Dear All,

I think the application class discussion we had is very important for a
useful ALTO service. This poses two requirements: the ALTO server to
understand application requirements and the ALTO client to understand
the ALTO response.

Let me try to summarize:

The ALTO service is not for application specific peer selection or
implements multiple selection algorithms. 
However, if both ALTO application and ALTO service provider should
benefit - otherwise it will not be used - it is important that both
understand each other while keeping the protocol simple:

- the ALTO server has to understand simple application requirements (as
discussed before), e.g., application class; I would propose to be more
abstract here

- the ALTO client (application) has to understand the response from the
ALTO server (has not fully been discussed before, see Zoran's email):
what is the basis for the ALTO information - as it is unlikely that the
ALTO server will reveal all the network internals, some abstract
notation is needed that can be easily understood by the client

Not all of the information is mandatory, but we might consider a
mandatory and an optional set.

Do we have this in the current requirements? I think this is very
important.

Regards,

Wolfgang

------------------------------------------
Dr. Wolfgang Kellerer
Senior Manager
Ubiquitous Networking Research
DOCOMO Communications Laboratories Europe GmbH
Landsberger Str. 312
80687 Munich, Germany

Tel. +49-89-56824-222
Fax. +49-89-56824-300
E-mail: [email protected]
http://www.docomolab-euro.com

Managing Directors (Geschaeftsfuehrer):
Dr. Toru Otsu, Dr. Narumi Umeda, Tsutomo Sakai
Amtsgericht Muenchen, HRB 132976
------------------------------------------
private URL: http://www.wolfgangkellerer.de


> -----Original Message-----
> From: [email protected] [mailto:[email protected]] On Behalf
Of
> Zoran Despotovic
> Sent: Donnerstag, 12. Februar 2009 09:39
> To: Stanislav Shalunov
> Cc: [email protected]
> Subject: Re: [alto] differences among applications
> 
> 
> Stanislav Shalunov wrote:
> > As others pointed out, an ALTO protocol is not expected to make peer
> > selections for the apps.  On a high level, it's expected to provide
> > information about the network and about ISP routing preferences.
> Makes sense. In fact, I guess that is the only correct view.
> >
> > While peer selection preferences vary from application to
application
> > substantially, the network itself is the same, and so the
information
> > about it remains valid.
> Same as above :). The point is to define precisely what the network
> should say about itself. Pure locality information may not be enough
> for
> streaming apps, or at least not as good as for BT. Or, we need to
prove
> or disprove that.
> 
> Best,
> Zoran
> >
> > -- Stas
> >
> 
> _______________________________________________
> alto mailing list
> [email protected]
> https://www.ietf.org/mailman/listinfo/alto
_______________________________________________
alto mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/alto

Reply via email to