Re: Relaxing Zope id rules for Zope 4 / Plone 5

On Wed, Feb 20, 2013 at 7:20 AM, Hans-Peter Locher
<[email protected]> wrote:
> I think object ids should still
> be strict enough, so they are url compliant
>
> ( rfc may be http://www.ietf.org/rfc/rfc1738.txt )

Is there a specific critique here of Wichert's proposal?  I read
http://www.ietf.org/rfc/rfc3986.txt section 2.5 updating the URI
specification to suggest using UTF-8 on all unreserved characters, and
only percent-encoding reserved characters defined as:

      reserved    = gen-delims / sub-delims

      gen-delims  = ":" / "/" / "?" / "#" / "[" / "]" / "@"

      sub-delims  = "!" / "$" / "&" / "'" / "(" / ")"
                  / "*" / "+" / "," / ";" / "="

Better questions we should ask:

(1) Does urllib/urllib2 support this encoding, assuming all browsers do?

(2) Are there known clients that will choke on utf-8 URLs vs.
known-to-work percent encoding?

(3) If we must have percent encoding, why do ids have to internally
use this?  I would not think that a requirement.

Sean

------------------------------------------------------------------------------
Everyone hates slow websites. So do we.
Make your web apps faster with AppDynamics
Download AppDynamics Lite for free today:
http://p.sf.net/sfu/appdyn_d2d_feb