I have been looking at this same issue for longer than I care to admit and I am at a sticking point and wanted to see if someone could help me:
I am putting together my teams build script using ant. I have gotten the basic build to work (i.e. it compiles the .swc for us...RSLs will be next). However my issue is when I try to integrate with flex unit. I have antennae faithfully building the suite class and running them according to the example (tried about 3-4 different ways on it). However when I exectute the unit tests, I get no report, just an indicator if all the test cases ran successfully. After a bit more research and trying the obvious dawned on me, that the reason why is because it is a flash web application, which due to security, can not write to the local file system. (or am I missing something with the flex unit in the ant script?) I saw a trick where people were compiling the flash code into an air application. However the argument was being made that this does not really simulate the enviroment the user will use (management may be willing to accept this one). Furthermore they are saying (reading into the statement a little bit) that you could not trust the results 100% because air is different (this is the real issue). What I am not sure of, is how different? I understand that air is deployed locally, security is a bit looser because the user basically flags it as a trusted app upon install...ect. But what I am not sure is how much of an issue it will be. For example if we write the code flex code and run the tests in air but deploy it as a flash web application will the same sequence of tests have the same results (other than security issues will there be any difference for the same method)? From what I can tell in theory, the same code in flex web app should faithfully behave the same in air (not neccessarly the other way...again due to security) but don't know if there are gotch ya's that are going to crawl up at me. I can of course have the swc with the unit tests launch and look at the results and detailed information that way but it defeats the point of automating our unit testing with our builds a bit (we prefer a detailed report emailed to everyone when our unit tests fail in a integration/QA/production build as opposed to failing the unit tests without know why). Or have I have traveled down the wrong path? Please let me know if I have. Any help would be appreciated. Randy ------------------------------------------------------------- To unsubscribe from this list, simply email the list with unsubscribe in the subject line For more info, see http://www.affug.com Archive @ http://www.mail-archive.com/discussion%40affug.com/ List hosted by http://www.fusionlink.com -------------------------------------------------------------
