Welcome to Pulse’s documentation!¶
Pulse is a resource version control system. I think you already know what a version control system is (GIT, SVN, etc..) but what the heck is a resource?
If you have already broken a project into smaller parts for team working, you already know what a resource is. A resource can be a prop’s modeling, a character’s rigging, a shot’s animation… A resource is versioned, and each version contains a work area and its products.
Here are concepts manipulated by Pulse with some examples.
Installation¶
Pulse is written in pure python, you don’t need to install any external libraries to use it. It has been tested with python 2.7 and 3.7
You’ll need to add /src directory to you python path. Then you can use the launch the standard test :
/tests/test.py
Test data will be generated under the tests directory, and they are ignored by git.
Adapters¶
To connect to a database or a file repository, Pulse has an “adapter” strategy. You can find them in /python/pulse/database_adapters and /python/pulse/repository_adapters In each folder you will find an interface_class.py you can derive from. An adapter determine the way you want to share data : the protocol you will use, and how it’s stored. It’s transparent for the user.
By default, API is set to use json_db.py adapter for database, and file_storage.py for file repository.
json_db.py is a very simple database, writing data to json files, stored in their class directory. It’s mainly used for testing the api, because it’s fast and easy to debug. But you could use it for simple projects where everybody work on the same network.
file_storage.py is a simple file repository system. It can stores files in any path writable for the user. This adapter is also used for the api test, because it’s fast and easy to debug. It’s also a good choice if you wish to store all the project’s resources under a single network path.
Custom adapters¶
You can write your own adapters, but Pulse also come with two custom adapters : a mysql database adapter, and a ftp repository adapter. Those adapters have their own test file : test_custom_adapters.py. Of course, you will have to setup a few things before using it.
First, you will have to install mysql connector, available with pip : # pip install mysql-connector-python
Then you will have to create a tests/custom_adapters_config.ini according to your own connection parameters. The mysql adapter has been test with mariaDB 10 server. The mysql account you will mention needs to have permission to create or remove a database. The ftp account needs to write and delete in the root folder you will mention.
[db]
host=192.168.1.2
port=3306
login=pulseAdmin
password=***
[ftp]
host=192.168.1.2
port=21
login=pulseTest
password=***
root=pulseTest/
Linked Directories¶
By default Pulse will create linked directories inside your work directory :
user_sandbox/{resource}/output -> user_products/{resource}/{current version}
user_sandbox/{resource}/input/{input product} -> user_products/{resource}/{chosen version}/{product}
It is very handy, since you won’t have to browse the user products directory to find the right one. And since the product path is abstracted, you can update any input very easily without the need to modify the work file.
But “With great power comes great responsibility”. You have to be very careful if you generate products which uses inputs (aka “not self contained products”).
In example, let’s say you’re working on a rig, you have texture input and its path uses the input linked directory. When you will generate the rig, if you don’t change the path you will refer to something like : user_sandbox/joe-rig/input/joe-texture.viewport
This directory won’t exists for anybody will use the rig product as input. When the rig product will be downloaded, it will also download the texture, but the product path will be user_products/joe_texture/v005/viewport. So you should always think to remap product path if you use the linked directories