Dated and undated references

By J. M. Woodgate B.Sc.(Eng.) C.Eng.  MIET SMIEEE FAES Hon FInstSCE MIOA

Dated and undated references

Terminology: IEC now refers to all sorts of publication as 'document' in running text, not 'standard' or 'technical report' or….

IEC allows documents that are referred to in other documents to include the edition number and publishing date, or not to include them. The latter are 'undated references'. Historically, before regulatory issues emerged, all references were undated and users of standards were '…encouraged to investigate the possibility of applying the most recent edition…'. This, as can be seen with hindsight, caused many problems, because people adopted different editions of the same standard (very often one that was many years out-of-date). So dated references were allowed, but only if a specific text-level reference was include in the referring standard, such as 'This shall be carried out as specified in Clause 7 of IEC 60000-9:1956.'

NOTE      IEC doesn't allow a reference quite like that; the date is given in the Normative references clause, usually Clause 2

This rule has been relaxed recently, so that committees can decide to have all references dated, but the IEC editors still tend to query it.

There has been much debate about the relative merits of dated and undated references, especially in IEC TC100. This is because the pros and cons are very nearly equally balanced. Some people wrongly believe that one or the other requires much more work by committee officers, but that is not so. The misconception is due to the necessary maintenance activity for each type of reference had never been thought through.

Care and feeding of undated references

In principle, an undated reference represents a big gamble on the part of the originating committee – how do they know what the 'owners' of IEC 60000-9 might do to it in the future (even withdraw it).  So someone has to check each new edition of IEC 60000-9 to see whether anything changed that invalidates the reference, vague though it is.

Care and feeding of dated references

A dated reference isn't a gamble, because the reference is quite explicit, and this is what, for many people, tips the balance in favour of dated references – the user of the standard always knows which edition to apply. (Outdated editions can still be supplied by IEC and some National Committees.) However, a maintenance activity is still required. Someone has to check each new edition of IEC 60000-9 to see whether anything changed that invalidates the explicit reference. This is a much easier task than checking all through an undated reference to look for changes.

Enter the European Commission's Legal Department

The legal eagles woke up last year to the 'uncertainty' created by undated references and it is now ruled that standards (since it's only standards that qualify) can be 'notified' in the Official Journal as providing prima facie evidence of conformity with a Directive or Regulation only if all its references are dated. Of course, this applies to ENs; IEC doesn't have to take any notice, so the dated references of standards adopted from IEC that have undated references have to be added as Common Modifications. To do this properly, someone who knows the standard very intimately has to determine why the undated reference was included so that it can be converted to a dated reference, or perhaps several different dated references. Clearly, this is no easy task.

It is often said that these tasks place a large additional work load on committee secretaries, but in fact the work can be shared with other committee members who have access to the relevant standards. Committee members from large organizations often have wide access, as their employer subscribes to costly 'on-line' libraries.