- Correct swagger api
- Hide and un-schedulable sizes from the Size selection (#188)
- Label all sizes to show if they will be schedulable on the cluster (#187)
- Updated docs to reflect changes in jsp-wrapper
- Updated the documentation
- Add error message on incorrect cm
- Updated annotations and testing
- Changed build status gathering to prevent issues
- Set notebook namespace to None if it is empty
- Change custom namespace gathering to allow unset namespace
- Fix to ensure there is an initial image selection (#177)
- include thoth.yaml configuration for release maintainence
- setup bot configuration for release
- Modified element accessing
- Fixed dropdown closing
- Rework the package structure; (#59)
- Moved dropdown value check to Dropbtn.js (#60)
- Removed page.html (#64)
- Fixed double ifs (#61)
- UI Definition (#51)
- Changed build source to opendatahub repo (#43)
- Update and lock OpenShift client version (#45)
- API definition (#39)
- Add a pod gpu config handling (#34)
- Use the resource parsing for profiles as well (#35)
- Escape user name to prevent OpenShift rejecting invalid names (#33)
- Added notebook image documentation (#24)
- Add user name to pod labels (#31)
- Make sizes parsing more error proof (#29)
- Implemented image list gathering based on labels (#21)
- Move image list implementation from JupyterHub Config (#17)
- Implemented support for enviroment variable injection (#14)
- Update README for dynamic profile loading from ConfigMaps
- Fix variable vs string issue
- Add support for dynamicallly loading profiles from configmaps
- Change the GPU enablement to support GPU_MODE env var and OCP3 and OCP4
- Use OpenShift v4 compatible api version
- Add GPU configuration to pod spec
- Add (t-shirt) sizes for Jupyter Singleuser containers
- Handle empty dict for data properly
- Enable extension of user config map
- Write user config map
- Load info from users configMap
- Add username as env to every container
- Quote username as it might contain @
- Add proper jsonpath implementation
- Add support for services
- Handle empty config map
- Filter also when image is not defined
- Fix README to reflect the last changes
- Update environment variables in apply_pod_profile method
- Emphasize that globals must be the first profile
- Remove pycache
- Add resources to the profile
- Add ability to merge multiple profiles (to support global/default) profile
- Update README to mention asterisk in user list
- Add README and filtering by users
- Add initial implementation
- Do not fail if a service template is not available - just report error and skip the sevice deployment
- Added CSS to size desc
- Fix typo in processed_template logging (#37)
- Clean up error handling for services and refactor new code a bit (#36)
- Generalize how service integration works (#26)
- Implemented kubernetes resource structure to profiles and sizes. (#22)
- Allow customization of notebook node affinity and tolerations
- Fix typo in dict access
- Move json loading into API code (#46)
- Use template as string rather than yaml
- Added volume gathering
- Added default image handling and label
- Separated documentation into files
- Added volume gathering
- Set the MEM_LIMIT env var to real memory limit
- Force the UI to use https for API calls
- Changed how envvars are stored
- Removed debug logs in EnvVarForm
- Rework how env vars and secrets are stored
- Restructure code for user info rework
- Deep copy the default CM to prevent pointer issues
- Added additional ids in VarForm for selenium tests
- Only set gpu if the key exists in profile
- Fix the logic around setting default image
- Allow getting image detailed info at once
- Forced https calls
- Use the cluster GPU count to set UI limits
- Added generated schemas to swagger
- add gpu number response
- Fix the empty-size UI issue
- Gather image annotation data
- Testing pydantic library
- Added ui_config
- Add api prefix to UI andpoint
- Add support for custom GPU taints
- Use JUPYTERHUB_LOGIN_URL instead of default and try get user info without cache
- Added a sort function based on label to images
- Disable GPU config by default
- Enable switching custom env variable to a secret
- Fix for hiding sizes correctly based on enabled flags
- Use ui/config for spawner options
- Update UI
- Change apply_gpu_config return value
- Update UI
- Refactor Images
- Remove resources from empty_profile to prevent overriding
- include requirements.txt in manifest.in for source build
- Update image order annotation to match others
- Avoid throwing exception when tag.annotations are not set
- Enable the JUPYTERHUB_LOGIN_URL for login redirect
- Default the display name to imagestream name
- Fix to save changes on env variable deletion (#164)
- UI: Indicate notebooks are still being built (#163)
- Handle multiple version of notebooks (#162)
- Skip imagestreams which don't have any tags
- Move buildstatus to ImageTagInfo
- Fix the UI config for GPU number dropdown
- Add segment key and clusterversion
- Added a build_status value for each notebook image
- Add default empty list before loop in tag_exists
- Added recommended tags and changed structure of images
- Handle config load errors and config loading state
- Use the default tag if there is no prev or default image
- Check for the empty status
- Append tag name to the default image
- Handle image tags with invalid semantic versions (#173)
- fix username read issue in frontend API call (#172)
- [RHODS-1641] Add notebook_namespace parameter for cm and secrets