Note: This article is my personal view. I don’t work for Red Hat / Mulesoft / WSO2 , neither they are so keen to hire me.
RedHat acquiring 3Scale... this is really interesting news for any integration developer who uses open source-based integration products and solutions.
I have been working with products like Fuse ESB, Apache Camel, Mule ESB, and WSO2 for more than 5 years. An interesting time lapse goes on in my mind when I look back, on how they have grown in terms of technical capabilities and in business. (Open Source Middleware market – revisited)
It was 2010 I started off with Apache Camel, It was so hard to do a simple file to file transfer, because there was no good documentation (I was annoying many members of the forum multiple times a day!). After I completed writing code using Apache Camel, I literally had just two lines of code to do the job, and from that moment on, Camel impressed me.
Doing the same with Mule was even faster, because for ‘File to File’ its all a matter of dragging and dropping icons. Camel code still looked simpler, but I am bit obsessed with Camel, forgive my bias ;)
Doing the same with Apache Synapse and WSO2 was bit harder, because if you are already an SOA person you will love WSO2, but not as a beginner. The interesting point to mention here is, if you want to do a simple file to file integration you don’t even need a IDE or write any code, you can do everything using WSO2’s web-based console. Now I realize, in 2010 what WSO2 had as platform features were visionary. I had told my friends and colleagues, "if you want to learn the entire SOA stack easily, just visit WSO2's site and go through their products."
Coming back to the topic: why this news matters. If you already read my article on open source market trends, you can observe some of the major changes in trends. I am sure this news is going to change the trend again.
The trend of the API Economy is the key reason. Mulesoft is a rockstar so far because of their API-centric middleware capabilities. Nowadays, people are less worried about integrating various systems, it's so easy with all these products. If it is only about integration, I will always prefer Apache Camel. Its my personal favorite, easily readable, and easy to use. But RedHat’s middleware product suite lacked a good API manager on top of their middleware stack. That's a key differentiator between Red Hat and Mulesoft from an integration perspective.
After working with Apache Camel and other Red Hat products for such a long time, I recently started working on Mule again. In the present version of Anypoint Platform, it is so easy to build and expose an API. Whether you are exposing it to a external party or whether you need to expose them with high end security, it's so easy to do. You don’t really need to be a core programmer to expose APIs using Mulesoft’s Anypoint platform.
This exact capability is what I wanted on top of Fuse ESB and Apache Camel. I am sure this particular aspect was a key decision-maker for many customers who have adopted Mulesoft. That's why the acquisition of 3scale is a big deal. Red Hat’s middleware stack is going to look more even attractive.
With the acquisition of 3scale, Red Hat has also announced that it's going to open source the code. Well, thats even better. Why? Imagine this combination: Open source 3Scale, Apache Camel, Spring boot, plus some necessary knowledge to roll them out into production! That's an attractive stack for any startup company who doesn't have sufficient funds to buy a production support license. And of course this will be a strategic driver for Red Hat’s sales!
What do I foresee ? In the next five years, open source-based integration technologies will become even more robust and easy to use. Also, it wont be a niche competency to develop, it will become as easy as developing an application in Salesforce. Also, the number of diverse protocols for integration will decrease if everything becomes API-centric (but the number of connectors will keep increasing for applications, even though everything is going to be REST-based).
Last but not least, there will be more annoying HTTP calls and connections. That's one drawback of this trending API-based integration. Well, let's hope ISPs build better networks so that API and cloud-based integration works seamlessly. Otherwise, we will have someone else to blame for slow response times. :)