By Bradley Friedman, FIE LLC

fiepipe is a set of scripts and libraries meant to be a pipeline/workflow system for people and legal entities to organize both complex and simple computer systems and networks.

It has its roots in the world of digital visual effects (VFX) and animation as practiced at companies large and small through the late 1990s through the mid 2010s. However, it is not simply for those industries. Rather, it is an abstraction and refactoring of general digital asset workflows. As more and more work becomes digital (not physical) in nature, the work de-facto becomes: digital labor on digital assets and requires digital workflow managment.

Therefore, all the best practices of working in digital VFX and animation become relevant. And therefore fiepipe becomes relevant.

fiepipe is the base-library. Two separate but sub-projects are important, and probably what you want.

-fiepipe_desktop is a subproject that implements desktop tools. Such as the all important fiepipe command. On a desktop, fiepipe_desktop is what you want most, and fiepipe is a dependency for fiepipe-desktop.

-fiellc my (Bradley Friedman) LLC's plugin/command package. For now, it also contains a lot of application specific development. It has fiepipe-desktop and fiepipe as dependancies. As they become standard or stable, they will be migrated to other sub-projeccts.

fiepipe is currently in pre-alpha stage.

I'm trying to build it and use it myself. I expect to code it while I use it. I guarantee nothing. Use it at your own risk. And probably, if you want to use it, you'd do well to contact me.

There is currently a master branch and a stable branch for the project, and the sub-projects. The stable branch isn't so stable. But it usually does at least run, and tries to represent some kind of 'currently usable' state.

For now, there are two recommended ways to install fiepipe directly. Either to just use it, or to develop it while you use it.

Installs are generally acquired via github and use a standard setup.py file. This should be secure because you are trusting HTTPS (not HTTP) to guarantee that github is actually github (and not an impostor or man in the middle). And in turn you trust github to guarantee that I'm me. And you in turn trust me.

To just install it system wide for use, issue a command to pip which tells it to install from github.

For example, the stable branch from my development repository:

pip install git+https://github.com/leith-bartrich/fiepipe.git@stable

upgrades can be acquired using pip's -U or --upgrade flag:

pip install -U fiepipe

As fiepipe is alpha software, you may want to install it in a python virtual environment instead, to keep your system python clean. Or to test new versions.

Development installs typically don't originate fully from pip because you're likely forking the source-code yourself. Meaning: you probably have to check it out yourself from a specific account or repository because you're developing.

Once you've checked out the code, you can use pip or the setup.py file.

pip is often better at getting precompiled versions of dependencies from pypi. A pip install for development might be:

pip install -e .

Or, running setup.py yourself might look like:

python setup.py develop

you may need to occasionally re-run the setup.py file in develop mode if you edit the setup.py file in a way that requires it.

Similarly to standard installs, you may find python virtual environments very useful here to keep a development version and production version separate from one another.

I myself use PyCharm for development and have two separate python venv environments that I use to both develop/test/use the two branches on my systems.