RE: Server Side Magic.

At 03:13  20/06/01 -0700, Jay Zylstra wrote:
>... project (which we wrote and contributed to Apache), includes many of 
>the tools necessary to do this...

There have been many good reports of Cocoon and I am considering using it 
for teaching e-commerce students at the Australian National University. 
However, this would be for sophisticated e-business applications. The 
technology is probably unnecessary for ordinary web pages, as discussed below.

>As for this approach's suitability "for public web pages", I'm unsure what 
>is meant by the phrase...

By "for public web pages" I meant web pages designed for use by the general 
public, in contrast to those for a limited group, such as employees in a 
company, where you  have some idea of the type of network, hardware or 
software used.

>If your client considers alternate Web clients to be just PDAs or the new 
>Phones+PDA, which have rudimentary HTML browsers, then HTML for a small 
>screen can work great (provided that you use HTML 3.2 and no client-side 
>JavaScript, imagemaps, CSS, frames, etc.) ...

If a web site is for the general public, then you have to cater for low-end 
devices, low speed links and consider accessibility for the disabled. It 
happens that doing this will also allow the web pages to work on small 
screens of PDAs. My (limited) experience is that accessible web pages tend 
to work okay on quarter-VGA screens and non-accessible ones don't.

As an example your company's web site <http://datachannel.com/> appears to 
have some accessibility problems. If this is intended for the general 
public and the information is not provided in alternative forms, then it 
may be unlawful in Australia and in countries with similar legislation. If 
you added the accessibility features then it would probably also run on a 
wireless PDA without further modification.

>But when your client is more broad-minded and demands (or will demand) 
>support for a variety of Internet-enabled devices, such as WAP, 2-way 
>pagers (RIM Blackberry), VoiceXML, Web Services clients, smart Web 
>printers, television set-top boxes, and so forth...

WAP is dead as a business proposition. I doubt that WAP 2, WAP NG or 
M-Services (WAP with GPRS and some consumer glitter) is going to do any 
more than hasten the end for WAP.

I don't have experience with 2-way pagers or VoiceXML, but expect they 
suffer the same problems as all small display text-only devices. These have 
such small displays that it is necessary to supply a completely different 
service for them. Telling the consumer they are going to get the web on an 
eight line 40 character screen just creates disappointment.

I am not familiar with Web Services clients and smart Web printers, so 
can't comment on them.

Televisions have a similar resolution to a PDA screen, and the set top 
boxes have similar cut-down browsers, so an accessible web page tested on a 
PDA should also work on a set top box (but in some ways the interface is 
more like a mobile phone). I haven't actually tried a set-top box browser 
as I am waiting for roll-out of the Transact high speed network
<http://www.transact.com.au/about_transact/> here in Canberra for a smart 
apartment <http://www.tomw.net.au/2001/sa>.

However, I did try your web site using Bobby <http://www.cast.org/Bobby> 
set to impersonate "WebTV 2.5". The resulting web page didn't appear to 
have been adapted for a smaller screen or limited browser. None of the 
browser setting I tried seemed to make a difference to the delivered page.

>... then a disciplined separation of presentation and business logic is 
>vital to keeping the task manageable, and XML is frequently the ideal 
>technology for the job.

XML has potential for separation of presentation and business logic. 
However, we have to have applications sufficiently complex to justify its 
use, a reasonable expectation the end user can use it and also take into 
account the needs of the broad range of users.

Rushing into technically sophisticated applications which give 
disappointing user experience will just create more WAP-type failures. 
I-mode provides a useful contrast and seems to be doing reasonably well, 
despite (or because of) using a relatively crude cut down version of HTML.



Tom Worthington FACS tom.worthington@tomw.net.au Ph: 0419 496150
Director, Tomw Communications Pty Ltd ABN: 17 088 714 309
http://www.tomw.net.au PO Box 13, Belconnen ACT 2617
Visiting Fellow, Computer Science, Australian National University
Publications Director & Past President, Australian Computer Society
--------------------------------------------------------------------
Australia's IT Landscape, 6 July Coolum: http://www.tomw.net.au/2001/eal

Received on Thursday, 21 June 2001 20:14:45 UTC