fix: png rendering api & timeout parameter now works as it should AND default render timeout increased from 30 to 60 seconds, fixes #8710

Fixes #5515. Flot gauge has its own threshold check that is different
from the threshold check in singlestat. This changes the flot check to
be exclusive (same as the singlestat check). E.g. if the threshold is
10, then the value 10 is over the threshold.

* heatmap: fix converting error when series contains 0 and log scale is selected, issue #8884

* heatmap: fix app crash when Y min set to 0 with log scale

* heatmap: fix tooltip for 'zero' buckets in log scale

* heatmap: fix tooltip histogram for log scales

* heatmap: fix flicker of the highlighted element

this was caused by too often fired mouseenter/mouseleave events

* heatmap: fix missing X axis option for log scales

* heatmap: fix missing zero bucket in tooltip histogram

settings(change): changed default value for allow sign up to false, closes #8743, and changed default option allow_org_create to false

If a dashboard has a null value for updated_by then the migration
for the dashboard version table fails. This change uses coalesce to
set -1 instead of null when inserting into the created_by column
which has a non-null constraint. Fixes #8783.

This reverts commit 8634c9d4.

Fixes #5356. Internet Explorer aggressively caches GET requests which
means that all API calls fetching data are cached. This fix adds a
Cache-Control header with the value no-cache to all GET requests to
the API.

* add alerting notifier: DingDing

* add alerting notifier: DingDing

* add dingding unit test

* add dingding unit test

* delete debug code & format code style.

* fix build failed: dingding_test.go

Fixes #8344. Documents the $__interval, $__interval_ms and $timeFilter
variables.

Adds docs for the new API endpoints, and for the dashboard history feature.

Merge branch 'refactor-basic-diff' of https://github.com/walmartlabs/grafana into walmartlabs-refactor-basic-diff