Re: Deja vu Again

[Juergen Schoenwaelder <schoenw@ibr.cs.tu-bs.de>]

>>>>> David T Perkins writes:

David> The cycle goes....  

[...]

obviously

David> I and others would like to see this cycle broken. Does using
David> XML-based techniques really make a difference? Is part of
David> reason for XML the old TEXT-based versus BINARY-based
David> application protocol religious argument? There is no need to
David> fight that battle again.

I guess it is part of the battle.

David> XML encoding of data specified in existing MIB modules is just
David> a tax with no added value.

You are certainly right from a very abstract perspective. However, the
number of ready to use tools for processing XML is much greater than
the number of ready to use tools for processing BER. But despite this,
it is easy to write a gateway which translates BER into XML so then
again it is not an issue...

I guess one of the main points underneath this is that the SNMP
community did not manage to produce widely available good frontends
that satisfy ISP network operators requirements (means ASCII or XML
formats which allow to process the data will all kind of existing text
based processing tools (cvs, diff, ...) rather than GUI).

In the enterprise market, where people run tons of bridges, the
approach with management products and GUIs has gained much more
deployment. You can actually configure most of the bridges via SNMP. I
am writing code at the moment to setup layer two VLANs on bridges from
various vendors via SNMP using scli. It works, although you have to
poke into proprietary MIB objects. The point however is that someone
has to implement an operator interface on top of the programmatic
interface. That takes time is definitely something ISPs would like to
not invest time in. (Another good example is SNMPv3 administration.
The MIBs are fine but in order to actually do it, you need to have an
operator interface which provides you task-oriented commands to do
what you want to do.)

David> Defining operations differently than that SNMP operations
David> affects the schema. Operations and schema go hand in hand.

David> Someone, please briefly describe the core motivation for this
David> new effort.

I guess it is as simple as this: The ISP operators are tired to write
expect scripts. (But perhaps I am wrong.)

/js

-- 
Juergen Schoenwaelder    <http://www.informatik.uni-osnabrueck.de/schoenw/>



--
to unsubscribe send a message to xmlconf-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.
archive: <http://ops.ietf.org/lists/xmlconf/>