Walter wrote: (or should it be Mr. Perry?):

> OTOH, from the client, or user's, point of view there is little here
> the HTML-viewed-in-a-browser paradigm does not offer.

I agree. Although just to emphasise your next point before dealing with

> The (admittedly enormous) labor savings in using XML to build
> these indices benefit only the server side.

This should not be underestimated. The evolution of our tools is
interesting. We developed our object-approach to relational data that I
described in the last message, actually before we discovered XML. Yet we
still found that we had loads and loads of ASP files, all looping
through our database, getting out the children of an object to display
them, etc. I feel it's a neat solution, but still (relatively) difficult
to maintain (if you are interested, see
for a site we did in which every single element of the site is coming
from our database).

Our new technique - part of which merges XML and XSL on the server -
involves about five stylesheets! That's it! We recently changed the
entire look-and-feel of the magazine we are working on, in an afternoon.
If you look at the MGB site we did before (URL above) it has about 30
ASP pages (still better than the old days) and about five CSS
stylesheets. But we still could not change the structure of pages
without a lot of work, and every developer seems to want to get the data
out of the database in a slightly different way, for the particular page
they are working on!

However, you know all that, and we agree, so on to your other points:

> My impatience with this is because Mr. Birbeck has the core of a truly
> distributed database and then settles for a static visual presentation
> of HTML to the client.

I agree again, but we should not underestimate the problems of
transition, and what we have achieved already. By transition, I mean
that everyone is still viewing with this old paradigm, so we have to
take them with us.

But more importantly, on what we have achieved already: we can get any
node of our database out as if it was an XML document. We can also -
albeit in a rudimentary way - select different sets of nodes using URLs
to our server that look like XSL. For example:[number=55]/article

gives us all articles for issue number 55. Not telling how we did that,
just yet!!, but I feel that this technique complements the XML-RPC work
that others are doing quite nicely; it is more compact for certain
situations, and actually fits the XSL proposal of using URI-style
references for data.

Now, for me, that development is far more important that what we then do
with the XML, since, to be honest, that is what everyone else is working
on anyway. At the moment we just happen to merge it on the server, but
when the rest of the world is ready for us ;) they can hook into those
URLs. The portals I mentioned in the last email could go for:[end()]/article[0]/title

to get the lead article; someone quoting an article could put a link to:[number=55]/article[title=R

and Mr. Perry, with his parser in his browser could go for:[number=55]/article[title=R

and display it how he wants. We are therefore already building the hooks
for Walter's ideas:

> Client-side, I could
> create complex data structures which incorporated, by database
> the data elements which Mr. Birbeck is already maintaining, together
> data elements of my own or data elements from other sites like Mr.
> Birbeck's, which he and his data know nothing of. To fill my data
> Mr. Birbeck would need only transmit a stream of XML equivalent to
what he
> already provides to server-side portals.

In fact the magazine we are putting online has statistics such as
population and literacy rates for countries, turnover and market cap for
companies, previous positions held for CEOs, when a head of state came
to power, etc., etc. All data I am sure you will want to have fun with -
all available in XML. You will be able to represent as figures, pie
chart, bar graph or musical notes, the following:[name=USA]/economic/GDP

However, before I get carried away with my own hype, there is still a
lot of work to do.

Again, I must emphasise that we must bring our 'static' page viewers
with us. They want the information as much as you, Walter. I share your
'impatience', and sympathise. All we need to do is have some simple,
predefined templates for presentation of the XML, for those who do not
want to format the XML themselves. The 'origin' of that XML is as
described above.

And for those who do want to play with the XML themselves, we still need
to work out how we tell you what information there is on our server. The
XSL sequences above all look great for those who like XSL, but how do
you know that a quote is a 'pullquote' and not an 'articlequote'? How do
you know that GDP is even available? We are some way towards solving
this, but I'm just pointing out that there are many problems to solve

(Not sure if we have gone off track there!)



Mark Birbeck
Managing Director
Intra Extra Digital Ltd.
39 Whitfield Street
t: 0171 681 4135
e: [log in to unmask]