a question of "Class"

Here is some background for discussing errata at today's WebCGM WG 
teleconference.  About the W3C "correction class" question, earlier I wrote...

At 10:50 AM 8/29/2007 -0600, Lofton Henderson wrote:
>[...]
>-- we need to sort out the "Class" for almost all of them.  This is 
>sometimes an issue, as we think there is some unclarity in the Process 
>document regarding how to classify an erratum.

And in the agenda...

At 04:07 PM 8/29/2007 -0600, Lofton Henderson wrote:
>[...]
>         - review for "Class" [2], [3]
>[2] http://www.w3.org/2005/10/Process-20051014/tr.html#correction-classes
>[3] http://lists.w3.org/Archives/Public/public-webcgm-wg/2007Aug/0018.html

In past discussions, we have talked about whether something is Class 2 or 
Class 3.  In a very strict reading of reference [2], almost anything could 
be considered Class 3, "may affect conformance."   (Btw, in the particular 
Process document usage of the referenced sentence, is "may" used in the 
RFC2022 sense?  Or is it just normal english language usage?)

A more pragmatic interpretation of Class 2 versus Class 3 would take into 
account the practical question, "is any good-faith implementation likely to 
misinterpret the standard and need change as a result of the erratum?"

The difference between class 2 and class 3 in W3C:  Class 2 does not 
require any consideration other than within the WG, whereas Class 3 seems 
to require that some sort of public review occur (minimum 30 days).  IMO, 
the latter is not a big deal for 1.0 errata, as we are not in a big hurry, 
and we only have to be concerned with W3C process.

For 2.0 errata, we need to keep in mind that the CGMO/OASIS errata process 
must be executed as well, and ideally should be somewhat coordinated with 
the W3C process.

Regards,
-Lofton.

Received on Thursday, 30 August 2007 14:42:36 UTC