I am recording window declarations for webpages using SilkTest 5.0. When I recorded these declarations, I had unchecked the option Window ID and checked the options Caption and Index in the "Record Window Declarations Options" dialog box. But Silk still records Window ID as a tag in addition to Caption and Index. (Because it is an HTML page, silk records the corresponding HTML link as window ID). Any idea how to make Silk not record window ID as a tag in spite of saying it not to do so?
I already reported this to Segue. But they say "when recording a multitag the idea was to give the object more than one way to be recognised to create a more robust recognition system which would lead to more stable and consistant playback over different platforms. When there are only a few options available for the object and multitag is selected then the Windows ID will be displayed." This reply does not solve the problem. In my case, Silk has two other options namely Caption and Index. But it still looks for the third one! My point is Silk should not do more that what I ask it to do!!
I emailed Segue this morning about this problem (I selected to record window declarations by INDEX only, and is giving me all kinds of different tags). I'll post their response if/when I get one. Glad it wasn't just us.
"The behavior you described is a known issue with recording tags in Silk 5.0. Development says this problem has been corrected in the Silk 5.0.1 release. The work-around is to enable multiple tags (including the index tag) when recording window declarations for an object, then in the application's frame file remove the tags you do not want to use."
Latest from Segue:
"My understanding from development was that the tag problem in 5.0 had been corrected in 5.0.1, but this is not the case. ...I notified development again about not being able to record single tags. I will recommend that a patch be created given that this is a problem that could affect many customers."
I also sent a mail to Segue about this annoying bug! I don't know what bugs they fixed in 5.0.1 but they have a long list of known problems and issues. As a QA engineer I wouldn't release any product with so many severe bugs!!
In 5.01 they were supposed to fix a very serious bug when dealing with Java apps. Donno they fixed it or not. Anyone who tries to use their Java extensions when testing Java apps should upgrade to this release and test whether or not this bug has been fixed. The problem was with Silk running its own thread rather then running on the AWT events thread which causes AWT thread block when calling an API that does some GUI updates.