[solved] auto response doesn't work

AschsLines

Hi there,

i have found out that auto response doesn't work on some computers. any ideas?
cheers, Bjorn

sebastiaan

Hi Bjorn,

Could you perhaps provide some more details? In what respect doesn't it work (i.e. do you get an exception, or does nothing happen)? Does it occur with all backends? Does it occur with all types of responses?

Cheers,
Sebastiaan

AschsLines

Hi Sebastiaan,

thx for your quick response!

I enable 'auto response' in the 'Run'-Menu, and then receive these errors:

PsychoPy:

'glGenBuffers is not exported by the available OpenGL driver. VERSION_1_5 is required for this functionality.'

OpenGL:

'Runtime Error. The media_player_plugin requires the legacy back-end.'

Legacy:
though 'auto response' is enabled, it behaves like being not enabled, so i have to click through the experiment by myself.

I have tested with some of the examples, and the same thing happens.

Cheers & Thx, Bjorn

sebastiaan

Hi Bjorn,

The problem is that auto-response isn't supported by the media_player plug-in. It ignores the auto-response setting. It should work for other items, like keyboard_response, etc. The errors that you report do not appear to have anything to do with auto-response per se, but the fact that, for technical reasons, the media_player only supports the legacy back-end.

So basically, auto-response is a bit half-baked at the moment.

Cheers,
Sebastiaan

AschsLines

Hi Sebastiaan,

thx again!

What is still puzzling me is that auto response works fine on two of my computers (both Win7 64bit), but not on my others (Win7 / XP Pro, both 32bit).

Could it be a matter of CPU-Performance or due to 32/64bit?

I will perform some more tests, that keeps away me from sleeping ;-)

Cheers, Bjorn

sebastiaan

Do you mean that the same experiment with the same version of OpenSesame doesn't show the same behavior when it comes to auto-response? Hmmm, no, I don't see how that could have anything to do with the architecture or operating system. Please let me know if you find out anything-but also get some rest

Bjorn

Jap, that's correct :-)

I'm sorry, i should have mentioned i'm using a dropbox account to sync my experiment file and my openSesame portable 0.25 folder with all my computers. So i'm always using the very same version of my experiment file and of opensesame.exe on different computers.

Really puzzling, isn't it? :-)
Cheers, Bjorn

AschsLines

Hi,

maybe i have found something what keeps openSesame from running an experiment with enabled auto response:

on those computers, wich doesn't support auto response as it seems, two things happen.
First, the debug window will show no reaction, it will only prints this once, no matter how often an experiment is run:

Python 2.6.6
Type "help()", "copyright()", "credits()" or "license()" for more information.
Type "modules()" for details about installed modules and version information.
Use the "print [msg]" statement in inline_script items to print to this debug window.
>>> 

It doesn't even show simple print statements in an inline script.

Second, interrupting an experiment by pressing ESC openSesame will show a tkInter - dialog, and not the "openSesame says..." dialog

I'm not shure wether this is really causal to the not working auto-reponse function, but i'm going on testing that :-)

cheers, Björn

sebastiaan

Ah, yes! These computers must have Run in a separate process enabled, right? (See Menu → Tools → Preferences) By enabling that, OpenSesame actually calls opensesamerun.exe to run the experiment. This is mostly a remnant from the time when OpenSesame tended to crash quite a bit, and by running the experiment in a separate process you could prevent the whole program from crashing. Nowadays I don't really use it any more.

opensesamerun doesn't support the auto-response option, which explains your problem. It should though, so I opened up a feature request: https://github.com/smathot/OpenSesame/issues/81

Thanks for tracking this problem down!

AschsLines

SPLENDID :-)

That's it, thank you so much! I disabled the "run in separate process" and now it works. You saved me hours of way and clicking the hole experiment manually (wich takes est. 45 min, so it is really no fun to test it again and again) ;-)

Thx for helping me so fast and so much!

Cheers,
Björn