We modified calendar_get_events in a non backwards compatible way
so instead of fixing that, an easier approch is to simply revert
it to the 3.2 implementation. This should be fine since the way
to get events is going to change in an upcomming release, and we
had to swap all occurences of calendar_get_events in core with
calendar_get_legacy_events to make sure the calendar and overview match.
Significant string changes:
* errorinvalidbyhour,core_calendar - correcting number from 59 to 23
* cron_help,core_admin - updated recommendation to run cron every minute
* tour1_content_end, tour1_content_welcome, tour2_content_end,
tour2_content_welcome all in tool_usertours - removing reference to
3.2
* invalidsesskey,core_error and invalidsesskey,mod_wiki - more
understandable error message
* pluginname_help and pluginnamesummary in qtype_essay - explaining that
the question type can be used for file uploads
* formatnoinline, mustattach and mustrequire in qtype_essay - changing
the word 'inline' to 'online'
Some auth plugins used to have a mix of the legacy style of plugin names
in config_plugins table (such as 'auth/mnet') and the new correct
style (such as 'auth_mnet'). Attempting to rename the setting plugin via
low level SQL UPDATE could lead to duplicate key violation.
The patch introduces a new helper function to safely migrate the old
settings to the new ones, eventually informing the admin about the
values mismatch.
Previously the assign submission page was using "lenient" logic
for overrides when more than one group override applied to a user
(i.e., use the earliest "open" date and the latest "due" date)
when really it should be using the sortorder as per the assign
grading table.
If the store is shared and the getAllKeys function is broken due to an
incompatability between libmemcached and memcached >= 1.4.23, then it is
not possible to purge the cache, and we cannot support the plugin.
This patch adjusts the isready check to additionally check if the
combination of libmemcached and memcached is affected by this issue.
In the detailed statistics view for one user attempt, the user response
for numerical question types was not being displayed.
This was only happening for incorrect responses.
