> calvin lash wrote of
> > deploying [mobile] content on multiple browsers and protocols.
> > It is like Bart Simpson always having to write
> > something 500 times on a blackboard:
> > "This punishment is not boring and pointless."
Heimo Laukkanen answered:
> In this point I praise my previous employer for giving me the knowledge
> of SGML, XML and storing data.
>
> How Christ's sake, it is no news to publish from one source to multiple
> templates in different medias.
SGML in Japan? It's going to be a while. It's hell enough here in the
tropical Japanese heat shoveling the coal into my steam-driven
Difference Engine while climbing up the ladders to keep the mechanical
switches well-oiled up on the 5th story of the banks of cams pushing the
data pins meshing with the Jacquard cards. ;-)
SGML is so far beyond the minds of mere mortals that XML had to be
invented. I see XML as the best hope and it fits well with the present
implimentation of the iMode (and less so with WAP). Unfortunately,
Japanese corporate practices center around content delivered to a web
staff locked down in PDF format, .doc files and paper hardcopy.
Think about it: When is the last time you saw a Japanese website* or
database with proper metadata. Sites here neither have HTML <meta> tags
to provide the Keywords and Description nor have any RDF/XML, suitable
for embedding in the <head>...</head> section of a page. The entire
concept of separating Format from Content is alien here. I am not even
going to try to talk about Schemata in the meta-info since
implementation is so problematic since no one agrees to what a Schema
should be.
Obviously, in the realm of Keitai I can't find any thought concern for
imbedding meta-information. The iMode specs only provide one <META>
(tag) [Attribute] charset=SHIFT_JIS but nothing about keywords or
content description. XML is the best way to manage a big site --- I
wish using XML with standard web tools was easier and that a broader
range of browsers would handle XML better.
GETTING back to Keitai....it's not that easy to maintain the mapping
between an XML document and all of the derived documents in different
formats.
Ya gotta produce multiple small data pieces from the original XML file.
A one-to-one mapping between HTML documents and iMode/WML ones just
ain't smart-----each HTML page is just so much larger the a mobile
card/page. The obvious solution is to split the content into smaller
linked decks that can be downloaded separately. Whether it's
server-side or client-side manipulation and the XML DOM methods to
dynamically isolate an XML subtree and apply transformation to it. Then
build links to subparts of the content..........
"M Y B R A I N H U R T S !" to quote to wisdom of my mentors, The
Gumbys, the handkerchief-bedecked, suspender-clad, thick-headed
characters of Monty Python.
<!--
calvino <calvin-lash@hitachi-cs.co.jp>
Hitachi Communication Systems Inc., TPS
-- translation/localization since 1961
-->
Footnote:
*The quickest way to glaze over a Japanese customer's eyes is to mention
the need for marking up their documents with meta-information. No one in
management has the foggiest idea of the importance of SGML, XML and
storing content in a rational manner. Case in point, none of the
following major Japan corporations have any META information for
Keywords and Description in their HTML sites:
http://www.hitachi.co.jp
http://www.toshiba.co.jp
http://www.sony.co.jp/index_ie.html
http://www.nec.co.jp
http://www.ntt.co.jp
http://www.nttdocomo.co.jp/
http://www.nttdocomo.co.jp/top.html
Received on Mon Jul 17 09:50:55 2000