W3C home > Mailing lists > Public > public-html@w3.org > January to March 2007

Re: declarative expressons in WF2

From: David Dailey <david.dailey@sru.edu>
Date: Mon, 26 Mar 2007 12:13:28 -0400
Message-Id: <6.2.5.6.0.20070326113217.01e23ec0@sru.edu>
To: James Graham <jg307@cam.ac.uk>,Dave Raggett <dsr@w3.org>
Cc: public-html@w3.org

At 10:50 AM 3/26/2007, James Graham wrote:

>Ignoring the issue of whether a declarative syntax actually delivers 
>on the claimed ease of use, I don't understand how is this target 
>population of authors is supposed to develop the server-side 
>component of their web-app? They could, of course, get a third party 
>to implement that, but then why not also have that third party 
>develop the client-side? I feel like I must be missing something 
>trivial because my interpretation is that the proposal is aimed at a 
>user base that does not exist.

A good question and a fair one, I think. Google's Dec. 2005 analysis 
(http://code.google.com/webstats/) of about a billion web pages 
revealed that about half contained a <script> tag. . It would be 
interesting to know how many of those are purely "presentational" in 
the sense of having no server-side return visits. If a page has 
<form> but no <script> then we may suspect (indirectly) that the page 
merely gathers info and ships it via submit to a server-side script 
(perhaps written by somebody else?). Google's data doesn't present 
that sort of cross-tabulation, though I'll bet if we asked real nice, 
they might generate such a report. I've certainly seen things like 
that -- my students do it from time to time. I'd guess between 2% and 
5% of the scriptless sites might fall into the user base, but that's 
just an out of the hat figure.

But it's not just the non-scripting public who "might" benefit. 
Side-by-side comparisons of the amount of code a big organization 
would need to build and maintain, would tend to suggest that nifty 
ways of doing this sans script could save some time and effort as 
well (subject to the constraint that the expertise base may be 
small). In the grand scheme of things, I suppose we'd need some real 
data on the efficiency of producing and maintaining code in a 
declarative environment. I went poking about the web but didn't find 
anything to help resolve the issue objectively.

In the arena of declarative animation, I can point to things that 
take 1000 bytes of code in a declarative environment that would 
require a minimum of five times that much script to make things 
happen. That's a bit different than declarative approaches to simple 
form elements, but if we consider, among future form elements, 
declarative controls attached to sliders, color pickers, scrollbars, 
sketchpads, graph-editors, levers, pulleys and the like, reduction of 
script could be significant.

regards,
David Dailey 
Received on Monday, 26 March 2007 16:17:04 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 26 March 2007 16:17:06 GMT