[Standards] UPDATED: XEP-0332 (HTTP over XMPP transport)

Peter Waher Peter.Waher at clayster.com
Thu Aug 21 14:54:52 UTC 2014

Hello Matthew

Thanks for your input. My responses to each one of your comments, below:

>Two small comments:
>  1) I see the examples have been fixed for streamid -> streamId.
>However it's worth noting that we've always conventionally used lowercase attributes and element names in XMPP.

I've noted a mix of syntaxes in different XEPs. Is there some kind of syntax convention extensions should adhere to? I know this has been discussed somewhat earlier. Haven't seen one, however. If the XSF wants to streamline syntax in new XEPs, it might be a good idea to start some kind of discussion regarding this, and include some guidelines in XEP-0001.

Personally, I've tried to use camel casing for attributes in all XEP's I've written myself. The main reason for this, is not style as such, even though it is easy enough to read, but to avoid the hyphen sign (-), which is often used as an alternative to separate words where lower-case-only names are used. The hyphen sign does not map well into other scripting languages, etc., and creates a lot of extra work, if you want to do more with the XML and schema, than simply use it for documentation, streaming and validation. (Other examples might be transformation between versions or formats, automatic code generation, searching, mapping in script, etc.)

>  2) Servers that don't route messages in order are violating http://xmpp.org/rfcs/rfc6120.html#rules-order - though we've ended up with similar attributes in XEP-0047 and XEP-0301, so there is some precedent. Also both these XEPs call the attribute 'seq', another opportunity for consistency.

Ok. Is this important to change? I would prefer not to change this, since there are various distributed solutions already in the field, and I would like to avoid making breaking changes, unless they improve or add functionality.

But the comment still raises a valid point: Should the XSF maintain a list of (commonly) used attributes and attribute names to promote reuse of the same?

Best regards,
Peter Waher

More information about the Standards mailing list