Re: Maxiumum Page Size Web Standards

zoran knezevic wrote:
> 
> You are right maybe this is not best place to talk about usability but

I'd suggest the w3c-wai-ig mailing list, as accessibility tends to 
include the issue of people who are running technology more than a year old.

> to add practical example from real life. Comments on articles on one
> of my sites have pagination 1 2 3 and link show all comments. After 3
> months we did analyze of Apache logs and result was shocking, 90% of
> people click "show all comments" instead of going to page 2 or 3...

You need both.  If you know that you are in the right place, you can 
generally wait for the document to load, especially if it incrementally 
renders properly, but, if you are coming in from a search engine, you 
want to know very quickly whether the page is going to be useful, and 
therefore want a small page.

(I tend to skim all FAQs, so I don't like FAQs that are paginated.)

Generally, though, it depends on the nature of the page, but most home 
page are way too large at the moment.

My feeling is that you should aim for intermediate navigation pages to 
load in about 2 seconds at 33kbps (note that contended ADSL will average 
  only 40kbps, although, if you are contending with sensible size web 
pages, the burst rate will be a lot higher).  I think that 5 seconds may 
be acceptable for the first one (to download common scripting and image 
resources, when a search hits the navigation page).

I'd suggest that homepages should display a non-trivial description of 
the nature of the site within about 5 seconds and complete in 10 to 15 
seconds.  Splash pages, of course, should be of zero size!

Leaf pages should try to display (incrementally rendered) the first page 
in about 2 seconds.

All of these are easily achievable, but are rarely achieved with typical 
sites.

At 33k, 2 seconds is about 12KB of text or 6KB of image.


-- 
David Woolley
Emails are not formal business letters, whatever businesses may want.
RFC1855 says there should be an address here, but, in a world of spam,
that is no longer good advice, as archive address hiding may not work.

Received on Saturday, 30 June 2007 21:49:21 UTC