10-26-2000, 07:13 AM
|
|
Junior Guru Wannabe
|
|
Join Date: Oct 2000
Posts: 68
|
|
Hi,
I'm trying to get the percieved download time of my page down and was wondering if anyone can throw in some useful advice.
What I want is the text of my page to download quickly, and then the graphical ilustrations to stagger in as the page continues to load.
The total HTML of the page is only about 15K but there are about 50K worth of graphics. I have placed all the text and graphics in one table cell and used some trick alignments to reduce the html from multiple tables and cells. I'm not currently online yet so can't test these theories out. But I'm guessing the html and text should come up really quick, with the graphics staggering in. Should this work? Or would it be better to put the text in one table cell, then the graphics in an adjacent table cell. What I don't want is the whole table cell to only appear after all the graphics have down loaded.
Your thoughts are appreciated.
|
10-26-2000, 12:39 PM
|
|
Web Hosting Master
|
|
Join Date: Jun 2000
Location: Southern California
Posts: 12,121
|
|
When the page doesn't load quickly due to a table (or all at once), it is usually due to having the table widths as percentages, and further complicating the whole thing, with tables inside tables inside tables -and mixing hard pixel values with percentages.
__________________
HostHideout.com - Where professionals discuss web hosting.
• Chicken
|
10-26-2000, 02:05 PM
|
|
Web Hosting Master
|
|
Join Date: Oct 2000
Location: Moldavia
Posts: 1,177
|
|
Chicken's right (as usual) on how the HTML interpreters in (most versions of) the browsers work.
The less work you make the interpreter do, the faster the page will "present".
That means explicit dimensioning of images and tables (avoiding percentages entirely if you can -- not always easy), because oftentimes page layout cannot be calculated by the interpreter until all the elements are parsed.
Nested tables are also a problem, because the interpreter needs to get through the entire "nest" before it can present.
Stacking tables with lesser vertical height is also good technique, since it speeds "top-down" processing by the interpreter.
|
10-26-2000, 05:52 PM
|
|
Web Hosting Master
|
|
Join Date: Apr 2000
Location: 80,000 feet under the sur
Posts: 2,736
|
|
As a side note, you can use the tools available at http://www.netmechanic.com to help you determine load time..
__________________
My 2 Cents.... (or is that 2.2 cents inc. GST...?) 
Have a think about this : Programming is like sex. Make a single little mistake, and you'll be supporting it for the rest of your life.
|
10-26-2000, 05:57 PM
|
|
Web Hosting Master
|
|
Join Date: Oct 2000
Location: Moldavia
Posts: 1,177
|
|
One caution on NetMechanic-like services: I have very definitely noted and confirmed that their evaluation of HTML code is often flaky, especially if there is a goodly amount of nesting. It also seems that many of them have a problem with certain types of FrontPage comments, typically those used with editorial includes.
|
10-26-2000, 06:31 PM
|
|
Web Hosting Master
|
|
Join Date: Jun 2000
Location: Southern California
Posts: 12,121
|
|
Also, it doesn't seem to actually test the download speed. I tested it on two servers, same file, same result. This would be fine, but one of the servers tested was faster (better server, better connection). So check download speeds but keep this in mind: results are not valid.
__________________
HostHideout.com - Where professionals discuss web hosting.
• Chicken
|
10-26-2000, 06:34 PM
|
|
Web Hosting Master
|
|
Join Date: Oct 2000
Location: Moldavia
Posts: 1,177
|
|
Most of these services don't actually test the speed -- they simply calculate it based on the size of the html file and the image resources used in the page. They also typically ignore the loading of things like JAVA applets, and any resources they might use.
But hey, whaddaya want fer free... lol.
|
10-26-2000, 06:44 PM
|
|
Web Hosting Master
|
|
Join Date: Oct 2000
Location: Moldavia
Posts: 1,177
|
|
Just a thought, by the way, but it's easy enough to build in speed testing on a page.
You simply grab the current time in the header with a JavaScript call; then use an onload= JavaScript call in the <body> tag to grab the time again (after the page has fully downloaded).
Subract the start time from the end time; then writeln the results discreetly to the end of the page.
A more elaborate example of this idea is exercised here:
http://webservices.cnet.com/bandwidth/
[Edited by etLux on 10-26-2000 at 06:49 PM]
|
| Thread Tools |
Search this Thread |
|
|
|
| Display Modes |
 Linear Mode
|
| Postbit Selector |
|
|
Posting Rules
|
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts
HTML code is Off
|
|
|
|
|
|
| Login: |
|
|
| Advertisement: |
|
|
| Web Hosting News: |
|
|
|
