But these days of course it has to fetch the associated CSS files and add them -- but wait, there's more, as they say, it should also fetch any images called by the style sheet and add them to the total.
Which is when the question arose, what is the actual size of a web page?
Good question.
Each browser has its own strategy -- there are no 'golden rules' when implementing a browser. What this means for us is that you you can really only calculate a webpage's maximum download size, assuming it downloads all referenced elements whether they apply to the the page or not.
The HTTP standard says a UA should make a maximum of six simultaeneous requests to a single server, which you would have to take into account if trying to estimate the time. Also don't forget for a single page the elements needing to be loaded can exist on different servers.
If you also consider caching between requests then it quickly becomes apparent there's little you can do for accurately estimating download size and times.
The best solution, as in most cases, is to actually run some tests. Get your target user's UA set up and capture the http traffic for each UA. You'd also need to record the display of the UA if you wanted to understand the loading of the visual elements.
I know that for Gecko based browsers background images defined in stylesheets (inline or linked) are deferred from loading until they are asked to display themselves. This also helps because many style sheet rules containing images often don't apply to the elements of the current page.
-- tim lucas
http://www.toolmantim.com
*****************************************************
The discussion list for http://webstandardsgroup.org/
See http://webstandardsgroup.org/mail/guidelines.cfm
for some hints on posting to the list & getting help
*****************************************************
