[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Tags (searching)



Greg Ferguson wrote:
> 

> 
> Why must we go beyond that? Other than stating the deprecated tags,
> and making sure they are not used (which I wholeheartedly agree with),
> the template should *guide* the author as to what should be used,
> and how it should be used (date and version number formats
> come to mind). I do not believe it should not limit the use of
> any non-deprecated tags (imo) that are accepted in the DTD.
> 
You guys are sliding over this too fast and easy. Tags mean
different things depending on what they're embedded in. For
instance, consider the following example:

<Article>

<ArtHeader>

<Title></Title>

<Author>
<FirstName></FirstName>
<Surname></Surname>

</Author>

<RevHistory>

<Revision>
<RevNumber></RevNumber>
<Date></Date>
</Revision>

</RevHistory>
<Abstract>

<Para></Para>

</Abstract>

</ArtHeader>

<Sect1>

<Title></Title>

<Para></Para>

</Sect1>

<Bibliography>

<BiblioEntry>

<CiteTitle><Emphasis></Emphasis></CiteTitle>

<Author>
<FirstName></FirstName>
<Surname></Surname>
</Author>

<PubDate></PubDate>

<Publisher>
<PublisherName></PublisherName>
</Publisher>

</BiblioEntry>

</Bibliography>

</Article>

Author appears twice. The first time it's in the article
header, and it means the author of the howto. The second
time it appears in a biblioentry and it means the author of
a book or journal article (BTW, DocBook does not have a
Journal tag, so I don't see a way to cite journal articles).
Likewise, Title appears twice. The first time in artheader,
where it's the title of the howto, and the second time where
it's a section title. PubDate is arguably best used in a
biblioentry, where revhistory specifically tags revnumber,
date, author initials, and revcomments, and arguably belongs
in the article header, where the information can be
extracted by a search engine that uses tags rather than
content format to separate revision number and date. I
thought the purpose of using SGML was to use the structuring
to make data available for search or extraction. OASIS means
something.

You guys are glossing over a difficult problem. You're
making it harder on yourselves by proposing a laisse faire
approach to using the tools while expecting them to perform
as if they were used by experts. If you look at the OMF
search engine, the categories you can search for obviously
have meanings that are set by tag context. Maybe we should
get a comment from Paul about what the OMF search engine is
really looking for.

Gary


--  
To UNSUBSCRIBE, email to ldp-discuss-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org