Re: [ARCH] two minor issues in current MMI-Arch spec

Hi,

I have been rechecking my findings and found another point that i would
like to comment on.

In contrast to what i wrote in my prior mail, not all .xsd files in Appendix
C
have attributeFormDefault and elementFormDefault set.

- mmi.xsd (Appendix C.1)

Neither has elementFormDefault = "qualified" on the schema element
nor form="qualified" on the mmi-element definition set. Since
"unqualified" is the default value all mmi elements are to be used
unqualified only.
This, as far as I know, makes most of the examples invalid as they
use the mmi element in a qualified way.

- mmi-datatypes.xsd (Appendix C.2)

Does not have attributeFormDefault nor elementFormDefault set. Which makes
it being set to the default value "unqualified". Types and Attributes
defined in this
xsd will have to be used without namespace prefix. But other .xsd documents
(e.g.
mmi-attribs.xsd (Appendix C.3)) reference them with prefix.


I am not a great expert on the XML Schema spec but I think someone should
look into
this. One thing I had to find out (while investigating my findings) was that
elementFormDefault and attributeFormDefault won't get inherited by child
documents
included in the parent documents.

Any comments are welcome!

Thanks & Regards,
Jakob.

2011/3/27 Jakob Sachse <jakobsa@gmail.com>

> Hallo,
>
> I have been reviewing the current specification. Even though it yet makes
> quite a mature impression I found some flaws that may should be revised.
>
> The .xsd files in Appendix C require a valid mmi-representation to have
> qualified attributes.
> ------------------------------------------------------------
> attributeFormDefault="qualified"
> elementFormDefault="qualified"
> ------------------------------------------------------------
>
> Given that it seems Appendix B is not in line with the spec.
> Examples there  (see exerpt below)  are not using qualified attributes.
> ------------------------------------------------------------
> <mmi:newContextRequest source="someURI" target="someOtherURI"
> requestID="request-1">
> </mmi:newContextRequest>
> ------------------------------------------------------------
>
> On the other side examples in Appendix E are valid to the xsd definition:
> ------------------------------------------------------------
> <mmi:startRequest mmi:requestID="1.237204761416E12"
> mmi:context="IM_dcc3c320-9e88-44fe-b91d-02bd02fba1e3"
> mmi:target="GUI">
>  <mmi:contentURL>login</mmi:contentURL>
>  <mmi:data>
>    <gui resourceid="login" xml:lang="de-DE">
>      <data id="back" enabled="false"/>
>      <data id="next" enabled="false"/>
>    </gui>
>  </mmi:data>
> </mmi:startRequest>
> ------------------------------------------------------------
>
>
> Another minor issue I came across was that in Appendix F.2 the spec states:
> ------------------------------------------------------------
> [...]The parameter timeout is optional and describes the maximum delay
> in milliseconds.[...]
> ------------------------------------------------------------
>
> Whereas the following sequence diagrams are using seconds as time unit.
> ------------------------------------------------------------
> timeout:
> e.g. 30 sec
> ------------------------------------------------------------
>
>
> That's what I came across so far. Overall it is fair to say that the
> working draft makes a good impression on me.
>
>
> Regards,
> Jakob Sachse.
>