Julian Hyde (in Chicago, May 2019)

Welcome back!

Almost six years since my last blog post, I’m sharpening my electronic pencil to start writing again.

From Blogger to Jekyll

Previously at Blogger, the blog is now self-hosted at blog.hydromatic.net. I figured that the readers wouldn’t miss the cheesy ads, I won’t miss the spam comments, and the conversations started by the posts can all happen on Twitter or in other people’s blogs.

I’ve ported all of the previous posts to Jekyll, Markdown and stored them in GitHub. They’re available in the index. I must say, having full control over my content is quite a relief.


A few things have changed since my last post, but many things have stayed the same. Around 2012, I was writing a lot about the Mondrian OLAP engine and olap4j API, but my work on Mondrian has since tailed off. I had a promising SQL parsing/planning project called Optiq that has since moved to the Apache Software Foundation and is thriving as Apache Calcite.

My work on streaming SQL, which started at SQLstream, has continued in Apache Calcite, found collaborators in Apache Beam and Flink, and resulted in a well-received paper at SIGMOD 2019. With luck, a future version of the SQL standard will have extensions for streaming queries.

At that time I worked for SQLstream and Pentaho; I have since worked on Apache Calcite, Hive and Hadoop at Hortonworks, and I now work at Looker (which last week completed its merger with Google).

Why blog?

Why did I stop blogging? At that time, Calcite was starting to grow really fast and my energies went into Calcite features, releases and conference talks. Oh, and I also had a newborn and a 3 year old.

Twitter played its part. As the leading microblogging service, it allowed me to vent my passion and bounce those idea off audience. But by the time I had blown off steam in 140 characters, I no longer had the passion or outrage to rework the idea into a blog post.

But I’ve come to realize that ideas need more room to breathe. In technology, why you are doing something is often more important than what you are doing. You can develop an idea over several posts, and bring your audience along. After the product is complete, the blog will show the thinking that went into it (and perhaps a few wrong turns along the way). That’s what I hope to do here.

The cool stuff

I am interested in databases and business intelligence (BI), especially from the perspective of relational algebra and query optimization. I want to extend the database paradigm, to areas such as streaming queries and geospatial data, and on language design, to make database technology more useful. Lastly, I want to bring the technology to a wide audience via open source software.

I’ll be writing about those things here. I’m especially keen to introduce you to Morel, a language that I am developing. It is a small functional language derived from Standard ML that is also an elegant and powerful database query language, and I think it has a bright future.

Watch this space.


If you have comments, please reply on Twitter: