Interesting to hear. I will do a formal test on our network sometime when I get the chance and report on my findings as well.
Out of curiosity, it wasn't an actual physical stop watch was it? Its far less error prone to use something like the Net tab of Firebug or an external plugin like "Charles" which can show you how fast segments are downloading etc. On Tue, Jun 10, 2008 at 3:41 PM, Peter Ottery <[EMAIL PROTECTED]> wrote: > Anton wrote... > --------- > In regards to "I'm guessing this sort of structuring comes at a cost > because a number of requests need to be made to the server." this is > generally untrue. In principle this is exactly how download > accelerators work. They split a large file into smaller segments and > sent multiple requests. Since the browser environment is completely > multi-threaded it should actually boost performance. (Note: I am not > 100% certain if this is the fact, but there is no evidence to suggest > otherwise either). > --------- > > If its a small site, with not much traffic I think you'd be hard > pressed to notice the difference. For large news sites that get > smashed with traffic, I've sat there with a stopwatch and timed the > difference (over different speed connections from dialup to broadband) > between separate css files, and all in 1. And just having 1 file is > definitely faster. > > in some cases it would bring the initial [1] load time [2] from > something like 6 seconds down to 3 or 4. and then bringing all the css > into the <head> of the page rather than a linked file chopped another > second off. > > as i said - only applicable if extreme performance/optimisation is an > issue - but it *does* make a difference. > > [1] - with an empty cache > [2] - the time taken for the page text to appear, the page might > continue loading for 10 or so seconds after this so loading in pics > etc. mileage varies > > pete > > > ******************************************************************* > List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm > Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm > Help: [EMAIL PROTECTED] > ******************************************************************* > > -- - Anton Babushkin ******************************************************************* List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm Help: [EMAIL PROTECTED] *******************************************************************
