eValid -- The Web Quality Suite
Browser-Based Client-Side Functional Testing and Validation
Page Timing/Tuning Transaction Monitoring.
WebSite Spidering & Analysis and Realistic Server Loading.
Summary
This page describes eValid's architectural constraints and assumptions.
Limited Context
The eValid recording cannot take advantage of any information OTHER than what
is sent from the server.
This means you can't have "object repositories" or other historical records
of facts or details about the pages being browser.
Script Contains It All
A fresh instance of eValid plus a script needs to produce reliable playback.
This means that all of the information needed for a future playback has to
be incorporated IN THE SCRIPT.
Low Overhead
The playback time overhead of any eValid action has to be < 1% of the
total time, and it would be best if it was < 0.1% of the total time.
If the overhead is higher than this the accuracy of the collected data
would be in question.
Small Footprint
The footprint of the engine has to be small, not significantly larger than
the IE browser upon which it is based.
The users will accept footprint growth like that of a browser, but not
if it is significantly larger.
Looks Like A Browser
The product has to look like a browser and behave like a browser even when
it is NOT doing its test recording, playback, spidering functions.
This is done to instill confidence in the user that the product really IS a browser.