At 3:17 PM -0400 2000-07-23, John Jacques wrote:
>Hello, I'm using the C API. Celeron 466, Slackware 7.0,
>This is a long message because I wanted to include anything that might
>When I pull a small amount of data it is instantaneous, but the more I
>pull, the slower it goes. It takes about 30 minutes to pull 1.3 megs
>using the following query. I'm using Netscape and my own cgi program.
>The status netscape shows at the bottom of the screen starts off at
>30K/sec and keeps dropping. After 640K of data has been loaded, it is
>down to 3.3K/sec and still dropping. After 800K the speed is 2.7K/sec.By
>the time I load in 1 Meg the speed is 2K/sec and still dropping.
>The same results from the command line, meaning it takes the same amount
>of time to display the 1.3Megs of data with my cgi-app and displaying
>the results on the terminal screen, so it's not a Broswer problem.
How does the command line client being fast indicate that it's not
a browser problem? If anything, it indicates that it *is* a browser
Are you displaying the data using an HTML table? If so, that may be
part of the problem. Netscape is pig slow with tables, particularly
big ones. 1.3MB is certainly big. So big that I'm kinda wondering:
how can it be useful to display that much data in a browser window?
(Or are you simply causing a file download rather than displaying
in a window; I don't think your message specified which.)
Paul DuBois, paul@stripped