On 10/1/2008 6:48 PM, Hedley Finger wrote: > Re the 1px deep repeating graphic, I thought that this would save > download time -- the usual delayer -- and allow the browser to quickly > render the page. If I use a larger graphic and it is not an exact > submultiple of the page depth, what happens then?
"Flat" graphics compress extremely well. Since my 7x16 GIF file consists of 16 identical rows, they'll compress to be only ever so slightly larger than a 7x1 image. Predictably, the largest of my graphics are the corners, which contain more variation. Even then, the largest of them is 163 bytes. Saving download time isn't a concern for these files. :) Even on a 2400bps modem, that largest graphic would take less than 1 second to transfer. Rendering can be significant simply because if a box is 1000 pixels tall, the browser has to render 1000 copies of a 1-pixel-tall image, whereas it only needs to render 63 copies of a 16-pixel image. I haven't played with the difference on modern browsers, but relatively few years ago, browsers would slow down considerably on pages with lots of image files, including pages with very large numbers of one tiled image. Having learned that lesson then, I've continued to apply it whether or not it's still necessary. For a 103-byte GIF file, it's not worth worrying about whether it'd be slightly more efficient if I made it 7x1 instead of 7x16. -- Erik Harris http://www.eHarrisHome.com - AIM: KngFuJoe - Yahoo IM: kungfujoe7 - ICQ: 2610172 - Chinese-Indonesian Martial Arts Club http://www.kungfu-silat.com ______________________________________________________________________ css-discuss [EMAIL PROTECTED] http://www.css-discuss.org/mailman/listinfo/css-d List wiki/FAQ -- http://css-discuss.incutio.com/ List policies -- http://css-discuss.org/policies.html Supported by evolt.org -- http://www.evolt.org/help_support_evolt/