FIRST DRAFT - C63 Negative Comments on using Dated References
October, 2016
ELLIOTT, MAC – ACS Test Lab
Is not what Ed Hare suggests what ETSI does to an extent? Was just perusing some of their standards and there seems to be a mixture of dated and non-dated standards (but mostly dated…).
The following verbiage is found in the reference section of the proposed 2016 Draft of 301 489-5 for instance. It is also found in a lot of other ones I looked at.
“References are either specific (identified by date of publication and/or edition number or version number) or non-specific. For specific references, only the cited version applies. For non-specific references, the latest version of the referenced document (including any amendments) applies.”
We could add the provision regarding the 1-year transition period that Ed mentions in cases that warrant it – especially if labs need to get the standard on their accreditation scope for Regulatory Purposes.
I agree with Ed Hare’s main points and suggestions of letting the Working Groups make the recommendations on a case-by-case basis and letting relevant Subcommittee and eventually the Main Committee decide if they agree with them.
Do we have any ETSI contributors that could comment on discussion regarding this and how ETSI came to their position?
GRIFFIN, ANDY – CISCO
With undated references within a standard (for example C63.4 has an undated reference to C63.5) when C63.5 is updated, it becomes a requirement ‘immediately’… Yes C63.5 will never be listed by the FCC in their regulations but because C63.4 is referenced, then C63.5 is indirectly listed as an undated version.
When thinking about dated/undated references, the response to issues should be based upon the users/tests labs/manufactures that are not an ‘integral’ part of the ANSI processes. There are thousands of manufacturers and test labs out that have nothing to do with developing the standards.
These are a few of the arguments against dated standards.
Issue/Group / Against Dated Standards / For Dated StandardsWG / Undated standards are easy for the WG developing the referencing standard because they can just ignore the changes without thought. / The WG has to make a ‘conscious’ decision to update the reference, hence has to assess impact of the changes.
WG / The WG should not have control over the requirement, these are dictated by the publication of the ANSI standard and hence must be adopt the requirement as soon as possible.
This assumes that everybody knows what is going on in all the other relevant committees and that the impact is easy to follow and understand. / The WG has control over when the revision is implemented, if there are major changes which require significant time for implementation then they could delay implementation until the next version.
Cal Lab / With undated references the cal labs have to rush around with urgent customer request to adopt the new requirement.
Test Lab / Test labs have to constantly monitor ANSI publications to see if any of the ‘undated’ standards have been updated and the follow the process defined in Section 1.
Accreditation / We shall use the accreditation bodies to manage the change from one version of a standard to the next. / Not all test labs are accredited.
Test labs do not need to be accredited to assess products against some of the FCC requirements.
Several versions / A possible problem.
C63.10 refs to C63.5:2014
C63.4 refs to C63.5:2009
C63.26 refs to C63.5:2016
I’m a test lab, do I need 3 different types of antenna calibration results?
/ The alignment of the various standards with regard to the version of C63.5 (for example) is kept as close as possible. With small amendments published if needed.
The FCC always has transitions to go from one revision of a standard to another, which will minimize the impact.
Revision timing / It takes too long for a standard to be published then important updates will take too long to be implemented. / If required, the new quick amendment process can be used to update the dated references.
Manufacturer / Manufacturers have to constantly monitor ANSI publications to see if any of the ‘undated’ standards have been updated. When they get products assessed they must use labs that have followed the process defined in Section 1.
Manufacturer / I’m a manufacturer, I establish a contract with a lab to have testing done to ANSI C63.4:2014. C63.5 is published on the 15th of July. The test lab tests the product on the 16th of July, if the process defined in clause 1 cannot be done on the same day, then the test lab is in breach of contract.
Manufacturer / The new version of C63.5 has new balance antenna requirements. It will take a while to implement the process defined in section 1, during this ‘period’ the results may be obtained with an ‘invalid antenna’…. the manufacturer will have to re-measure the devices with the valid antenna.
Manufacturer / When assessing a report against C63.4, and the test lab has not put which version of C63.5 was used to assess the antenna. Is the test report valid?
Hare, Ed – ARRL
Mandating that all references must always be dated will remove flexibility that is better left to the Working Group and balloting body.
Three potential pitfalls with respect to the use of undated standards:
1.In the United States, and presumably in Canada, if regulators change rules, they must put rule changes out for public comment. Incorporating a standard by reference is very much making the language of that standard into regulation, and so the content of the regulation, and the content of the standard, must be subject to public review, comment, and even various federal appeals process. I am not even sure an undated reference can be incorporated by reference, and if it were, and changes are made to the standard, that has the de facto effect of writing new regulations without the opportunity for public comment.
2.As a standard is created, very often, references to other standards are normative, meaning that the content of those other standards are part of the requirement of the original standard. As that standard is balloted, the balloter also determines that he or she approves the requirements in normative references. If undated normative references are used, and those standards change, this has the effect of changing the requirements of the standard that refers normatively to other standards, very much making an essential change to the requirements of a standard without that original balloting pool being able to approve or disapprove.
3.If a standard refers to undated standards, in effect, what would stop someone from using ANY version that other standard? If C63.5 is specified, a revision three-years old could be used. If changes are made to a normative reference, how long can the old one be used? Must the new one be used? Right away? Within a year? Yes, all of this could be incorporated into each standard.
It is my opinion that rather than having C63 make an absolute determination on undated vs dated references, we allow these decisions to be made on a case-by-case basis by each Working Group as it develops a standard. In some cases, it makes sense to want to mandate dated references, and in other cases, it may make sense to permit an undated reference, while in other cases, it may make sense to say that the current version of a changed reference may be used, but for a <1> year period , the previous version may be used, etc.
I’d like to hear from those with audit experience, to understand how undated references could impact their ability to make an audit, in either direction.
Ed Hare in response to Mac Elliott
If that is the case, Mac, then a NO vote will NOT implement a mandate that dated references must be used. Working Groups and balloters will be able to decide that on a case by case basis, just as they do now. I actually prefer dated references most of the time, so regulators can incorporate them with regulatory certainty, but I am leaning towards a NO vote to give the WGs the flexibility they have right now. The way the Motion was worded, a NO vote does not change the present status quo, which is mute on the subject.
Ed Hare –again in response to comments from individuals
The Motion is: The C63 Committee shall adopt a policy for the use of dated references in all C63 standards.
The way the Motion is worded, a YES vote will mandate that dated references must be used.
A NO vote will not mandate that dated references must be used, but a NO vote to this Motion will not prohibit dated references. With a NO vote, it can still be up to the Working Group and ballot pool as to whether a reference shall be dated or undated.
Hare, ED – in response to Dennis Ward
We are, of course, an ANSI-accredited body, not an IEEE one, so the IEEE guidelines do not apply.
The motion was accepted and not declared out of order, and the motion as written mandates that dated references always be used in C63 standards. Dan Hoolihan had mentioned an earlier motion to mandate that we always use undated references, but I was unable to find it going back to 2010. I may have missed it.
If the first motion doesn't exist and the second motion fails, then each WG and balloting body is free to use dated or undated references as they deem appropriate. If this motion fails and the first motion exists, C63 is mandated to never use dated references.
Heirman, Don
There was a major reason why the 2006 edition of C63.5 was used in C63.4 2009 as in the pastI you could use the 1988 and the 1998 versions where were quite different and did not keep up with the latest international calibration techniques. One of these required antenna factors recorded for both polarizations and at two distances. The other did not but was still not in full synch with the international approach in CISPR. The 2006 edition made the harmonization with international use. So if you used a dated earlier edition, you would have to calibrate your antennas twice: once for the dated version you used and another time when making emission measurements for CISPR standards. That was considered intolerable which lead to the need to use the latest edition of C63.5 (happened to be the 2006 edition) for one calibration method for worldwide use. The 2006 date was kept and referenced as it was the latest edition and made it clear that the 1998 and 1988 editions could not be used. The FCC emphasized this in one of their KDBs. If no date is given AND the introduction to the normative section did not say that if undated the latest edition must be used, they you are correct that any edition can be used. But we do not let that hanging as we clearly state in the introduction to the normative references for example in C63.42014 this: For dated references, only the edition cited applies. For undated references, the latest edition of the referenced document (including any amendments or corrigenda) applies.
We seem to continue to forget that C63 is in complete control of what is in a newer version of our standards in what we for or against approval. This was the reason in the first place for using undated standards when C63 approved that approach a couple of years ago. If the newer edition would bring on the problems that are being cited in these exchanges, simply to vote to disapprove as that is the responsibility of the balloter.
But we need to motivate Working Group members to improve our standards. One of the best ways to acknowledge their work and the interest of C63 members is that what the worked on will go into effect as soon as it is published. Why do more work if the new work might be left for years before it will be even a candidate for using?
Our standards are voluntary as are all standards including the IEC for example. Regulatory authorities have to state which standard is to be used or referenced. Even today, there are 6 clauses /subclauses that the FCC, for example, does not accept from C63.4-2014 and past editions. The point is that if the issue is using the latest edition (via the undated standards route), at least our regulatory agency will decide which clauses to use and which not to use. That would be a check and balance as the FCC does listen to those that they regulate for the most part. I pointed out the “cons” white paper what they said about this subject in the NPRM that recognized referencing only the 2014 edition of C63.4 They recognized that they will have to deal with the next edition (which is the same as not using dated standards). I hope you all read that well-crafted explanation of their position. And, note that they are on C63 and participate in our standards work.
I realize that the above is not a majority opinion on this subject. But, at least it should be on record to show that there are positive outcomes also from using undated standards.
FCC
The subject Report and Order
In the Matter of Amendment of Parts 0, 1, 2, and 15 of the Commission’s Rules regarding Authorization of Radiofrequency Equipment Amendment of Part 68 regarding Approval of Terminal Equipment by Telecommunications Certification Bodies
ET Docket No. 13-44 RM-11652 REPORT AND ORDER
Adopted: December 17, 2014
Released: December 30, 2014
Clause applicable for dated/undated standards as it applies to C63 standards
81. We also disagree with those commenters who assert that we should not adopt the new ANSI standards because they include cross-references to other undated standards. These commenters believe that this practice could inadvertently result in new compliance requirements by introducing revised editions without the opportunity for comment or defined transition periods. We recognize that when standards incorporate other standards by reference, the use of undated references may be unclear to users – particularly when there are several versions of the referenced standard. However, requiring that only dated standards be cross-referenced would not always result in certainty regarding compliance requirements and would prevent us from realizing the benefits in adopting the new ANSI standards. Standards development organizations have sought to find the appropriate procedures to clearly deal with these issues. ANSI-ASC C63 has decided to use undated references to other ANSI-ASC C63 standards, carefully reviewing the effect of any revisions as part of the standards development process. We accept this convention and we realize that under this approach, there could be a revision to a standard cross-referenced referenced in ANSI C63.4 or ANSI C63.10 that would result in a substantive change in requirements. In such a case, the revised cross-referenced standard would not take effect until the Commission or OET on delegated authority completes a rulemaking adopting that change. To the extent that parties are unclear as to which version of a particular standard is currently in effect, OET will continue to provide guidance via the KDB on the use of updated references in ANSI C63.4 and ANSI C63.10.
Here is the KDB from the FCC on C63.5
Hofmann, Bob – comments
The words that follow were presented by Mr. Dennis Ward, and I agree that they allow the needed flexibility for our standards.
"Most if not all ANSI standards now state “For dated references, only the edition cited applies. For undated references, the latest edition of the referenced document (including any amendments or corrigenda) applies.” This clearly leaves it open to use either when necessary. So, to say that it is mandated to use undated standards is not correct."
Ward, Dennis – PCTEST
There is basically only one advantage to a dated reference in standards. That being “legal certainty”. Both the regulators/legislators and the users of the standard know exactly what is to be done.
That however is frankly outweighed by problems introduced when using dated references. While it may introduce a legal certainty as to what must be used/done, “nevertheless, the disadvantages of such a method prevail…….every time a standard is adjusted to the latest state of the art or even entirely replaced, legislation using that standard will also have to follow if it wants to respond to the latest state of the art.” This later is basically what the FCC ET Docket No. 13-44 RM-11652 REPORT AND ORDER infers and what is clearly explained in the EU Commissions report Enterprise Guidelines from whence the quote comes.
Using dated references means that if, for example, C63.5 was updated, C63.4; C63.10, C63.26, and any standard referencing C63.5 would have to be adjusted/updated it the referencing standard wanted to “stay up to date” on the latest methodology in C63.5. How many countless other standards and regulatory requirements would also then have to be updated and reissued giving the latest date of reference for C63.5.
And, if there were also numerous other referenced standards in any of these, then those too would have to be reissued/adjusted with new dated references. For example, C63.4 and C63.10 list 5 ANSI standards and C63.26 lists 6 other ANSI standards. If anyone of the listed standards changed, then each would have to also change. If several changed, but not at the same time, then you could wind up changing “adjusting” the same standard several times in one year.