[ba-ohs-talk] Web engines with XML and JavaScript
I am beginning to notice the emergence of Web engines based almost purely
on scripting languages rather than hard-wired core infrastructure. I have
discovered two such projects (thus far) (01)
The Tamber Project http://tamber.com/index.asp (LGPL)
"The Tamber project is a free, componentised n-tier website engine that
uses open languages such as XML and JavaScript. In a nutshell, content is
stored in separate XML files, in databases, or other data objects; business
functions are carried out by JavaScript and ASP; and presentation is
controlled by an XSL transformation, which allows for delivery over
multiple channels such as HTML, WAP and MHEG. Currently Tamber contains
modules that support; automatic locale detection, search engine
optimisation, e-commerce shopping carts, secure sign in, data access and
conversion services and advanced session management" (02)
and (03)
Whitebeam http://www.whitebeam.org/
(http://www.whitebeam.org/RPL/RPL-1.0.rhtm license which looks like LGPL
(maybe))
"[Whitebeam] is an Open Source project that provides a complete rapid
development environment for Web applications allowing sophisticated, robust
and secure applications to be built using only XML and server-side JavaScript.
The Whitebeam architecture integrates two popular and robust Open Source
platforms - Apache and Mozilla's SpiderMonkey JavaScript engine - with it's
own powerful XML environment and security architecture. The XML environment
provides both event driven and DOM based manipulation of XML trees along
with the powerful XPath querying language - all controlled via simple
server-side JavaScript interface.
Providing an extensible XML based interface to back end objects on any
platform, the client model securely segregates thousands of customer
applications within a single server." (04)
My comments:
I'm not sure what to make of this, but it makes some sense to me. What I
am learning from the NexistWiki experience is that, in order to make things
scalable and evolvable, the more functionality you can move out of the core
engine and into a plug in architecture, the better. (05)
Thus, Version 2 of NexistWiki will dump all the Extreme Programming RAD
code and move to pure XML-XSLT. The only code needed (I think) to support
that is enough code to convert database queries into the XML string ready
to paint by way of XSLT. It's not entirely out of the question that
something along the lines of Whitebeam or Tamber could lead the way to such
a remake of NexistWiki. (06)
Comments always welcome.
Jack (07)
---------------------------------------------------------------------------
XML Topic Maps: Creating and Using Topic Maps for the Web.
Addison-Wesley, ISBN 0-201-74960-2. (08)
NexistWiki
http://www.nexist.org/wiki/ (09)