Tuesday, June 27, 2017

Cirrus: Versioning control over the cloud embedded in 3D app ( Python )

I usually work on very different type of assets such as Digital assets in Houdini, meshes, textures etc.
I wanted to have a easy versioning control over these objects just like my python code on github. That is why I decided to create my own tool for that directly embedded in 3D apps using Python.

I decided to go with Amazon Web Service S3 simple storage, it allows you to keep former version of your assets and have a decent pricing model as well. It comes with a neat python API and it's also easily scalable.

This tool works on a base of "project", a project is just a folders hierachy where you save your assets and scenes. These folders also exist on the Amazon Cloud. You can then send and get data from / to the cloud direclty you app.
As it is fully python code only ( with PySide2 for UI and boto3 amazon API ), it can be used in many 3D packages which support python ( Maya, Nuke, Houdini, ... ) or even as a standalone app. The video of example bellow is made in Houdini only because it's my main 3D package.

It also support a lock system where you can "lock" a file you want to work on. The file is then locked on the cloud. If another artist want to work on that asset he / she will have to wait until you finished your task  on that asset ( until you submit your asset on the cloud or unlock your file ). This process ( very well known as "checkout asset" in Perforce system for instance ) allows you to safely work with multiple artists on the same project without overlapping your work.

It also comes with a "plugin" system which allows you to execute python scrits after an action such as, Get a file, Lock a file, Save a file or Click on icon. This, according a filters ( type of files, software etc.).
This allows you for instance to reload a geometry after getting a new version, reloading an digital asset etc.

This tool is designed to work with Amazon S3 system but the in / out interface is one python file, which can be changed to use another cloud provider as long as they have python API available.

Here is an example of the user interface of the tool direclty embedded as a python panel in Houdini where you can browse your project and get or send data from / to the cloud:

Here is a video explaning a bit how it works and what features are available:

The tool is still work in progress, there are still some bugs I have to get ride of. I also have to work a bit more and the UI and features.
As usual, the code is available and free on my github: