Re: Proposed XMPP Extension: End-to-End Encrypted Contacts Metadata

Kevin Smith <[email protected]>
Newsgroups gmane.network.jabber.standards-jig
Message-ID <[email protected]>
On 15 Jan 2026, at 10:52, Daniel Gultsch <[email protected]> wrote:
> 
> Hi,
> 
> On Thu, Jan 15, 2026 at 11:37 AM Daniel Gultsch <[email protected]> wrote:
> 
>> Title: End-to-End Encrypted Contacts Metadata
>> Abstract:
>> This specification describes how to encrypt contacts metadata to
>> minimize server exposure.
>> 
>> URL: https://xmpp.org/extensions/inbox/contacts-e2e.html
> 
> honestly I’m not entirely sure I like this. My naive implementation
> would just stick a bunch of <item xmlns='jabber:iq:roster'
> jid='[email protected]'/> in some sort of wrapper element into on
> singleton roster node. No multiple items, not separate node for
> groups. Just re-use the same <group/> children from jabber:iq:roster.
> I reckon for most medium sized rosters the encryption overhead is the
> bottle neck instead of the bandwidths constrains of one items vs
> multiple items.

For ’normal’ rosters, I think the many-items thing is an irritation rather than helpful. For a stpeter roster, though, you could start running into issues with size limits trying to put everything into a single item, I think? Of course, if you *do* have a multi-thousand item roster, one item per roster item is also unfortunate, so I suspect neither of these approaches is truly ideal.

On a different note - I’m curious why, in a spec that’s concerned with reducing metadata leakage, it’s requiring that you advertise to to world that you support this. I suspect the disco is both redundant and counterproductive :)

/K
_______________________________________________
Standards mailing list -- [email protected]
To unsubscribe send an email to [email protected]
lmpx.com only provides a reader for public news (NNTP) servers. It is not affiliated with the servers or forums shown here and is not responsible for the content of articles, which is written by their respective authors.