brah002 wrote:Is there a big difference between your "browser users" and the other guys' "virtual users"?
That is very easy to describe.
In eValid a recorded test is played back directly into the browser object -- that is, the actual automated browser itself -- so from a server's perspective it is impossible to distinguish an eValid run from a real user run.
The "Virtual User" concept is to make an abstraction of the test in terms of the way the browser communicates to the server via HTTP/S, and then to apply that same sequence of URL transfers to simulate a user's actions...that's why it is called a "virtual user".
The fact is, if all your test is doing is navigating from page to page this is kind of simulation can be pretty accurate. But if there are user interactions involved, or if the state of the browser changes based on activity or interaction between the user and the server, then the sequence of HTTP/S URLs not long is realistic.
eValid dodges that issue completely by never worrying about the specific HTTP/S sequences to use. Instead, eValid commands direct the browser to navigate to a certain page and then the underlying browser does all of the work. No faking any communication protocol, and total client-side state preservation.
So yes, there IS a big difference when you are dealing with anything other than the simplest kind of web applications.
The eValid Team
by brah002 » Mon Dec 06, 2010 4:39 pm