[jdev] Re: CSpace instant messenger integrated in Jabber

thomasasta at gmx.net thomasasta at gmx.net
Sat Aug 5 10:24:48 CDT 2006


Hal, thanks for the ideas for jabber

jabber replacing email with Link offline communication and integrating p2p waste-like filetransferare innnovative ideas.
http://www.jabber.org/jeps/jep-0174.html

Yes, Cspace has no servers and then no offline messages, as storing them in throd party nodes would be too insecure and difficult.
I suggested a sent-queue, which then allows sending messages to offline buddies, while the transfer is then when they are both online.
http://tachyon.in/pipermail/cspace-users/2006-August/000234.html

You all see, discussing both methods (server/DHT) brings up a lot of good ideas to both...
 
Kind regards..

-------- Original-Nachricht --------
Datum: Sat, 5 Aug 2006 11:05:34 -0400
Von: "Hal Rottenberg" <halr9000 at gmail.com>

> I agree, this was not the request, just to discuss technically, how they could technicall join. For Python GAJIM, where the discussion started, it is easy to add CSpace, for all others with enxryption the question is rised, why still serverbased. And all others can profit from this discussion to built modern messengers!
>
> The binding between Cspac and Jabber is the enxryption: Every jabber client with encryption does not need servers. Its a WASTE routing !

That's nice, but it has nothing to do with Jabber.  You are solving a
different problem perhaps, or you just need to change your mindset to
look at it from our POV.  We are replacing email.  Server-centric is
where the value is added to a piece of dumb client software.  There
are a lot of use cases where the best solution involves a server.
Aside from that, we do have an ad-hoc serverless protocol called
Link-Local Messaging [1].  Not the same as what you are doing, but it
is serverless.

[1]: http://www.jabber.org/jeps/jep-0174.html

As you see, we can't just drop the server model.  However, we are
lacking a strong encryption method right now (although there are
plans).

> If you go to a fiends pc you can easily log on with the CSpace-ID. well it is a help construct and serverbased, but you get your account open without your buddies, right, you need then to insert them again. But: There is the option to take your profil with all your buddies on a USB stick and then you can anter any PC. In the next release this profil is a ZIP file, readable from linux and windows, not matter which PC you use..

Right now I can pull up Agile Messenger on my mobile phone and get my
buddy list.  :D

> As there are python jabbers and enxrypted jabber there is basis for a discussion, and in general: why do we need jabber with servers, if we can call cspace jabber without servers? As CSpace-ID is optional, we could make jabber servers working as well with kademlia DHT and RSA-Keys additionall, so if any jabber server is offline, the network would still work, as it is decentral

I'm sorry you are just apples and oranges here.  Don't get me wrong, I
think WASTE was a novel idea that could have gone much further, and it
seems you are trying to take it there and that's great.  But what you
are talking about is not just some trivial change, it's scrapping it
all and starting over.  There is no compelling reason to even do that.
 But read up some more aqbout Jabber.  You might find that we have
some common problems that need solving (like encryption) with which
you can lend your expertise.

I think a WASTE-over-Jabber S2S negotiated, but closed-loop P2P file
transfer would be a neat thing.  Not sure how or if it would work but
there are possibilities to consider.  Note the "over jabber" part.
Your negotiation traffic would have to travel as XML from jabber
server to jabber server, and then clients which expose to the server
or other clients that they have e.g. the "WASTE capability" could then
go P2P.


-- 
Psi webmaster (http://psi-im.org)
im:hal at jabber.rocks.cc
http://halr9000.com

-- 


Echte DSL-Flatrate dauerhaft für 0,- Euro*. Nur noch kurze Zeit!
"Feel free" mit GMX DSL: http://www.gmx.net/de/go/dsl



More information about the JDev mailing list