Forum Moderators: phranque
However, if you're apprised of that quirk ahead of time and just use it for relative figures, it's very useful.
However, there are at least 3 good reasons to make use of an independent download estimator:
(1) You may be developing for a platform which has a lot of its own HTML built in, e.g., Yahoo! Store. In such a scenario, you won't know about the download time until you visit the published page (or point the download estimator to it).
(2) The most efficiently designed page is still at the mercy of the server which parcels it out. Using an external download estimator is a great way to determine whether your Web host is giving you good enough file throughput.
(3) "The proof of the pudding is in the tasting." For any kind of development effort (from architecture to, uh, making pudding), you won't have complete confidence in the effectiveness of your design until you experience the final product or --better yet-- have an objective third party experience it and report back to you. An external download estimator is that kind of objective third party.
...You know, it occurs to me that it might be useful to have a few Web-based PING tools too. These won't help with testing download time for an entire page, but they will help test the relative speed of your Web server compared to others. I try to test with three servers to (hopefully) determine an average server response time.
TestMySpeed.com's Alert Site Net Test Tool
http://www.testmyspeed.com/tmsalertsite.htm
CNET's SiteSpeed Meter
http://webservices.cnet.com/Ping/
Network-Tools.com's Ping and more
http://www.network-tools.com/