LTI embedded timed quizzes throws javascript error and breaks MathJax (impacts on STACK questions most)

Re: LTI embedded timed quizzes throws javascript error and breaks MathJax (impacts on STACK questions most)

by Stuart Anderson -
Number of replies: 1
Just discovered this issue is more unhelpful than I initially thought. Even if the timer is not used, if the "Close the quiz" option is used instead, a timer will still automatically appear 1 hour before the date/time is reached and the use of the timer again triggers this bug which breaks any mathjax rendering. 
The right solution here is not to remove the navigation block at all. It's a really bad idea.
In reply to Stuart Anderson

Re: LTI embedded timed quizzes throws javascript error and breaks MathJax (impacts on STACK questions most)

by Stuart Anderson -
I thought I would follow up with one final post about this in case anyone in the future needs help resolving it. Because of the impact the issue was having, I decided to dive in and work out my own fix. I have essentially changed the theme files I was using (boost theme) to update the embedded layout rule to point to different versions of embedded.php and embedded.mustache which add in the blocks column.  I've carefully documented this process and included some sample files for an older and newer version of moodle.  See attached zip file.

In the long term I would see the best fix in core to be either my fix, although I don't know all the contexts in which embedded layout is used and therefore cannot comment on undesirable side effects.  OR an additional layout is created for LTI specific embedded based on my fix.

I checked the impact of not having the blocks show against a range of tools.  Quizzes are serious impacted.  The Book resource tool will not show the Chapter block.  I was unable to test the Workshop tool due to the amount of effort required to set it up.  All other tools/resources did not have any dependencies on the side blocks and therefore aren't impacted by this issue.

I've also updated my bug report with the attached with the hope that a fix makes its way into the core code in the future.

Thanks to Daniel for giving me the initial pointer in the code so I could start to unpick the problem.

Regards,

Stuart