The CReW Macros - A Retrospective

As I close the book on Alteryx it feels like a good time to look back and reflect on what I have often called one of my greatest successes and greatest failures during my time working with Alteryx -

The CReW macros.

For anyone not familiar with them: The CReW macros are a community supported set of additional tools for Alteryx Desktop Designer, first released on my personal data analytics blog (www.ChaosReignsWithin.com) back in June 2014 and very widely used across the Alteryx community. It used to be a joke with my colleagues at the annual Inspire conference that people were more excited to meet me as the author of the CReW macros, than for all of the work I had done on the Alteryx engine and product over the years!

Original Goals

Before we jump into looking at the success and failure of the project it is worth revisiting my original goals for the macro pack. At the time I started the project I was a full time C++ developer and was inspired by a very well known library in the C++ community namely the Boost Libraries (www.boost.org). These are a community built and supported additional set of functionality that a C++ developer can use when writing C++ programs. But as well as that, a lot of ideas that start in the Boost libraries eventually end up being added to the official C++ standard library, once the ideas have been tested out and refined as part of Boost.

This gave me 3 initial goals for my project:

1. Community built and supported

2. Macros should be useful and showcase best practises

3. Ideas that prove useful, get refined and added to the core Alteryx project

So let us look at the success and failure against each of these:

Community built and supported

This one is a mixed result: there have certainly been multiple contributions from members of the community over the years, and of course in the end I handed over the CReW macros to the Alteryx legend that is MarqueeCrew (aka Mark Frisch), but in reality the majority of the macros have been built by myself and Mark. I think at the beginning I had hoped there would be many more macros built by different members of the community, but in the years between I have realised that getting open source contributors for any project is difficult. People have limited free time to contribute to a free and open source project.

Overall I'll rate this category as a partial success.

Macros should be useful

"Macros should be useful and showcase best practises"

I think in this category we can call the the macros a huge success. Not only are they useful, but they are very widely used. Further they are well documented and provide a great illustration of techniques for building Alteryx macros. Myself and Mark have used them for multiple conference talks and as training examples over the years.

So category two is a definite success.

Ideas get added to Designer

"Ideas that prove useful, get refined and added to the core Alteryx project."

Ah yes... This did not go according to my original hopes and dreams... Remember that I took inspiration here from C++ and the Boost Libraries, but C++ is a different beast to the Alteryx language and what worked in one place did not work in another.

I think that the difficulty came because there was no forcing factor to add the new functionality into the product. Just because people were using it wasn't enough. And in actual fact I wonder if because the CReW macros provided certain functionality then customers stopped asking for that functionality so much. Which then was actually a counter pressure. But that is largely speculation. The facts are that very little functionality from the CReW macros ever made it into the product.

Category 3: a definite fail.

Conclusion

One fail, one success and a partial fail/success. So yes a mixed result. If I could go back to the start would I do things differently? Probably not. It was a fun journey and I am proud of what we created together.

Chaos Reigns Within