Hello Peter Thanks for the links. We are aware of those attempts, and the IEEE working group had also unearthed them.
However, the proposed solution is not sufficient. For EXI to be successfully implemented, there needs to be a handshake where both parties agrees on a set of parameters and a set of schemas to use. Also, an interchange of schemas between parties would be very good, so introduction of new schemas could be done without having to update software. Even though EXI can work in a schema-less mode, it's compression ability is best when both parties have access to the same XML schema. I've on my list to create such a proposal and mail to this list. Sincerely, Peter Waher -----Original Message----- From: Peter Saint-Andre [mailto:[email protected]] Sent: den 11 mars 2013 19:17 To: XMPP Standards Cc: Yusuke DOI; Peter Waher Subject: Re: [Standards] Question on valid XMPP -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Please be aware that we had discussions about this four years ago: http://xmpp.org/extensions/inbox/compress-exi.html http://mail.jabber.org/pipermail/standards/2008-February/017938.html (and follow-up messages in that thread) http://mail.jabber.org/pipermail/standards/2008-March/018185.html See also more recent discussion: http://mail.jabber.org/pipermail/members/2012-March/006712.html (and follow-up messages in that thread) On 3/7/13 7:40 AM, Yusuke DOI wrote: > Dear Peter, > > Yes, for sensor networks such as one on 802.15.4, XMPP/EXI should be a > good choice (if we can implement it efficiently). Please let me join > your effort. Thanks! > > // Yusuke DOI <[email protected]> Corporate R&D Center, TOSHIBA > Corp. > > > (2013-03-07 20:55), Peter Waher wrote: >> Hello Yusuke We have been experimenting with EXI and find it an >> exceptional and efficient way to compress XML. We want to use it in >> sensor networks, where RAM and packet size is an issue. We hope to, >> within our effort to create XEPs for sensor networks, to include a >> XEP for EXI enablement of XMPP communication. If you want to join us >> in this work, you're welcome. Note to XSF >> members: Anybody interested in participating in such an EXI XEP, >> please contact me. Sincerely, Peter Waher >> >> Från: Yusuke DOI >> <[email protected]<mailto:[email protected]>> >> Datum: 6 mars 2013 07:32:55 CET Till: XMPP Standards >> <[email protected]<mailto:[email protected]>> Ämne: [Standards] >> Question on valid XMPP Svara till: XMPP Standards >> <[email protected]<mailto:[email protected]>> Hi, >> >> My name is Yusuke Doi. I'm wondering if it is possible to put EXI[1] >> with XMPP. >> >> During my experiment to encode some XML instances captured from my >> tcpdump to EXI, I have got unique particle attribution (UPA) problem >> as described in [2]. Although EXI works with schemaless XML, it's far >> better to use schema-informed EXI in terms of efficiency and >> validation. Schema-informed EXI is for valid XML, but current spec >> does not allow validation due to UPA. Changing local schema >> (suggested by Peter in [2]) breaks interoperability of >> schema-informed EXI. >> >> Is there any possible way to make a 'valid XMPP spec' with some >> (backword-compatible) specification update? I guess there are three >> choices. >> >> 1) change XMPP schema/spec to avoid UPA. This may break >> inteoperability. 2) use different XMPP schema for EXI communications. >> This looks awkward. 3) use XSD 1.1 for weak wildcard. I'm not sure >> XMPP community wants/hates 'cool' features of XSD 1.1. >> >> [1] http://www.w3.org/TR/exi/ [2] >> http://mail.jabber.org/pipermail/jdev/2012-June/089069.html >> >> Regards, >> >> // Yusuke DOI >> <[email protected]<mailto:[email protected]>> >> Corporate R&D Center, TOSHIBA Corp. >> >> >> >> -- Hälsningar >> >> Joachim Lindborg Teknisk Chef >> >> Sustainable Innovation AB Adress: Box 55998 102 >> 16<tel:55998%20102%2016> Stockholm Besöksadress: Storgatan 31 >> (Malmgården) Email: >> [email protected]<mailto:[email protected]>, >> www.sust.se<http://www.sust.se> Tel +46 >> 706-442270<tel:%2B46%20706-442270> >> -----BEGIN PGP SIGNATURE----- Version: GnuPG/MacGPG2 v2.0.18 (Darwin) Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/ iQIcBAEBAgAGBQJRPmYHAAoJEOoGpJErxa2pArIP/0TgL6wTndj/Lc0Xi3c5z5Yp comTxaO/iCIrgnhnvu4qYGYZS9zRDIc4QL6VZm4aDSXiu0wJBuOnS11GJN4qDTpf 8wLaehG+6gTdqNsY8lKL+vGTNVMqtBmQIOrftJ4q+N+/fip3tKSwojP8DHwwCScI 3QDj3f9m0/rtTLdllgFrlP8O7HySbqh3tyeGy7qTSJCMvdYM+EJNt5lHg9Vmjnnr iZdE9sPlhfJIWLAb2LSL/ZS921kC5C1phUPKWZOBWKvRccJwee1oOnWoJLskt1c5 pYt014eJBzrQnRbCrqJjzq8cx66vZ7l7oONInB9aqIG0i5o4x+V9Dr0mRq0chGP4 LVrwCijfoBvtHHELMmGnUlWw5YMRHca9HiBgIacMb9vo7/MfI+Gv1EVsXiDcCdjm x4uAH+8yUnV/VUS6DdG0l/uzbTyZta2zbfvXtFM8i1uoDKHIObkPeoeTIKozxN/b ksziXsoyTPqbwMNJzP+/9JD+MJEWoVT1I/v9cfHtX49n8OzBY15VNpVo8ukbPR7X Ec6sVUUQv+R3/5VZw0BSllBuI28ev874mFUQFZgRP18/ltbdWIDboq9lPgyq7QKs gwNeIneqs6IXoPpAhzzxrNnZRBdPhWul9gThLKU5RCQIj3M26BmKZTV/Jbcm/WPq OfSERCnYdBT00MR5W1bS =3LVI -----END PGP SIGNATURE-----
