There is some discussion of this on the sakai-pmc list which I am not a member of and cannot post to.
Here is the history (in code) of this patch:
First pre-2.9 implementation is in
https://source.sakaiproject.org/svn/sam/branches/SAK-17243/ (SAK-17243 because SAM-1943 when T&Q JIRA issues moved to SAM)
Then there is a branch of 2.9.x with EMIs added which is what UCT has been running in production for a year:
https://source.sakaiproject.org/svn/msub/uct.ac.za/sam/branches/SAM-2.9.x/
Bugfixes that came up from testing and production experience were committed to this branch.
Then as preparation for merging EMIs to trunk (i.e. this process), OpenCollab created this 10.x branch of T&Q with EMIs:
https://source.sakaiproject.org/svn/msub/uct.ac.za/sam/branches/SAM-2.10.x/
from which the patch is created.
The design of T&Q is such that adding a new question type (with full support for import / export, printing, preview, delivery, authoring, etc.) does touch a large number of files.
It appears there are some changes in this patch which are not strictly EMI-related, and they probably arose from general T&Q improvements in the 2.9.x UCT branch. It might be possible to split them out further, but if they are valid on their own merit, then it is probably worth just committing them as well.
This is a complex question type, and we do not expect people to be able to cherry-pick a few commits to merge back to their 2.9.x branch, as T&Q has changed too much to make that a simple process. (It was complex updating the 2.9.x EMI code to the 10.x T&Q codebase).
UCT has done testing on the 10.x msub build, i.e. effectively trunk T&Q with EMIs, and has not identified any regressions so far. We would do further testing once the EMI code is committed, and also intend to branch 10.x code to run in production in our 2.9.x Sakai instance if feasible.
Stephen Marquard
![]() |
Monday 10 Feb 2014
Saturday 08 Feb 2014
Monday 27 Jan 2014
|