Advanced Configuration

Warning about the configuration

The configuration can be tricky, especially for complex cases. There is currently a lack of documentation and tools to assist and ease the configuration. Some complex use cases may not be achieved out of the box without augmenting Spil itself.

If you read this and are considering the use of Spil, please don’t hesitate to contact us at spil@xeo.info.
We will be glad to help.

This documentation is work in progress.

Configuring from scratch

To start with, there are two options:

  • you configure Spil for usage in an existing pipeline

  • you create a new pipeline

Configuration for an existing pipeline

  1. spil_sid_conf

This files contains the templates for the Sid. The Sid is an abstract representation of your pipeline, it describes the data you manipulate at a high level.

The typical sid pattern looks like this:

'asset__file':            '{project}/{type:a}/{assettype}/{asset}/{task}/{version}/{state}/{ext}',
'shot__file':             '{project}/{type:s}/{sequence}/{shot}/{task}/{version}/{state}/{ext}',

Some tips/questions to guide your configuration decisions:

  • use the naming convention of your existing pipeline, eg tasktype or step? asset_type or category?

  • use abbreviations if your naming convention contains some, eg seq for sequence

  • by default use short, industry adopted, naming, eg “asset”, “shot”, “task”, “version”

  • keep one case for all, preferably lowercase

  • stick to singular (prop vs props) unless your naming convention states otherwise.

  • match the hierarchy of the filesystem (given the filesystem plays an important role in your system). Is {version} above {state} or vice-versa? Eg. animation/work/v002 or animation/v002/work? (The latter is the typical cgwire config)

  • Keep the sid config simple, but complete. Do not cut out what you may need, but don’t carry around invaluable fields. Eg. Do you use tasktype (aka step) and task, or only task? Or task with optional subtask? (see the “kumquat” config example) Do you need an asset variant (for clothes, lods, etc.) or is it included in the asset name ?

  • whenever possible, use enforceable patterns. Examples: sq010, v002 are easier to resolve than named sequences or just numbers.

  1. spil_fs_conf

This files contain the path templates. For all templates that exist in spil_sid_conf and that have a path representation.

Testing the configuration with an existing file system

  • Use parse_sids_from_fs.py to parse existing sids, and check if they get correctly parsed

Configuration for a new pipeline

So, you are about to create a new pipeline ? :)

Some tips/questions to guide your configuration decisions:

  • build a strong naming convention with short, industry proven terms. Eg. use Shotgrid terminology, check out the spil glossary

  • define the data you need, and structure it in a hierarchy

      - new pipeline - choices
      - overall hierarchy definition and glossary
      - episodes or not ?
      - step / tasktype
      - named or numbered shots and sequences
      - state over version (cgwire)
      - publish/work version matching
      - push or pull updates ? (publish permalinks, eg. a "valid version" or explicit dependent version update)

(See also questions for an existing pipeline)

Testing the configuration

  • Create or adapt the generate_example_sids.py script (as found in spil_hamlet_conf/hamlet_scripts) which generates correctly formatted test Sids.

  • Run save_examples_to_mock_fs.py: this will create dummy project files and folders on disk

Creating a new type

Create a new Config entry.

  • First, we define the data in the spil_sid_conf.

  • Then, we define it in spil_fs_conf.

Testing a new config entry

  • Checking the config

  • Testing the Sids

  • Testing the Finders

Check out: Testing the configuration.

This documentation is work in progress. Do not hesitate to get in touch if you are interested in using Spil.