Drop markdown for monospace
With markdown, we're only allowing about 1/10 of what is possible.
This creates a false reality and expectation. Lets not suggest this.
Nor do we have in-editor preview. So the user has to save, look at it,
then go back and edit. And seeing a bunch of sanitized HTML or just
missing sections isn't firendly.
Monospace, what you type, is what you're going to get. It gives the
presenter enough power to build a readable abstract, but not so much
that they can break the page and it's CSS, nor the ability to confuse
onselve through not getting what you expect.
We keep bleach sanitation and we should probably run linkify on this in
the long term. (Turn links into clickable links)
This is all noop in Py3, and Py2 is broke now in various places.
Dropping Py2 code as it will not be a thing going forward.
Django 2 is the next release, Py2 support will be dropped, as such,
dropping this is forward looking.
* Remove markitup (to be replaced with Ace editor)
* Use DUA decorators
* Removed custom signup bits
* Upgraded dependencies
* Added migrations
* Namespaced template locations
* Removed html5parser/sanitizer (for now) - parsing functionality
should be moved out entirely to a hooks
* Replaced ProposalScoreExpression object with a function that returns
F() expressions
- forms for labels and helps
- views for success or warning messages
- apps and models verbose_names for admin
Signed-off-by: Hiroshi Miura <miurahr@linux.com>
Requires that a 'reviewers' group exist, and that there are people assigned to it (otherwise it won't find any reviewers to add!).
It will assign up to 3 reviewers, where those reviewers are NOT one of the speakers, and that reviewer has not opted out of reviewing that talk. It will choose random reviewers from those with the fewest existing not-opted-out assignments.
When a review was not the latest it would cause a bug on deletion of the
latest. We changed the logic to work with the latest vote versus the previous
to fix this issue. Now all review deletion ensure a consistent latest vote
state.
The original default value of one assumed ProposalResult was being created
when a Review was created which meant a comment and a vote was made. This is
no longer the case from changes after the original change and has caused off
by one errors in the denormalized count.
