FW: Possible problem with TC46 revising its XPS spec to track 29500:2011 (or 29500:2012, for that matter)

[John Haug]

Forwarding to WG 4 as FYI for a potential discussion topic.

Apparently there are different schools of thought on namespace versioning.  Agree with Makoto - something for WG 4 to discuss.  At least we can be aware of the concern from TC46.

John

-----Original Message-----
From: eb2mmrt at gmail.com [mailto:eb2mmrt at gmail.com] On Behalf Of MURATA Makoto
Sent: Wednesday, May 16, 2012 3:13 PM
To: John Haug
Cc: MURATA Makoto (FAMILY Given)
Subject: Re: FW: Possible problem with TC46 revising its XPS spec to track 29500:2011 (or 29500:2012, for that matter)

John,

It appears that the MS implementation is a single program that has many conditional statements for namespace checking.  Then, can is it possible to use a version attribute instead?

Historically, there have been a lot of discussions about namespace changes.
Some people argue that whenever an element or attribute is added or removed, a new namespace should be introduced.  Others are like me: namespaces should not change unless almost every application will inevitably crash.
I would argue that 29500-2 should continue to use the same namespace even if a few auxiliary elems or atts are added.

I think that this issue should be discussed in WG4.

Regards,
Makoto

2012/5/17 John Haug <johnhaug at exchange.microsoft.com>:
> Not sure what you mean by same implementation.  My understanding is that work was done to the apps for each change to the standard.  We increment the date part of the Microsoft-specific URIs as we change their contents.
>
> -----Original Message-----
> From: eb2mmrt at gmail.com [mailto:eb2mmrt at gmail.com] On Behalf Of MURATA 
> Makoto
> Sent: Tuesday, May 15, 2012 5:54 PM
> To: John Haug
> Cc: MURATA Makoto (FAMILY Given)
> Subject: Re: FW: Possible problem with TC46 revising its XPS spec to 
> track 29500:2011 (or 29500:2012, for that matter)
>
> John,
>
> I did not know the removal of contentType elements from OPC.  I read  Appendix I of the first edition of ISO/IEC 29500-2.  Another change is the addition of value elements.
>
> Should we use a new namespace if an element is dropped or added?
> No, addition of elements do not necessarily need new namespaces.
> (Consider the HTML family)  IMHO, new namespaces are needed when most of the existing applications are expected to behave totally incorrectly.
>
> Has Microsoft used the same implementation for pre-29500 OOXML,
> pre-29500 OPS, post-29500 OOXML, and post-29500 XPS?  If this is the case,  I think that we should continue to use the same namespace.
>
> Regards,
> Makoto
>
>
> 2012/5/16 John Haug <johnhaug at exchange.microsoft.com>:
>> Hi Makoto -
>>
>> Rex brings up a good thing to consider as a COR for Part 2.  It seems 
>> to me that any namespace in 29500 would need to be updated if 
>> anything defined in that namespace changes. Does that sound right to 
>> you?  If so, we probably needed namespace updates for each reprinted 
>> edition,right?  I think the only namespace changes in 29500 were around the Strict vs. Transitional breakup.
>>
>> John
>>
>> ________________________________
>> From: Rex Jaeschke [rex at RexJaeschke.com]
>> Sent: Saturday, May 12, 2012 2:01 PM
>> To: John Haug
>> Cc: Jim Thatcher
>> Subject: Possible problem with TC46 revising its XPS spec to track
>> 29500:2011 (or 29500:2012, for that matter)
>>
>> John, a few days ago, I was reviewing some very old TC46-related 
>> notes and found the following "Deferred" issue in TC46's issues list:
>>
>> "IS 29500 Part 2 (OPC) removed element contentType element from the 
>> Core Properties Part schema, but did not change the namespace for 
>> that schema. As a result, there is no way to distinguish between
>> ECMA-376:2006 and IS 29500 w.r.t Part 2. So, at its Geneva meeting,
>> TC46 decided to refer to
>> ECMA-376-2006 instead of IS 29500.
>>
>> The status of IS 29500, especially w.r.t changes in the schema 
>> namepaces to indicate breaking changes (possibly as a result of the 
>> Defect Report filed by CH) should be monitored, so the XPS spec can 
>> be changed to point to IS
>> 29500 at some future point."
>>
>> For some time now, I've been nudging TC46 in the direction of 
>> pointing to the OPC and MCE specs from 29500:2011., so we (Ecma via
>> TC45) can legitimately claim in WG4 that there are other, important users of 29500.
>> However, as I read this issue now it seems to me that nothing has 
>> changed in that respect, which may well mean that TC46 will not be 
>> able to follow my suggestion. If that is true, we'd need an amendment 
>> rather than a COR (or a
>> revision) of the OPC to provide some distinguishing facility. Do you concur?
>>
>>
>>
>> Rex
>>