RE: Bobby Applications (DTD comments)

I should apologize for not having posted a follow-up to explain this
document that I posted. This document is still in draft stage, it's part our
implementation of the Techniques document; as the techniques evolve, so does
this, including the document structure. I posted it for discussion in the
Monday teleconference just to contextualize some things we were talking
about; I'll make a point to make that purpose clear next time I post a
version. I haven't put time into making the DTD completely correct - just
sufficient to get the document to display (in IE5). That said, your
suggestions are helpful to me as a relative XML neophyte (we inherited the
file from Josh who knew what he was doing more than the rest of us) and, as
we finalize the document in the next month or two, I'll make sure to take a
closer look at those structural aspects. As far as the specific questions
you raised in the revised DTD, they relate to aspects of our implementation
that are still evolving and it might be more helpful to explain, and get
feedback on, those items a little bit later. Michael

> -----Original Message-----
> From: w3c-wai-er-ig-request@w3.org
> [mailto:w3c-wai-er-ig-request@w3.org]On Behalf Of Harvey Bingham
> Sent: Tuesday, August 10, 1999 9:00 PM
> To: Michael Cooper; w3c-wai-er-ig@w3.org
> Subject: Re: Bobby Applications (DTD comments)
>
>
> At 1999-08-09 09:06 AM, Michael Cooper wrote:
>
> >Here they are - the XML file and DTD file must be in the same
> folder, then
> >open the XML file in an XML browser such as IE5. Alternatively,
> view the XML
> >file in a text processor, somewhat less readable but it should be ok.
> >Michael
>
> Recall that names of elements and attributes in XML are case-sensitive.
> The DTD uses lowercase for these, a common practice.
>    waigl1990505r19990807.dtd
>
> The XML sample of CAST use has those names as uppercase.
>    waigl1990505r19990807.dtd
>
> Some of the textual commentary uses uppercase names as well.
>
> The DTD needs a DOCTYPE. I add it below, with representative content.
> It includes the name of the root element: guidelines, the reserved word
> SYSTEM, and the SystemLiteral:
>
>     <!DOCTYPE guidelines SYSTEM "waigl1990505r19990807.dtd">
>
> The SystemLiteral is the URI of the local system version of the DTD,
> in the same node of the file system as the documents that are
> tagged to it.
> More generally the PublicLiteral is announced by the reserved word PUBLIC,
> and provides the URI where this DTD is publicly available (but
> unregistered).
> The PublicLiteral may have following it the SystemLiteral without
> the need
> to include the SYSTEM, as indicated below:
>
>     <!DOCTYPE guidelines PUBLIC
>          "-//CAST//DTD Bobby Documentation of W3C Web Accessibility
>          Guidelines 19990505//EN"
>         "waigl1990505r19990807.dtd">
>
> Whitespace is normalized to a single space if a separator, else omitted,
> in the PublicLiteral.
>
> The DTD has ID attributes, appropriate as targets for references,
> but no way to reference them within the same document via any
> attributes with IDREF. Likewise, there is no way to reference external
> URLs, such as in the WAI guidelines/checkpoints themselves.
>
> I'd like a comment <!-- ... --> included before each element indicating
> its intended semantics. There is no use of aprompt in the sample, and its
> semantics are unclear to me.
>
> Here is the DTD, with my augmentations and questions in comments:
> ----
>
> <?xml version="1.0"?>
> <!DOCTYPE guidelines PUBLIC
>     "-//CAST//DTD Bobby Documentation of W3C Web Accessibility Guidelines
>         19990505//EN"
>     "waigl1990505r19990807.dtd">
>
> <!ELEMENT guidelines (guideline+)>
>
> <!-- Is there a meaningful guideline without a checkpoint? -->
> <!ELEMENT guideline (checkpoint*)>
> <!ATTLIST guideline id CDATA #REQUIRED
>               title CDATA #REQUIRED>
>
> <!-- Does any checkpoint have no technique for realizing it? -->
> <!ELEMENT checkpoint (technique*)>
> <!ATTLIST checkpoint id ID #REQUIRED
>              title CDATA #REQUIRED >
>
> <!--A technique without application has little meaning. It has no
> attributes,
> so it can only be a placeholder marker point. -->
> <!ELEMENT technique (application*)>
>
> <!-- What is "partial" conformance? Can partial conformance still get a
> rating? -->
> <!ATTLIST technique id CDATA #REQUIRED
>              title CDATA #REQUIRED
>              conformance (full|partial|none) #REQUIRED
>              rating (p1|p2|p3|none) #REQUIRED >
>
> <!ELEMENT application (ease,rationale,aprompt*)>
> <!ATTLIST application id CDATA #REQUIRED
>              headline CDATA #REQUIRED>
>
> <!ELEMENT ease EMPTY>
> <!ATTLIST ease
>               easetofix (easy|hard|medium) #REQUIRED>
>
> <!--I believe that there is much more structure under rationale
> than just #PCDATA. Examples have paragraphs and lists, possibly
> any of the HTML stuff?
> -->
> <!ELEMENT rationale (#PCDATA)>
>
> <!--What are the aprompts for? Do they need attributes?-->
> <!ELEMENT aprompt (#PCDATA)>
> ~
>
>
>
> >
> >Michael Cooper
> >CAST, Inc.
> >39 Cross St.
> >Peabody, MA  01960
> >Tel 978-531-8555 x265
> >TTY 978-538-3110
> >Fax 978-531-0192
> >Email mcooper@cast.org
> >http://www.cast.org/
> >http://www.cast.org/bobby/
> >
>
> Regards/Harvey Bingham
> Tel 781-862-6908
> Email hbingham@acm.org
>

Received on Thursday, 12 August 1999 10:32:15 UTC