So, until there’s a Kubernetes Operator for an Ethereum test chain (or something comparable where you click a button to get everything running and perfect), it seems like you’ll have to do things a little more manually when it comes to setting up an Ethereum development environment in the cloud. Why’s that? Because with a nascent tech like Ethereum, everything keeps changing rapidly, and a Docker image with an embedded Ethereum testchain and dev tools will be outdated in just a few months. So, in the end, you’d have to do the same thing anyway with just a basic Docker image: download the latest packages, make any needed adjustments to get a testchain working, and then refactor your code (Solidity, etc.) so that it’ll now compile without warnings or errors.
So, when I heard that Microsoft had created a service that would wrap around Ethereum, my eyes started to sparkle. Finally, I wouldn’t have to worry about playing the role of admin or devops, so that I can just develop…
…but then I promptly stopped and thought “Wait a minute? Microsoft? Creating a PaaS with Ethereum? Hmmmm…I don’t know about that…” Still, though, I’m willing to try anything once, so I decided to give Blockchain Workbench a chance.
So, I tried out Workbench to get a quick impression, and after a few hours, I think that I have a fair assessment of it. Now, if you’re a beginner, I’d recommend it: it’s a good place for an Ethereum beginner with an Azure account. But (and I’m trying not to be too harsh here), I would say that’s where its potential would be maxed out. Beyond being interactive teaching software for enterprise officers to learn about blockchain, I doubt that it’s something Ethereum developers or proponents would ever use. Not only is the Ethereum node completely hidden and inaccessible, the Workbench is an abstract wrapper around the node, so abstract that it becomes an hindrance to even communicate with the node. In the end, it’s completely unfriendly to a developer, an opinion which seems to be shared by others.
However, I will agree with Dimitrios on a second point: Microsoft’s template for Ethereum PoA could be useful. After experimenting with that one for an hour, I could totally see a consortium opting to try it out as a shared Ethereum solution.