[Standards] mobile optimizations

Peter Saint-Andre stpeter at stpeter.im
Wed Feb 20 16:15:29 UTC 2008

juha.p.hartikainen at nokia.com wrote:
> Hi!
> Regarding optimization we get good figures from compression zlib (at
> least good start).
> Also in some comments:
> Optimizing roster: Good start, but also should be considered vCard based
> traffic,
> which causes traffic load. Caching the data to devices is good start and
> resurrect the 
> xep-0150 could be easiest and straight forward action. Also adding use
> case "Caching and 
> retrieving vCard" with etagging. Of course Pubsub could be one approach
> for vCards, 
> but requires extra implementation.

Thanks for the suggestions. Etags (or something very much like etags)
will definitely be on the agenda for the XMPP devcon this weekend,
because we can use it for rosters, vCards, and perhaps other domains too.


> Cheer's
> <Juha>
>> -----Original Message-----
>> From: standards-bounces at xmpp.org 
>> [mailto:standards-bounces at xmpp.org] On Behalf Of ext Joe Hildebrand
>> Sent: 19 February, 2008 17:01
>> To: XMPP Extension Discussion List
>> Subject: Re: [Standards] mobile optimizations
>> Could this be added to XEP-198?  Basically a pause command 
>> that would cause stuff to get buffered on the server.  There 
>> would need to be an ack that comes back to avoid races.  
>> Additionally, I could imagine the thing doing the buffering 
>> could toss old stale presence stanzas when new ones are 
>> received from the same full jid.
>> Combined with XEP-198 quick-reconnects (or BOSH) with long 
>> server-side timeouts, and it seems pretty efficient for this 
>> use case, at the expense of some DEFLATE performance.
>> On Feb 19, 2008, at 2:47 AM, Fabio Forno wrote:
>>> On Feb 16, 2008 6:10 AM, Peter Saint-Andre 
>> <stpeter at stpeter.im> wrote:
>>>> It would be good to analyze the relative percentage of presence to 
>>>> messages. My sense from watching the XML console in Psi is 
>> that most 
>>>> of the "noise" is presence.
>>>> In general I think it might be difficult to determine which stanzas 
>>>> are worth waking up for (some messages might just be PEP 
>>>> notifications), but ignoring presence seems like a good start.
>>> I must study how privacy lists work (don't know if they need some 
>>> extension or we can use them in this way), but the best solution in 
>>> this case seems a set of privacy rules allowing asymmetric presence 
>>> distribution. When I'm online with a mobile there may be at least 
>>> three groups of users:
>>> - symmetric presence exchange (as it works now, I want being 
>>> constantly informed about their status and let them know mine)
>>> - asymmetric presence: let them know my status, but don't update me 
>>> with theirs until they do some action (e.g. a message) or I query 
>>> explicetely
>>> - no presence at all: don't bother with these contacts when 
>> I'm with a 
>>> mobile
>>> --
>>> Fabio Forno, Ph.D.
>>> Bluendo srl http://www.bluendo.com
>>> jabber id: ff at jabber.bluendo.com
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/x-pkcs7-signature
Size: 7338 bytes
Desc: S/MIME Cryptographic Signature
URL: <http://mail.jabber.org/pipermail/standards/attachments/20080220/5b19b704/attachment.bin>

More information about the Standards mailing list