TAPRegExt Erratum

[Paul Harrison]

On 2014-10 -07, at 02:00, Markus Demleitner <msdemlei at ari.uni-heidelberg.de> wrote:

> Brian,
> 
> On Mon, Oct 06, 2014 at 05:43:44PM -0600, Brian Major wrote:
>> One thing I noticed about this guinea pig is that the status of the note
>> is not captured in the document.  As per Francoise's draft on the
>> proposed procedure, the note goes from "Draft Proposed Note", to
>> "Proposed Endorsed Note", then to "Endorsed Note".  As it stands,
> 
> Yeah -- as I said I'd have to update ivoatex to support the various
> new types, which I'd rather do only if we are reasonably sure this is
> where we want to go; imagine the status is "Draft Erratum" for the
> time being.
> 
>> On a similar note, having a section named "Proposed Change"
>> would not make as much sense once the note is endorsed, and
> 
> True.
> 
>> could be interpreted as something yet to be accepted.  I wonder if
>> language could be chosen that would make sense in all three
>> statuses so that the document template need not change upon
>> status change.
> 
> Well, turns out I'm wondering, too.  I've tentatively made the
> section header "Erratum Content" now.  Improvement ideas are welcome.
> 


Hi,

I have not been involved in the discussions on errata at Interop meetings recently, but it seems to me that the whilst it would be useful to have “endorsed notes” that are of an explanatory nature, I believe that they should be a separate concern to errata. Is the proposal that the erratum note goes through the relatively fast track process and then it stands separately to the original standard document, with that never being revised? - if so that would not really serve the needs of all of the types of readers of the original standard. Sure the experienced implementor will have a concise document that explains the differences to the original standard with which they are familiar, but the new convert to the VO who reads the original standard for the first time will then be dismayed and possibly confused to discover that there might be a separate erratum published on the document and they have to read in a non-linear fashion.

I think that it would be better that errata were placed directly into the original text standard document, with the nature and reason for the erratum update being noted in a “Changes” section (as is common practice today for other upgrade style changes). In this case the document numbering could be altered to use a 3rd level - e.g. 1.0.1 identifies the 1st erratum revision of the document, and perhaps the more “lightweight” approval process could be used for any document that had this numbering. During this approval process the document could be known as a “Proposed Erratum Recommendation”, and once approved revert to a “Recommendation”.

It seems to me very “old fashioned” to be publishing errata as separate documents, as that mechanism was invented by publishers as a way of saving money on having to re-layout the type of the whole book, when today the “cost" of altering the original document is equivalent to creating the separate erratum document. 

Paul.