Channels are the fundamental organizational unit in Kolibri Studio. Each channel represents a curated collection of educational content that can be published and imported into Kolibri for offline distribution to learners.
What is a channel? A channel is a standalone educational content collection with:
Hierarchical content structure - Content organized in a tree of topics and resourcesMetadata and branding - Name, description, thumbnail, and language settingsVersion control - Track changes and publish new versionsAccess control - Manage editors and viewers with role-based permissionsPublishing workflow - Separate staging and published versions
Channel model fields The Channel model defined in /models.py:1004 contains the following key fields:
id = UUIDField( primary_key = True , default = uuid.uuid4) name = models.CharField( max_length = 200 , blank = True ) description = models.CharField( max_length = 400 , blank = True ) tagline = models.CharField( max_length = 150 , blank = True , null = True ) thumbnail = models.TextField( blank = True , null = True ) thumbnail_encoding = JSONField( default = dict ) language = models.ForeignKey( "Language" , ... )
Content trees Channels maintain multiple content trees for different purposes:
main_tree = models.ForeignKey( "ContentNode" , related_name = "channel_main" , ... ) staging_tree = models.ForeignKey( "ContentNode" , related_name = "channel_staging" , ... ) trash_tree = models.ForeignKey( "ContentNode" , related_name = "channel_trash" , ... ) chef_tree = models.ForeignKey( "ContentNode" , related_name = "channel_chef" , ... ) previous_tree = models.ForeignKey( "ContentNode" , related_name = "channel_previous" , ... ) clipboard_tree = models.ForeignKey( "ContentNode" , related_name = "channel_clipboard" , ... )
Each tree serves a specific purpose:
main_tree - The working version of your channel contentstaging_tree - Content prepared for publishingtrash_tree - Deleted content (recoverable)previous_tree - Previous published versionchef_tree - Content imported via ricecookerclipboard_tree - Temporary storage for copy/paste operations Permissions editors = models.ManyToManyField(settings. AUTH_USER_MODEL , related_name = "editable_channels" ) viewers = models.ManyToManyField(settings. AUTH_USER_MODEL , related_name = "view_only_channels" )
Publishing and versioning version = models.IntegerField( default = 0 ) last_published = models.DateTimeField( blank = True , null = True ) public = models.BooleanField( default = False , db_index = True ) deleted = models.BooleanField( default = False , db_index = True )
These fields are calculated when a channel is published:
published_data = JSONField( default = dict ) total_resource_count = models.IntegerField( default = 0 ) published_kind_count = models.TextField( blank = True , null = True ) published_size = models.FloatField( default = 0 ) included_languages = models.ManyToManyField( "Language" , ... )
Channel lifecycle Creating a channel When a channel is created (models.py:1235):
A unique UUID is generated as the channel ID
A main_tree root node is created automatically
A trash_tree is created for deleted content
Content defaults are initialized from user preferences
def on_create ( self ): if not self .main_tree: self .main_tree = ContentNode.objects.create( title = self .name, kind_id = content_kinds. TOPIC , content_id = self .id, node_id = self .id, original_channel_id = self .id, source_channel_id = self .id, changed = True , complete = True , )
Staging vs. published
Edit in main_tree
Content creators work in the main_tree, making changes without affecting the published version.
Prepare for publishing
Content is copied to staging_tree when you initiate publishing. This allows validation before making content live.
Publish
The staging tree is exported to a SQLite database and made available for Kolibri instances to import.
Archive previous version
The previous published tree is saved to previous_tree for rollback capability.
Channel versioning Each time you publish a channel, the version number increments:
version = models.IntegerField( default = 0 ) last_published = models.DateTimeField( blank = True , null = True )
The ChannelVersion model (models.py:1548) tracks version-specific metadata:
class ChannelVersion ( models . Model ): channel = models.ForeignKey(Channel, related_name = "channel_versions" ) version = models.PositiveIntegerField( null = True , blank = True ) version_notes = models.TextField( null = True , blank = True ) date_published = models.DateTimeField( null = True , blank = True ) resource_count = models.PositiveIntegerField( null = True , blank = True ) kind_count = ArrayField(JSONField(), ... ) included_licenses = ArrayField(models.IntegerField( ... )) included_categories = ArrayField(models.CharField( ... )) included_languages = ArrayField(models.CharField( ... ))
Secret tokens Channels use secret tokens for secure distribution:
secret_tokens = models.ManyToManyField(SecretToken, related_name = "channels" )
Two types of tokens are created (models.py:1394):
Primary token - A human-readable proquint string (e.g., “lusab-babad”)Channel ID token - The channel UUID itself
These tokens allow Kolibri instances to download channel databases securely.
Public vs. private channels public = models.BooleanField( default = False , db_index = True )
Private channels - Only visible to editors and viewersPublic channels - Listed in the public library, discoverable by anyone
Channels in the Community Library cannot be marked public. They use a separate submission workflow.
Channel storage Channels track storage usage for quota management:
def get_resource_size ( self ): tree_id = self .main_tree.tree_id files = ( File.objects.filter( contentnode__tree_id = tree_id) .values( "checksum" , "file_size" ) .distinct() .aggregate( resource_size = Sum( "file_size" )) ) return files[ "resource_size" ] or 0
Next steps
Content nodes Learn about the MPTT tree structure for organizing content
Creating channels Step-by-step guide to creating your first channel
