Re: RFC3728 (1788) (RFC Errata System)

I am afraid that such a change is not possible in an errata without
deprecating the object and defining a new object. 

According to RFC 2578, Section 10.2 

(6)  A UNITS clause may be added.

It is not clear if the UNITS clause can just be changed as Menachem
proposes. If both dB and dBm are valid units it looks like a change in
the semantics of the object. Even if this was just a typo I doubt the
change would be acceptable.

Deprecating an object and defining a new one requires a revision of the
MIB module. Such change cannot be made in an errata. 

I would like to hear Bert's opinion. 

Dan




> -----Original Message-----
> From: [email protected] 
> [mailto:[email protected]] On Behalf Of Menachem Dodge
> Sent: Monday, June 01, 2009 9:59 AM
> To: Scott Baillie; Bert Wijnen (IETF)
> Cc: [email protected]
> Subject: Re: [Adslmib] RFC3728 (1788) (RFC Errata System)
> 
> Hi,
> 
> So I believe there is now consensus that:
>  
>   The UNITS clause should be changed to "0.25dB" (instead of 
> "0.25dBm").
> 
> No other changes are needed to any of the other clauses.
> 
> 
> 
> Are there any objections to accepting this errata? 
> 
> Best Regards,
> Menachem
> 
> 
> 
> -----Original Message-----
> From: [email protected] 
> [mailto:[email protected]] On Behalf Of Scott Baillie
> Sent: Thursday, May 28, 2009 8:07 PM
> To: Bert Wijnen (IETF)
> Cc: [email protected]
> Subject: Re: [Adslmib] RFC3728 (1788) (RFC Errata System)
> 
> Hi All,
> 
> I think the UNITS clause should be changed to "0.25dB"
> and this is not an incompatible change.
> 
> The MIB object in question holds a noise margin value and is 
> expressed in dB. The MIB object does not represent a power 
> level, it represents a ratio of power levels.
> Because the MIB object represents a ratio there is no need to 
> mention power levels at all in the UNITS clause or in the 
> DESCRIPTION clause.
> 
> I have a few comments about some of the previous posts :
> 
> >Without an explicit or implicit reference level, decibel (dB) is 
> >undefined as it represents a ratio.  In our case (as with the ITU 
> >document), the reference level is explicitly specified in the UNITS 
> >clause.  If it were not, it would be implied by the context. 
>  That is, 
> >we are measuring changes in audio levels in telephone 
> circuits.  This 
> >has a VERY long and established history (the bel is named after 
> >Alexander Graham Bell, a decibel being one tenth of a bel).
> 
> I understand that it is convenient to use dBm when expressing 
> power levels in telephone circuits but the MIB object in 
> question is a noise margin, not a power level.
> 
> >However, if one felt forced, compelled, to be more specific, then, 
> >changing "dB" in the DESCRIPTION clause to "dBm" would be the more 
> >accurate change.
> 
> That change would be completely incorrect.
> 
> >Burt the reported proposal for the fix was going to change 
> Integer32 to 
> >Gaug32, or so I understood. That would be really bad I think.
> 
> No need to change the SYNTAX clause.
> 
> Regards,
> 
> Scott.
> 
> On Thu, 2009-05-28 at 00:01 +0200, Bert Wijnen (IETF) wrote:
> > I can possibly live with updatinmg the description clause.
> > Burt the reported proposal for the fix was going to change 
> Integer32 
> > to Gaug32, or so I understood. That would be really bad I think.
> >  
> > If the SYNTAXes and the value-ranges) all stay the same, 
> then I guess 
> > that formally it would still be a semantic change, but it 
> is a matter 
> > of wording change that I could probably accept if the WG really has 
> > consensus on this.
> >  
> > Bert
> >         ----- Original Message ----- 
> >         From: Randy Presuhn 
> >         To: [email protected] 
> >         Sent: Wednesday, May 27, 2009 8:36 PM
> >         Subject: Re: [Adslmib] RFC3728 (1788) (RFC Errata System)
> >         
> >         
> >         Hi -
> >         
> >         > From: "Ray, Robert E. (MSFC-NNM05AB50C)"
> >         <[email protected]>
> >         > To: "Menachem Dodge" <[email protected]>;
> >         "Romascanu, Dan (Dan)" <[email protected]>;
> >         <[email protected]>
> >         > Cc: <[email protected]>
> >         > Sent: Wednesday, May 27, 2009 10:50 AM
> >         > Subject: Re: [Adslmib] RFC3728 (1788) (RFC Errata System)
> >         ...
> >         > Because of this, I disagree completely with the 
> Bert and Dan
> >         on this because
> >         > I don't think the semantics are changed at all.  
> I think the
> >         context, and hence
> >         > the reference level, is completely stated.  We are, after
> >         all, dealing with
> >         > telephone wires.
> >         
> >         Agreed.
> >         
> >         > However, if one felt forced, compelled, to be 
> more specific,
> >         then, changing
> >         > "dB" in the DESCRIPTION clause to "dBm" would be the more
> >         accurate change.
> >         ...
> >         
> >         Agreed.  It's a case of the notation used in the DESCRIPTION
> >         clause
> >         being well-understood within the specialist community, but
> >         being
> >         potentially ambiguous outside of that community.  If the
> >         change
> >         is deemed necessary, it's nothing more than a clarification.
> >         
> >         Randy
> >         
> >         _______________________________________________
> >         Adslmib mailing list
> >         [email protected]
> >         https://www.ietf.org/mailman/listinfo/adslmib
> >         
> >         
> >         
> >         
> ______________________________________________________________
> >         
> >         
> >         No virus found in this incoming message.
> >         Checked by AVG - www.avg.com 
> >         Version: 8.5.339 / Virus Database: 270.12.42/2137 - Release
> >         Date: 05/27/09 07:50:00
> > _______________________________________________
> > Adslmib mailing list
> > [email protected]
> > https://www.ietf.org/mailman/listinfo/adslmib
> 
> _______________________________________________
> Adslmib mailing list
> [email protected]
> https://www.ietf.org/mailman/listinfo/adslmib
> _______________________________________________
> Adslmib mailing list
> [email protected]
> https://www.ietf.org/mailman/listinfo/adslmib
>