm still uncomfortable with
that kind of data being placed into it.

So I'm still wondering if defining a query string protocol isn't better
here, even if the spec text ends up largely equivalent work.

Also, it seems to me that the implication of using '+' as the separator
(which I take here to mean an unencoded value), means that we need text
precluding the present of that character in the actual IDs. There's no
escaping mentioned, so there's no way to have a '+' inside an ID.

-- Scott
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 5778 bytes
Desc: not available
URL: <http://lists.iay.org.uk/pipermail/mdx-iay.org.uk/attachments/20130930/aa9d0780/attachment.bin>