With a great pleasure we will welcome Chris Engelpert for the second time here in Bulgaria for #jPrime2017! A small interview with him just before our conference:
Hi Chris! Can you please introduce yourself?
Hey, I’m Christoph Engelbert, Manager of Developer Relations at Hazelcast. That said right now, I’m doing exactly what my CEO expect from me: I’m in a hotel somewhere in the world, answering interview questions from somewhere else in the world, flying to yet somewhere else afterwards.
In general, the probably best way to describe my job is a lot of airports, airplanes, taxis, hotels, conference center, and trolling on twitter :-)
This will be your second visit to Bulgaria and our JUG. First time we were talking about JCache. What about will be your talk on this edition of jPrime?
Yeah it’s the second time and Bulgaria is really nice. Looking forward to come back!
At jPrime I’ll talk about Jet Streams. That means the fast flowing air currents around the world. In a more technical fashion, it’ll be about Hazelcast’s new stream processing framework, Hazelcast Jet. Given that, we’ll do some cool live coding (if you’re fast enough also hands-on) for our take on a distributed java.util.stream implementation, as well as a little bit of the much more powerful, underlying DAG API. For people that don’t know yet what that means, join the talk and learn about DAGs. If I get the chance and I’m still trying to make this happen, we will do live calculations with Hazelcast Jet from a live data feed of the World’s Jet Stream using java.util.stream APIs, doesn’t that sound like a good bridging of topics? ;-)
You were very active in the sun.misc.Unsafe debate recently. What is your take on the upcoming Java 9 in general and particularly in project Jigsaw?
I’d say good and bad. Given that I’m answering these questions relatively late (or because I’m good at fortune telling) I know that the first Jigsaw JSR proposal (JSR376) was down voted by the JCP EC yesterday, whereas the umbrella JSR for Java 9 (JSR379) was accepted. So the overall feature set of Java 9 is perfectly fine to the EC, however the specification for Jigsaw still is not yet clear enough or some issues still have to be fixed. The expert group now has 30 days to fix the spec and resubmit.
I was one of the people to down vote it in the EC and I stated my issues with the current specification, mostly based around two facts. The missing consensus inside the EG itself, as well as problematic “features” like the warnings if you activate “—permit-illegal-access”.
In general I think Jigsaw is a good idea, however I would like to see it as an “implementation detail” of the JVM. I don’t think it solves too many real world problems, as without versions support it is only meaningful to people that control all bits of the system. That certainly is the JDK team itself and maybe a handful of users building monolithic-styled applications. For Hazelcast, as a library vendor, we need versions. And to make sure we’re all on the same page, I’m talking about simple versioning, like “our Hibernate-Integration version X works with Hibernate version Y to Z” and not the famous multiple versions in the JVM at the same time because log4j 1.x was so badly designed - that is a whole lot of a different story.
That said, I don’t think Jigsaw will have too much influence and the adoption rate for Java 9 might still be low, but thanks to our “kill-switch”, slightly higher than it was expected a few months ago. JShell… nah, don’t get me started. HTTP2, yep definitely nice, would be a back port for Java 8 too (I think), so nothing too wow for Java 9. My personal favorite though is the Stack Walking API (http://openjdk.java.net/jeps/259)! I guess I would call my relationship to Java 9 as “acquaintance” but there are relationship goals to meet ;-)
You work as “developer relations”. What are the challenges in that new type of job that we see in a lot of companies nowadays?
I would say the biggest problem, traveling the world which is a big part of the job, is, that there’s still no way to beam myself from point A to point B in the blink of an eye. Weren’t we supposed to have this by now, looking back at the movies and series of the 80s? Maybe the USA did already invent it and is just not telling us? :-( The second biggest challenge is community management. A lot of the work is manual and people want to be entertained but also being showed appreciation for their work, which is just fair. I love my job and I love to see when people take something and make it their own, no matter if we’re talking about an issue, a pull request, a blogpost or whatever. And, as mentioned, I love to show appreciation, that’s where the Hazelcast Hero program comes in to specifically honor the most active community members.
Apart from that, I don’t think there are too many other problems or they all derive from what was said above, like in Developer Relations you need to show the effectiveness of the program because traveling is expensive. You see, beaming would help :-)
Is there something you like to do except programming?
I love cooking and barbecue, or let’s better said, I love food in general. Since I don’t see too much of the actual places I travel to, I always love to get the real local food experience and it hurt my soul going to India and knowing you cannot get street food without ending up ill. A lot of people wouldn’t believe it but I really enjoy eating and especially trying new, yet unknown, food. Apart from that, my life’s pretty boring and mostly consists of working on one project or the other. Fortunately there’s one more point to my life I really don’t want to miss, my wife. She’s a huge support since traveling means you’re not home too often and I guess there are not too many people to be ok with that.
Thank you very much! See you soon in Sofia!