Re[2]: MIB comments 1

"Rozman, Chris" <crozman@usr.com> Mon, 31 January 1994 20:54 UTC

Date: Mon, 31 Jan 1994 14:42:18 -0500
Sender: ietf-archive-request@IETF.CNRI.Reston.VA.US
From: "Rozman, Chris" <crozman@usr.com>
Message-Id: <9400317600.AA760056138@robogate.usr.com>
To: Receipt Notification Requested <modemmgt@telebit.com>
Subject: Re[2]: MIB comments 1

     US Robotics Modems use 255 as the 'disable' value for S10.  We could 
     certainly do the translation you speak of below if we have to, but it 
     makes our agent much more generic when we don't need to do such 
     things.  If others out there use 255 (or if nobody uses 0) as the 
     'disable' value, I would like to define it to be 255 in this MIB.
     
     -Chris
     

______________________________ Reply Separator _________________________________
Subject: Re: MIB comments 1
Author:  Steven Waldbusser <waldbusser+@CMU.EDU> at Internet
Date:    1/31/94 12:56 PM


>Upon reflection, I believe that '0' is not used since it takes a 
>finite tie to recognize that carrier has been lost. Also, I believe 
>that it is more common to use '255' to disable this function, 
>therefore I would propose that we drop the mdmLineCarrierLossControl 
>object and use a value of '255' to disable.
     
Is 255 really in common usage?  Use of zero seems much cleaner design. 
For example, should a reason emerge for making these times greater 
than ~4 minutes, we would have a distinguished value in the middle of 
the range.
     
If there are implementations that use 255, the agent can always 
translate before placing the value in the S-register.
     
     
Steve

Re[2]: MIB comments 1 Rozman, Chris