Re: Starting an issues list?

[Stephen Green <stephen.green@documentengineeringservices.com>]

Or I'm considering changing it to a better schema expression

	<xs:complexType name="sourceDocument_type">
		<xs:simpleContent>
			<xs:extension base="xs:normalizedString">
				<xs:attribute name="revision" type="xs:normalizedString"/>
				<xs:attribute name="version" type="xs:normalizedString"/>
				<xs:anyAttribute namespace="##any" processContents="skip"/>
			</xs:extension>
		</xs:simpleContent>
	</xs:complexType>

to avoid the mixed content, in which case the spec will read

<sourceDocument
  revision ? = 'xsd:normalizedString'
  version ? = 'xsd:normalizedString'
{any attributes with non-schema namespace . . .}>
 Content: xsd:string
</sourceDocument>

This seems better and improves the mapping to the model.
	
Best regards

Steve
---
Stephen D Green




On 4 February 2010 17:36, Stephen Green
<stephen.green@documentengineeringservices.com> wrote:
> I'd also like to make another minor correction where,
> debatably, a discrepancy exists between schema and
> XML Representation for sourceDocument (the last
> element to receive a mod and so the most likely to have
> such an error). I'd like to change
>
> <sourceDocument
>  revision ? = 'xsd:normalizedString'
>  version ? = 'xsd:normalizedString'  {any attributes with non-schema
> namespace . . .}>
> Content: xsd:string
>  {this element also allows mixed content so text can be included
> directly as part of the main element's content}
> </sourceDocument>
>
> to
>
> <sourceDocument
>  revision ? = 'xsd:normalizedString'
>  version ? = 'xsd:normalizedString'  {any attributes with non-schema
> namespace . . .}>
>  {this element also allows mixed content so text can be included
> directly as part of the main element's content}
> </sourceDocument>
>
> by just removing "Content: xsd:string" as it is the
> mixed content which allows the text to be added,
> not an XSD 'base=xs:string' - as the schema has:
>
> <xs:complexType name="sourceDocument_type" mixed="true">
>                <xs:attribute name="revision" type="xs:normalizedString"/>
>                <xs:attribute name="version" type="xs:normalizedString"/>
>                <xs:anyAttribute namespace="##any" processContents="skip"/>
> </xs:complexType>
>
> Best regards
>
> Steve
> ---
> Stephen D Green
>
>
>
>
> On 4 February 2010 17:28, Stephen Green
> <stephen.green@documentengineeringservices.com> wrote:
>> All I will do with these non-substantive changes
>> to the schema (separate and inline) is change
>> those places where we have
>>
>>        <xs:complexType name="***Shared_type" mixed="true">
>>                <xs:simpleContent>
>>                        <xs:extension base="***_type">
>>                                <xs:attribute name="conflict" type="***ConflictCode_type"/>
>>                        </xs:extension>
>>                </xs:simpleContent>
>>        </xs:complexType>
>>
>>
>> to
>>
>>        <xs:complexType name="***Shared_type">
>>                <xs:simpleContent>
>>                        <xs:extension base="***_type">
>>                                <xs:attribute name="conflict" type="***ConflictCode_type"/>
>>                        </xs:extension>
>>                </xs:simpleContent>
>>        </xs:complexType>
>>
>> (where '***' is 'predicate', 'target', 'var', etc)
>>
>> This change has little effect - it only improves the
>> clarity of the schema and the fidelity to the spec
>> representation. The 'mixed=true' is just ignored
>> by schema processors/parsers. Just as well to
>> remove it though I think.
>>
>> Best regards
>>
>> Steve
>> ---
>> Stephen D Green
>>
>>
>>
>>
>> On 4 February 2010 16:06, Stephen Green
>> <stephen.green@documentengineeringservices.com> wrote:
>>>> I passed it
>>>> through the W3C online schema validator and found I got
>>>> a few warnings because there are places where simpleContent
>>>> is specified as having mixed content (mixed content when
>>>> there are no child elements, I think that means).
>>>
>>> OK, I found the causes of these 'warnings' and could correct
>>> them easily enough but that would warrant another draft of just
>>> the markup spec and the schema. I think it's worth this very
>>> minor change just to avoid misunderstanding of the important
>>> schema, even though it does not affect instances.
>>>
>>> I'll send out new drafts of the schema and the spec in a few
>>> hours time.
>>>
>>> Best regards
>>>
>>> Steve
>>> ---
>>> Stephen D Green
>>>
>>>
>>>
>>> On 4 February 2010 15:23, Stephen Green
>>> <stephen.green@documentengineeringservices.com> wrote:
>>>>
>>>> Dear TAG TC,
>>>> I guess we will need an issues list. Any progress/thoughts
>>>> about having and using a Jira account?
>>>> I'm not sure how our TC comments will need to be separated
>>>> from external/public comments. I will have a few comments
>>>> of my own on how things might be improved after public review.
>>>> After this month I may have to submit comments via the
>>>> public comments list. Are we supposed to keep TC comments
>>>> separate and try to make such comments before the review
>>>> rather than during it?
>>>> My comment for now is that I think there would be special
>>>> benefit in having the schema itself reviewed. I passed it
>>>> through the W3C online schema validator and found I got
>>>> a few warnings because there are places where simpleContent
>>>> is specified as having mixed content (mixed content when
>>>> there are no child elements, I think that means). Some
>>>> review of this and any other related schema design features
>>>> might be warranted and might require minor changes to the
>>>> model as well as the schema (but these should not, I hope,
>>>> affect XML instances).
>>>> Best regards
>>>>
>>>> Steve
>>>> ---
>>>> Stephen D Green
>>>>
>>>
>>
>