- From: Dan Connolly <connolly@w3.org>
- Date: Mon, 21 Mar 2005 23:49:15 -0600
- To: www-tag@w3.org
On Mon, 2005-03-21 at 23:33 -0600, Dan Connolly wrote:
> I just discovered...
>
> <definitions name="GoogleSearch" targetNamespace="urn:GoogleSearch">
> -- http://api.google.com/GoogleSearch.wsdl
>
> urn:GoogleSearch isn't registered.
> http://www.iana.org/assignments/urn-namespaces
>
> I'm pretty sure google would be embarrassed to use
> an HTTP URI with an unregistered DNS domain... why aren't
> they embarrassed to use an unregistered URN namespace identifier?
>
> I think the fact that the lookup mechanisms for DNS names,
> and hence http/ftp URIs, are ubiquitous creates a useful
> pressure to not use bogus names. The pressure doesn't
> seem to be there for URN namespaces, and the only way I can
> see that pressure increasing is if URNs duplicate the
> functionality of http... or at least if URNs get connected
> to *some* communications protocol that causes bogus names
> to get noticed routinely.
Hmm... the anti-pattern seems to apply to URI schemes too...
<operation name="getWind">
<soap:operation
soapAction="capeconnect:AirportWeather:com.capeclear.weatherstation.Station#getWind"/>
--
http://www.capescience.com/webservices/airportweather/AirportWeather.wsdl
Not to mention the over-use of POST...
print AirportWeather.getTemperature("KLAX")
-- http://aspn.activestate.com/ASPN/WebServices/SWSAPI/pytut
--
Dan Connolly, W3C http://www.w3.org/People/Connolly/
D3C2 887B 0F92 6005 C541 0875 0F91 96DE 6E52 C29E
Received on Tuesday, 22 March 2005 05:49:16 UTC