GSQL Now and In The Future

Since the introduction of the TigerGraph Developer Edition in June 2018, the GSQL language has generated a lot of interest and a lot of valuable feedback.

In general, the feedback has been positive for GSQL’s ease of use, high performance, and high expressive power. Here is a comment from a graph database expert and developer on GSQL – “Personally speaking, I am not concerned about current lack of Cypher support. I’m actually finding GSQL quite expressive and powerful.”

These comments come not only from graph database application developers, but also from SQL developers who are using relational database in their daily work.

To further help SQL users to embark on the exciting journey of graph analytics, we have open-sourced a set of classic, high-demand graph algorithms coded in GSQL. In the next few months, you’ll see significant improvements and additions to GSQL (described below). And we’re part of the joint effort to set a standard graph query language.   

GSQL Graph Algorithms Open-Sourced

Our graph experts have spent the last few months to create an open source graph algorithm library using GSQL. The library is available on github. We encourage you to download, customize, contribute to, and run it. We have a video demonstrating one application for our community detection algorithms, and of course we have documentation.

Graph Query Language (GQL) Standardization Involvement

The TigerGraph language team has been actively working with other industry experts to discuss establishing a property graph query language standard. The INCITS working group now is focusing on the SQL 2020 graph query language extension project. The goal is to support a property graph model in the SQL language, which includes property graph DDL, read-only pattern match query syntax, etc. We will participate in a face-to-face discussion of this working draft in Brisbane, Australia in Jan 2019.

In the same meeting, there will be discussion on establishing a stand-alone GQL project and its scope. There are proposals regarding the technical content coming from Oracle, Neo4j and TigerGraph.

We will work closely with the group on this topic for the coming years. Our latest submitted documents can be found at opengsql.org.   

What Will GSQL Look Like In 6 Months?

Here’s a preview of the product roadmap for GSQL.  We’re adding an interactive, interpreted mode for GSQL, so you can create-and-run big data graph queries in one step, without pause.  This is an exciting offering, as you’ll no longer need to wait for code generation and auto-deployment when you are prototyping or ad hoc exploring a big graph. The current compile mode will only be necessary when you’ve tested out your queries and want to optimize them for production.

We’re also adding support for multi-hop paths in one SELECT-FROM-WHERE block. This much-requested feature will make it simpler and more intuitive to describe multi-hop patterns for graph pattern matching.