[Standards] XMPP URI usage in "HTTP over XMPP"

Peter Waher Peter.Waher at clayster.com
Wed Jul 10 14:21:56 UTC 2013


Hello Peter 

Thanks for your comments. I'll respond to each in turn:

>> Hello Ralph
>> 
>> Thanks for your mail. You're absolutely correct. It was sloppy of me 
>> to propose a URI scheme with the same name as the previous xmpp 
>> scheme. I'm sorry.
>> 
>> I've therefore renamed the scheme to httpx (§4.3.1.1 ff). This URI 
>> scheme is through to work seamlessly with existing applications, 
>> without requiring changing the web applications themselves (except if 
>> full URL's including schemes are used inside the application).
> 
> Defining a new URI scheme is not an effort to be taken lightly. Please see RFC 4395, in particular Section 2.1:
>
>    The use and deployment of new URI schemes in the Internet
>    infrastructure is costly; some parts of URI processing may be
>    scheme-dependent, and deployed software already processes URIs of
>    well-known schemes.  Introducing a new URI scheme may require
>    additional software, not only for client software and user agents but
>    also in additional parts of the network infrastructure (gateways,
>    proxies, caches) [11].  URI schemes constitute a single, global
>    namespace; it is desirable to avoid contention over use of short,
>    mnemonic scheme names.  For these reasons, the unbounded registration
>    of new schemes is harmful.  New URI schemes SHOULD have clear utility
>    to the broad Internet community, beyond that available with already
>    registered URI schemes.
>
> Does your proposal pass that test?

Yes, definitely. It is basically two use cases which is (will be) of clear interest to the broad internet community:

* Extending the semantic web onto peer-to-peer XMPP-based networks. I'll write a paper on this shortly, to better describe the use here. However, since the semantic web is based on URL's data needs to be addressed by URLs. Data hosted on xmpp devices therefore need a unique URI scheme to identify how to get the data (if data points to a semantic data document).

* Hosting web applications in private environments using XMPP (for example using Apache on a plug computer at home).

Note however, that I only propose the URI scheme to be registered as a provisional URI scheme, not a permanent one, as per BCP 35. This allows for modifications during the experimental phase. When (if perhaps some would say) the XEP is lifted to draft, the URI scheme registration could be registered as a permanent URI scheme.

>> I've also added an URI Scheme Registration Template to the document 
>> (§8.1), as per BCP 35, that we can mail to IANA to register the new 
>> provisional URI scheme,
>
> I appreciate the fact that you appear to be saying that the XMPP Registrar would coordinate with the IANA here, rather than having you contact the IANA directly. These inter-SDO things are best handled that way, and this is also XSF policy:
>
> http://xmpp.org/extensions/xep-0053.html#iana

I would be happy to do this. I didn't want to presume that I could do it in the name of the XSF. I still would like XSF approval of the idea. However, I can make sure to send the registration and answer all comments/suggestions/critique.

>> when the XSF has approved the proposal and moved it to experimental.
>
> The XMPP Registrar will most definitely *not* register a new URI scheme with the IANA when a XEP is merely published as Experimental. Instead, the Registrar will do so only when the XEP is advanced to Draft in the XSF's standards process, so that we can make sure the registration reflects community consensus. This too is consistent with XSF policy:
>
>http://xmpp.org/extensions/xep-0053.html#registries

Note that I suggest we register this as a provisional URI scheme, not a permanent one. As I read BCP 35, this is how they suggest it should be done. I also interpret BCP 35 as saying they recommend (and invite) people to register provisional URI schemes this way. They explicitly state they "encourage registration".  I see nothing wrong with this. It registers the URI scheme name httpx. When the XEP moves to draft later, and there are multiple implementations in open frameworks, the URI scheme can be re-registered as a permanent URI scheme.

Have I understood something wrong?

Sincerely,
Peter Waher



More information about the Standards mailing list