bug-gne
[Top][All Lists]
Advanced

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

Re: [Bug-gnupedia] just HTML??


From: Hook
Subject: Re: [Bug-gnupedia] just HTML??
Date: Wed, 24 Jan 2001 17:15:43 +0800

Christopher Mahan wrote:
> > > This discussion of databases versus XML is
> > > ill-advised. Databases can not
> > > organize content in the same way as a markup
> > > language (how would you make
> > > things like paragraph breaks represented in a DB
> > > schema?)
> >
> >Easily! I'm not sure, but I have a feeling you can do
> >this in a mySQL database anyway. I'll look into it.
> >And even if you couldn't, as the database will have to
> >be parsed by a perl script, you just have to put in
> ><br> in for line breaks when you database it, and then
> >read those when you parse it again for display. Any
> >other markup you need is done in a similar way, and
> >interpreted by the perl scripts youd have to have with
> >or without an XML respository.
> >
> >
> > > However, things can be XMl and then stored in or
> > > indexed by a database. A
> > > good way to do this is to make sure the entire DB is
> > > reconstructable from the
> > > plain XML files, and use the DB for speed of access,
> > > searching, etc.
> >
> >Yes you could have a DB for catalogue individual XML
> >files. But as one DB would be faster, easier to
> >maintain and easier to construct (it is so much easier
> >to put a submission straight into a database, rather
> >than formatting it and saving it as an XML file), why
> >have XML files at all? Until somebody gives proper
> >proof of the advantages of the XML/DB idea (ie not
> >just attitude, a logical document or a demo) I'm not
> >at all convinced.
> >
> >Tom Chance
>
> In my opinion, the only two advantages the XML has over db records is that
> it can be edited offline with a text editor, and that it can be stored
> anywhere on the web, using a server that does not need scripting
capability.
>
> For example, I could host my page.xml document at tripod, and they would
be
> sent to the client as-is. This could be useful for people who are going to
> write javascript pages that are xml capable, and for non-browser
interfaces
> (such as an application able to retrieve html and xml pages off the web)
and
> SOAP-type systems.
>
> As another example, it could allow for storing of the different elements
of
> the article in mutiple places, on multiple servers, like:
>
> <article>
>    <article_id>UTX9839B</articleID>
>    <article_body>
>      <article_body_location>http://bodyserver.com/articles/UTX9839B.xml
>      </article_body_location>
>    </article_body>
>    <article_video_clips>
>      <video_clip>
>        <clip_id>884iyhr5562.mpeg</clip_id>
>        <clip_location>http://clipserver.com/clips/884iyhr5562.mpeg
>        </clip location>
>      </video_clip>
>    </article_video_clips>
>    <other_stuff>
>    ...
>    </other_stuff>
> </article>

But all this "either/or" argument for and against XML and databases is
misplaced.  It's *not* an exclusive thing - as far as I can see anyway.  You
want to index one article or a million, then you use a database.  You want
to define structure within an article, then you use XML.  Both tools are
useful.

Paul




reply via email to

[Prev in Thread] Current Thread [Next in Thread]