Andrew, e.a.,
> > 1. How did we get to a point where the web tier became totally > > untestable? I don't think it became totally untestable. (see my quote on js below) We simply were not able to easily rerun the tests and keeping them up to date. > > 2. Did we ever make a decision to require JavaScript to do even > > minimal things like registering users? During the template design, several tests were done with javascript deactivated. Actually, we want jspwiki to behave fairly decent still within old browser. Javascript is only used to improve user experience, but should gracefully degrade. I know this is not fully the case today. EG View, Edit, Info, Login is fine; Registering users, Viewing Attachments is not ok cause you need to switch TABs which requires javascript. > > 3. How should we fix web tier testing? In the short term, we can't > > even "cheat" and access the registration page directly. I think we should be able to fix all those functions which ''should'' run without js, rather easy. Let's log'em in JIRA, and fix them. > > 4. Do we care enough to delay 1.6.0? I do, but perhaps I am the > > only one. I would not delay the 1.6.0. much longer. I think the current jspwiki is pretty stable. As Janne suggested, let's run those 16 cases manually (with js on and off), just to make sure. I agree we need to improve on automated testing of the web tier. And selenium looks fine, but it will get us some time to get familiar with. I would also like to add javascript unit tests based on mootools spec driven tests platform. my 2 cents dirk Note on mootools: apparently selenium is using prototype, which seems to conflict with mootools. Let's automate web-junits of those core functions which should run without js support. I think for Javascript related stuff, we should be unit-testing with a js testing framework.
