OASIS Open Document Format for Office Applications (OpenDocument) TC

  • 1.  Suggestion from the ISO meeting in Seattle

    Posted 09-24-2009 11:50 
    Greetings!

During the ISO discussions in Seattle last week the following suggestion 
was made for cases where the standard does not or cannot define precise 
behavior.

> From ISO 9075-1:2008 SQL/Framework:
>
> * *
>
> *7.2 Implementation-defined elements*
>
> Every part of ISO/IEC 9075 contains an Annex that lists every element 
> of SQL and its processing that is specified
>
> in that part, and is permitted to differ between SQL-implementations, 
> but is required to be specified by
>
> the implementor for each particular SQL-implementation.
>
> * *
>
> *7.3 Implementation-dependent elements*
>
> Every part of ISO/IEC 9075 contains an Annex that lists every element 
> of SQL and its processing that is mentioned,
>
> but not specified in that part, and is thus permitted to differ 
> between SQL-implementations, but is not
>
> required to be specified by the implementor for any particular 
> SQL-implementation.
This came up in the context of discussions about corrections to ISO 
26300. Particularly areas where in retrospect we were not completely 
precise and to offer more precision now might create incompatibilities 
with existing implementations.

I think it is an interesting suggestion and one that could prove to be 
useful.

Hope everyone is having a great week!

Patrick

-- 
Patrick Durusau
patrick@durusau.net
Chair, V1 - US TAG to JTC 1/SC 34
Convener, JTC 1/SC 34/WG 3 (Topic Maps)
Editor, OpenDocument Format TC (OASIS), Project Editor ISO/IEC 26300
Co-Editor, ISO/IEC 13250-1, 13250-5 (Topic Maps)


  • 2.  Re: [office] Suggestion from the ISO meeting in Seattle

    Posted 09-24-2009 12:55 
    Patrick,

I don't think we can mark something as "implementation-defined" in 
Approved Errata, since that would be adding a new requirement (a 
requirement for the implementation to specify its behavior) on existing 
implementations, something we are not permitted to do under OASIS rules. 

However, "implementation-dependent" would be fine. 

Of course, in ODF 1.2 we are free to say "implementation-defined" if we 
want.

-Rob


Patrick Durusau


  • 3.  RE: [office] Suggestion from the ISO meeting in Seattle

    Posted 09-24-2009 17:39 
    I concur with Rob's analysis with regard to approved errata.

Going forward, I recall seeing ISO annexes that provided cross-referenced
enumerations of 

  * all implementation-defined items (and these might include matters of
SHOULD and MAY)

  * all conformance statements with identification of the party responsible
for satisfaction of the condition - e.g., format, consumer, producer, and
other cases.

The matter of responsibility might apply to implementaton-defined too,
particularly with respect to SHOULD and MAY, but I am not sure I've seen
that covered.  I don't think CAN has to be dealt with because it is a
statement of fact, not permission or requirement, as far as I can tell.

The OASIS recommendations for conformance provisions in specifications
suggest that something along these lines is also desirable.

 - Dennis


    
                                                
    
					
                                                    

        
                                                
				
    
                                                
    
                                                    
                                                        
                                                        Original Message


  • 4.  Re: [office] Suggestion from the ISO meeting in Seattle

    Posted 09-24-2009 23:24 
    Rob,

robert_weir@us.ibm.com wrote:
> Patrick,
>
> I don't think we can mark something as "implementation-defined" in 
> Approved Errata, since that would be adding a new requirement (a 
> requirement for the implementation to specify its behavior) on existing 
> implementations, something we are not permitted to do under OASIS rules. 
>
> However, "implementation-dependent" would be fine. 
>
> Of course, in ODF 1.2 we are free to say "implementation-defined" if we 
> want.
>
>   
Good point!

I was thinking more of a good future practice as I am sure there are 
going to be things we are better off to simply require applications to 
define what they do, rather than our trying to wade into the weeds to 
sort some problems out.

Hope you are having a great day!

Patrick
> -Rob
>
>
> Patrick Durusau


Related Content