Script now allows individual event_proc test to be disabled as there are two cameras (so far) that crash when testing event_proc based set_focus() tests.
No need for retesting anything but those two cameras
Which are those?
Based on what we have been able to tabulate so far (
50 out of 121 cameras) the following cameras shutdown while trying to do the event_proc part of the tests :
As mentioned above, the version 2.11 of the test script allows individual sections of that test to be disabled so we will try to collect more detailed data.
As I already mentioned, the various MFOn event procedures should be equivalent (they should call the same firmware function, checked this in the a460 fw). If more than one of them exist on a camera, they are PT_MFOn and MFOn. SS.Create and RegisterShootSeqEvent are only needed to register MFOn and its friends. Since MFOn and MFOff have no arguments, it's probably possible to call them directly (if they ever become part of CHDK).
I have done some grepping.
Thanks - that's very useful. Some of the older ixus cameras do not have any MFOn type events ( ixus30_sd200 , ixus40_sd300 , ixus50_sd400 , ixus700_sd500 ). Fortunately, it also seems that older cameras will do a
set_focus() without any additional setup needed so perhaps that will not matter.
Other than those old cameras, it would be very nice if the solution to all this comes down to adding a script function to call an MFOn type event_proc (registered or direct) prior to using
set_focus(). Using
post_levent_for_npt("PressSw1AndMF") is a toggle function - you don't know if its on or off when you issue it. And just using set_focus() or set_aflock() crashes some cameras.
I'm not sure we are safely there yet but I think the results are encouraging?
Update : current spreadsheet of all phase 2 tests attached for reference
