Hi Chris,
I work in Lindsey Glickfeld’s lab and we have problem loading the xml with the newest nightly build. The xml file is attached here.
Thanks,
Ziye
Hi Chris,
I work in Lindsey Glickfeld’s lab and we have problem loading the xml with the newest nightly build. The xml file is attached here.
Thanks,
Ziye
Hi Ziye,
Thank you for sharing your experiment. I’ll be out of the office all of this week, but I’ll take a look at it when I return.
In the meantime, can you confirm that you’ve recompiled any custom plugins against the new MWorks build? (I’m mostly thinking of the LabJack plugin, but maybe there are others.) Attempting to use old plugin binaries with the new build definitely could cause the server to crash.
Thanks,
Chris
Hi Chris,
You are right about the old plugin. I didn’t compile LabJack plugin for the new MWorks build. But when I tried it, I got compile error saying the “pulseScheduleNode->kill()” no matching function for kill.
Thanks,
Ziye
Hi Ziye,
I got compile error saying the “pulseScheduleNode->kill()” no matching function for kill
Yes, the kill
method no longer exists (and should never have been used, even when it did). You can remove that line.
It’s possible that you’ll see other compilation failures that require minor changes to the plugin code. If so, please let me know, and I’ll guide you through them.
Cheers,
Chris
There is another error I got:
strcmp(optic_device->getValue().getString(), “led”)==0
“No matching function to call for strcmp”
Is getValue().getString() still available? optic_device is a string variable.
strcmp(optic_device->getValue().getString(), “led”)==0
You can replace that with
optic_device->getValue().getString() == "led"
Chris
That works! When I loaded xml file, there were two errors:
ERROR: Unknown attribute: “interruptable”
ERROR: Unknown attribute: “_error”
When I loaded xml file, there were two errors:
Yes, MWorks is now stricter with validating the parameters of certain components. You should do a textual search through your XML file, removing any instances of interruptable="..."
and _error="..."
(as well as any other unknown attributes). Hopefully, there won’t be too many of them!
Chris
I removed those attributes and there is error about Matlab interface file:
ERROR: Experiment packaging failed: The following parser errors were encountered:
/Users/hullglick/Repositories/BehaviorCode-Glickfeld-Hull/BehaviorCode/NewSkeletonCode/MATLAB_window_files/HoldAndDetect_Frames.m:1: parser error : Start tag expected, ‘<’ not found
function [retval] = HoldAndDetect_Frames(data_struct, input)
I removed those attributes and there is error about Matlab interface file
It looks like you’re attempting to load your MATLAB script as an experiment, which certainly isn’t going to work. MATLAB scripts should be selected in MWClient’s MATLAB window. Click the button labeled “Select .m file”, and choose the file you want to load. For more info, see Using the MATLAB window.
Chris
I didn’t realize I was uploading MATLAB script as experiment – they have the same name. I tested python script bridge and I was getting error for “keyError constD[‘rewardStaircaseWindowWidth’]”. I changed the rewardStaircaseWindowWidth variable name to rewardWindowWidth and it fixed the error message. I wonder if it has anything to do with length of variable name?
I changed the rewardStaircaseWindowWidth variable name to rewardWindowWidth and it fixed the error message. I wonder if it has anything to do with length of variable name?
There shouldn’t be any limit on the length of variable names. Are you sure the experiment contains a variable called rewardStaircaseWindowWidth
? (I don’t see either of the variables you mentioned in the experiment you sent me.)
Chris
I attached the experiment I was testing.
Hi Ziye,
Thanks for the additional files. I’ve reproduced your Python script error. This may be related to Mark’s Python bridge issue. I’ll investigate and let you know when I have more info.
Cheers,
Chris
Hi Ziye,
I’ve reproduced your Python script error. This may be related to Mark’s Python bridge issue.
I was mostly right, except that it was a different issue than the one I referenced previously. In any case, the fix for this problem should be in tonight’s nightly build. When you have a chance, can you download the new build and see if it fixes the issue for you?
Thanks,
Chris
Hi Chris,
The error is fixed in the new version. Thank you.
Ziye