Re: draft-ietf-tsvwg-iana-ports-09: How we have resolved WG last call comments

[Lars Eggert <[e-mail]>]

Hi,

On 2010-12-2, at 14:52, Eliot Lear wrote:
> This is a distinction without a difference.  Nobody cares anymore about
> the port range, and the argument is circular, if it was every really
> stable to begin with.  This distinction was made at a time when perhaps
> the only port in use over 100 (not 1000) was RIP, with the assumption
> being that services would be registered.  That flew out the window
> largely with X, and then out of the planet with Windows, and out of the
> solar system with mobile devices.  Please.  Let's move this distinction
> into history, where it belongs.

Assume we got rid of the distinction between ports larger and smaller than 1024 like you're asking us to. IANA just treats them the same, i.e., FCFS with Expert Review.

Assume that someone asks for a port number. IANA rolls the dice and gives them 842.

They put that in their code and all of a sudden realize that on many platforms, they need to ask the user for admin rights in order to bind to that port.

They won't be happy.

Right?

Because in practice, on many systems, it DOES matter if you have a low port or not.

(So instead what we do is we limit the FCFS range to 1024-49151 and make people who think they want a low port go via the IETF.)

Lars