Itamar Tzadok

I think it has more to do with the particular page than with server performance. Some pages in moodle load more than others even with the extensive caching. So it may take the browser some time to process everything.

I too consider that an issue. Not everyone agrees though.

Usability-wise I think that editors display mode should be made more flexible. Many cases of text input do not require the full blown editor. I would prefer to be able to set the default to open a textarea and where needed click to open an editor for the textarea I'm currently on (the format value which accompanies every editor may be used to make the preference sticky).

Anonymized restore should not defaultly create anonymous user accounts and should allow assigning restored data to the restorer

(MDL-32792)

It's ok to let the quiz be only a test. It's wrong to think that a test does not involve learning. Every choice we make in life is a test of a possible route. How else do we learn?

Indeed the number of tries and extent of variation affects the effectiveness of learning by quizzes or any other component for that matter. Instructors who cannot or are not willing to make the necessary effort required for variation, feedback etc. should not blame the quiz. The often taken route of open ended (aka vague) assignments (essays in particular, shorter or longer) is much easier but rarely a mark of effective teaching and learning.

The overview tab under the course groups lists all groupings and for each grouping the grouped groups and their members. If the overview had some ui for moving groups into and out of groupings it would have been fairly easy to manage even with obscure group names.

The course groups import feature allows you only to create the groups with your choice of naming but it doesn't add the users to the groups. You would have to that manually. So, it's not much of a solution.

The obvious problem with auto-creation of groups per participants with participant name for the group name is that group names have to be unique and participants' names are not necessarily so. So name duplications have to be resolved one way or another.

I've just hacked the autogroup code to to auto create groups with participants names. For uniqueness I added to the name the participant's user id in parentheses (e.g. 'John Doh (3)' where 3 is John's user id). User ids may be long so the current auto numbering could be used instead to suffix the participant's name. It's not the most elegant solution but not necessarily less elegant than the current naming method. And it's just 3 more mostly harmless lines of code.

The docs seem to suggest that it is already implemented in earlier 2 versions but I'm not sure.