, ,

Hey Google Cloud, ever heard of Boto3? Come on.

Why can’t GCP come up with their own Boto3?

First, let’s set the record straight. GCP is better than AWS. This will be clear to anyone who has used both services for a reasonable amount of time. GCP was built with the developer in mind, the services and tools offered work better, are cleaner, and way simplier. But, there is one thing that is totally annoying. Where is GCP’s answer to AWS’s Python Boto3 library? I mean seriously. Boto3 is the one stop shop to plugin and interact with pretty much every AWS service available, and the documentation is reasonable. Seriously GCP, where you at?

Google Cloud has no answer for Boto3.

Want to work with Google Kubernetes Engine? Here…. no problem.

pip3 install google-cloud-container

Oh, you want a little Stackdriver Logging to go along with your app?

pip3 install google-cloud-logging

Oh, and don’t forget… probably going to need some Cloud Storage.

pip3 install google-cloud-storage

And so it goes…. on and on…. and on. There is no continuity for a single Python API interface into GCP. And yes, you need a great Python package API into any Cloud offering or you will never get the adoption you could otherwise.

Oh and don’t forget, the documentation… or lack thereof for the desperate Python GCP packages. Take the google-cloud-container documentation. I dare you to go check it out and try to figure out how it works. It’s a joke.

The terrible GCP Python library documentation.

Most of the documentation for the Python libraries are just a terrible list of methods and properties for the different Classes. There is logical organization or layout.

Boto3 is more Pythonic than anything GCP offers.

Oh you want to work with AWS resources? No problem here is our sweet documentation, it was actually written by a human.

pip install boto3

You want a little cloud storage?

s3 = boto3.resource('s3')

Need a instance to work on?

client = boto3.client('ec2')

I think you get the point. The documentation of Boto3 is laid out well, you can tell a lot of time and thought went into it. Overviews, quick starts, samples. Even for beginners, Boto3 and it’s documentation makes working with AWS resources easy and straight forward.

Is this why AWS is more prevelant than GCP?

There is nothing that will hurt a great product more than a crappy user experience. GKE is an amazing tool that has made working with big data fun. Having to write a custom Python wrapper around a crappy Google provided library was not fun.

Sure GCP has a great Console experience, better than AWS, but at some point real software projects move away from the Console and into the code and libraries. This is where GCP starts to fail.

If GCP wants to take a real swing at AWS they need to provide Developers some great tools and libraries, on par with Boto3.