Re: [DotGNU]Re: A Proposal for a Solution (Jabber)

[Barry Fitzgerald]

Adam Theo wrote:
> 
> Norbert Bollow wrote:
> >
> > > Adam Theo wrote:
> > > >
> > > > I have a proposal to make. I have been thinking about this all day, and
> > > > believe I am ready to advocate such a position.
> >
> > Let's move this discussion to the developers mailing list, so
> > that it won't get in the way of the work of the auth list (which
> > is to quickly create that that quick-and-dirty first incarnation
> > of DotGNU virtual identities. )
> 
> yes, i agree, and am starting to do so now. from now on thius list will
> be about advocating using Jabber as a basis or in *some* way for DotGNU.
> 
> >
> > "Dan Kuykendall (Seek3r)" <address@hidden> wrote:
> > > I never agreed with those that think this can all be done peer to peer.
> > > Sometimes it is best for companies to implment a trusted auth server in
> > > their environment
> >
> > Yes.  That's why the long-term implementation of "DotGNU virtual
> > identities" will be done on the DotGNU platform, the
> > http://dotgnu.org/arch.htm for some details.
> >
> 
> yes, i feel that Jabber has many benefits, and is powerful, flexible,
> and with some work, more than secure enough for our needs at DotGNU. I
> can easily see Jabber not only solving the Virtual Identities problem in
> the long run, but also solving some web services and distributed
> computing needs of DotGNU in general.
> 
> > The key point is that then the auth system, just like any other
> > application that runs on this platform, can run in P2P mode, and
> > it can run with a single trusted auth server, and it can run with
> > a network of auth servers where the auth database is replicated
> > on each auth server, and the copies of the database are
> > automatically kept in sync.
> 
> not sure what your point is here, sorry. probably just because it is
> after midnight here, and i'm tired a bit....
> 
> > Yes, XML has a lot of mindshare.  It will be wise for DotGNU to use
> > protocols that are built on XML.
> 
> yes, this is something i agree with completely! XML is the way to go,
> and pre-existing protocols are the way to go, as long as they fit in
> perfectly with what you need to do. no need to re-invent *any* part of
> the wheel just because you want to create something grand and new.
> 
> so, at the very least, XML. let's try to have DotGNU be a 100% XML
> system. it is certainly possible, and has many benefits.
> 
> and, to hype my main point, let's take a look at how the Jabber protocol
> can take care of what we need now. we are getting an idea of what we
> need in this DotGNU system, so let's see how Jabber can fill in as many
> of those needs as possible, so we don't have to take the massive amount
> of time to do it ourselves.
> 
> after all, we *need* DotGNU to be nearly done and working well within a
> year. or at least have something great to show for it. if we don't, we
> stand a good chance of being left behind, even if we have a superior
> technology (just think of all the times this has happened in the past to
> others... oodles and oodles of times.). making use of technologies and
> protocols that already exist and fit our needs while being 100% open
> source/GPL will save us *so* much time.
> 
> >
> > > > Now, if all of you are still skeptical about how Jabber can be used, I
> > > > propose that this list choose one or two people from among us to learn
> > > > about Jabber. Give them a week or so, and have the report back to the
> > > > list on what they think. This way all of us can see how Jabber looks
> > > > from a FSF and DotGNU perspective, without all of us having to learn
> > > > about it.
> > >
> > > I agree with this plan. I would jump in and start studying jabber in
> > > more depth if you guys want.
> >
> > Sounds good to me.  However before you do this, please have a
> > good chat (perhaps on #dotgnu) with Myrddian <address@hidden>
> > and Barry Fitzgerald <address@hidden>.
> >
> > Myrddian and Barry are working on some ideas and specs that you
> > should be aware of when you look at the Jabber for checking
> > whether the protocol would be useful for DotGNU.
> 
> ok, will do. i don't use IRC, but i will contact them by email.
> 
> >
> > > 1) It exists, and is a working distributed solution
> > > 2) Its been gaining acceptance over time
> > > 3) It has several companies backing it as well as many in the free
> > > software community
> > > 4) Its got plenty of 'coolness' factor, which cant be overlook even tho
> > > it may sound silly to outsiders, 'coolness' is a key ingrediant to a
> > > successful free software project
> >
> > I agree that these are good reasons for looking into Jabber.
> 
> yes, especially the first and fourth.
> 
> Myrddian wrote:
> > Well Jabber as far as I see it is a good way for the Virtual Identities 
> > system to work.
> > However, the scope of DotGNU is far more larger and is a tad bit more 
> > scaled that
> > Jabber
> > was intended for ( I feel any how ). So extensions as such and a better 
> > mechanism
> > might be needed. I'll have to read more on it any how, but it's quite good 
> > so far :)
> 
> ok, p2p: when jabber was first really being designed (back in 1998, i
> believe), p2p was just hitting the streets. jabber was originally
> designed with a server/client model, but now that p2p is proving itself
> to be a worthy model, jeremie miller and the other founders are starting
> to turn it into a p2p system, at least much more so than it already is.
> and this will not take much work. the jabelin server (the OSS
> implimentation of the protocol) is very small. small enough to be
> embedded into a p2p client itself with minimal work, bringing it closer
> to a true p2p model.
> 
> scaleability: jabber is very scaleable. it is currently being used by
> many different organizations on many different networks for many
> different needs. one person i know just installed it at his office, for
> his co-workers to communicate with each other. he has easily set it up
> on their internal network, and the boss knows they are not using it to
> chat with friends or family while on the clock. the jabber.org local has
> i think thousands of users on it, and it is not a top of the line server
> at all, if i recall correctly.
> 
> anyone can run their own jabber server, like they can run their own
> sendmail or nntp server... and it can talk to the outside world or only
> be used for internal office talk.
> 
> it uses xml, and is open to embedding other xml specs in it's messages.
> therefore it is easy to do SOAP and/or XML-RPC calls over the jabber
> system, or to send custom formats only the recipient is intended to
> understand and use.
> 
> one of the neat things about jabber is the username and resources. the
> jabber user name (or JID as we call it) looks like an email address:
> address@hidden but it can also (note: optionally) have a 'resource',
> which comes after a slash at the end of the JID:
> address@hidden/resource. a resource is like a 'sub-account' (or you
> can think of it that way). a user can log into a resource under their
> account for special purposes. and you can be logged into your account
> simultaniously as long as it is from different resources. so say you
> have 3 jabber client s running right now: your desktop comuter one,
> named 'desktop', the one embedded into Mozilla which is embedded into a
> kitchen net appliance next to your stovetop named 'kitchen', and another
> in your car (for when you are on the go) named 'auto'. now, someone can
> send to a specific resource, and you will get it only there, or can just
> send to your account in general (normal) and get you wherever you are at
> the moment.
> 
> i see the above thing of resources as having lots of potential, which i
> explain below:
> 
> for DotGNU, you will need an addressing or ID scheme, right? some sort
> of identifier system to use so the components and applications in DotGNU
> know where each other are at. now, you could use the normal
> http://www... URL system, except that depends on the www and web
> servers, as well as the http protocol, correct?
> 
> how about using jabber's naming and addressing infrastructure? it
> already exists, and already has support for it. so your DotGNU
> application's 'name' could be address@hidden or if you want to
> use resources: address@hidden/wordprocessor. it would log into it's
> jabber/dotgnu account and send a SOAP-ized jabber message to some other
> application, residing at address@hidden/tryme or whatever.
> the other application could then reply, sending back the spell-checked
> document.
> 
> this is just a beginning. but the advantage to using his, is that the
> addressing, resolving, and identifier systems are already specified and
> in place.


Hello, 

Good ideas here, please see the RLS spec at
http://www.dotgnu.org/spec/rls.htm ... it should be relatively
self-explanatory.  I will be releasing a companion to this spec that
explains the token scheme tomorrow.  It needs a little cleaning up
before I release it.  Have fun. :)

        -Barry