IESG proposed statement on the IETF mission

[Eric Rosen]

> The critical difference  is whether or not multiple  *operators* will have
> to communicate using the protocol.

> As soon as  the lights in NYC (owned, provisioned and  bought by NYC) have
> to operate  with the  light controller  in Chicago, then  we need  an IETF
> standard.

You  are   making  a  host  of  unwarranted   assumptions.   It's  perfectly
conceivable that NYC has, say, five  different networks, one for each of its
five boroughs, and that each network  is run in an autonomous fashion.  It's
equally  conceivable that  the  main  street light  control  facility is  in
Manhattan, and it needs to talk to the lights in each of the five boroughs.  

If  one thinks  of  street lights  in a  set  of smaller  towns, it's  quite
conceivable that  each town has its  own network, but that  light control is
done  from  a  regional center  jointly  funded  by  the towns.   It's  also
conceivable that  a bunch  of towns  in one state  might contract  out their
light control functions  to a company whose monitoring/control  center is in
another state entirely. 

So I don't think your criterion rules out the street light protocol. 

There are  some things which  are clearly ruled  out by your  criterion: for
example, OSPF would be ruled out, and  so would DHCP.  I think SNMP would be
ruled out as well (one operator managing another's routers?). 

I'd say then that your criterion  does not reflect any existing practice nor
does it reflect a desirable practice.