- Comments are stored in HTML as well as Markdown, so that conversion
only happens when saving (rather than when viewing).
- Added 'markdown' Jinja filter for easy development. This is quite
a heavy filter, so it shouldn't be used (much) in production.
- Added CLI command to update schemas on existing node types.
A POST will create a short link (if it doesn't exist already), whereas
a GET will return the short link for the node.
The endpoint will return a dict like:
{'short_code': 'XXXXXX',
'short_link': 'https://blender.cloud/r/XXXXX',
'theatre_link': 'https://blender.cloud/r/XXXXX?t'}
This also updates Cerberus to 0.9.2 and simplejson to 3.8.2.
I've also changed the way we get to the application object, by replacing
from application import app
with
from flask import current_app
SCST tokens are now stored in the 'tokens' table.
This unifies old token handling and new subclient-specific tokens.
Also ensures the BlenderID expiry of the token is taken into account.
Removes use of httpretty, in favour of responses.
Authorised users can now create projects. The before and after insert
projects hooks take care of stripping unwanted urls and attaching
default node_type and permissions, as well as initialising a storage
bucket per project. We are temporarily switching to the development
version of the gcloud library, since it allows the creation of EU-based
buckets.
It is now possible to specify an encoding backend (at the moment only
zencoder) to take care of video variations encoding. Files transfer
happens directly on CGS (although any storage backend can be
supported). New requirements is the Zencoder Python library.
We introduce a new node_type, storage, which can support different file
storage backends and makes them available for browsing via dedicated
entry points in the application. We intend to keep pillar storage
agnostic and support both internal/local storages as well as
remote/hosted solutions.
