JavaOne Keynote: Oracle Talks Blockchain, Bots and Serverless

The developer keynote at JavaOne on Tuesday focused on communication and Oracle’s cloud offering.  Oracle provided an update on Oracle Developer Community, the relaunched OTN that was announced at JavaOne last year, and provided a number of demos and overviews of Oracle Cloud including showing the integration with two companies Oracle has acquired, Apiary and Wercker, talking about their bot platform, Blockchain, and their “serverless” offering.  Slack’s Buster Benson also appeared to speak about Slack’s API and customisation.  The full keynote video is available on demand; below we’re providing a summary of the key information. 

The keynote kicked off with Bruno Borges and Stephen Chin talking about Oracle Code. Oracle Code is a series of events that started in San Francisco a year ago. Since then, it's traveled to 17 countries and 20 cities. 10,000 developers have attended, and they've had 350,000 online viewers of videos recorded.

After talking about Oracle Code, Chin switched to discussing the Oracle Technology Network, often called OTN. The OTN team decided they wanted more engagement with people, community, and developers. Rather than talking about technology in general, they wanted to publish more content on JavaScript, Java, DBAs, and DevOps. Therefore, they renamed OTN to Oracle Developer Community. The website for this community, developer.oracle.com, launched last year at JavaOne. It regularly publishes technology articles and examples of how to implement and use different languages with Oracle Cloud and other products. It's a great place to learn about what Oracle Code is and all the events.

They've also migrated social media accounts to use this branding. You can find @OracleDevs on Twitter, Medium, YouTube, and Facebook. In addition to articles, they're offering a number of other learning tools:

• MOOCs (Massive Open Online Courses), all free.
• Code Online webcasts and DevLive interviews.
• A Developer Newsletter, with monthly updates.
• Oracle Zip Labs - Hands-on labs (with beer!) for developers with challenges and giveaways.
• Interviews with Developer Legends, Champions, and Luminary Speakers.

Oracle has recognized that communication is critical, hence the reason for many of these efforts.

To help underscore this notion, Chin invited Buster Benson, head of platform product at Slack, to come to the stage. Benson started by asking "What is Slack?". He then described it as a messaging platform with 50,000 paying customers, who work 10+ hours/day on average. It's the interface of work.

The reason that so many customers have found Slack valuable is that it provides an excellent way for human-to-human communication. It's at the heart of what Slack does. At a high-level, Slack helps you 1) find people and information, 2) build shared understanding, and 3) make good decisions quickly.

Slack has over 1,000 Apps. And 90% of Paid Teams actively use apps.

Via Slack Enterprise, they've discovered that every enterprise company is special. However, there are some things they've found that every company wants to do better.

• Use case #1: Get timely notifications about important things. How's the business doing, not just when things are on fire.
• Use case #2: Know when something important happens. Approve/Reject directly from Slack.
• Use case #3: Pull relevant context into the conversation. It's often tough to pull information behind the links. The solution for this has been what they call "app unfurls."
• Use case #4: Pass structured info around seamlessly. Slack's solution: dialogs.
• Use case #5: Get ready for bots. Donut is one of his favorites. It pairs you up with someone for coffee or donuts. It knows who you talk to, and connects you with people that are new.

You put all these things together, and you have a conversation.

Slack aims to provide a myriad of options for customization, making it easier for the people within your organization to build the processes and tools that everyone uses. You can find Slack's API docs at api.slack.com and its apps directory at slack.com/apps.

Next on stage was Amit Zavery, senior vice president of Oracle Cloud Platform.

Zavery started by saying it's a great and exciting time for developers. The world has changed quite a bit in recent years. We used to have hard-coded elements; now we have a dynamic assembly of blended experiences. There used to be a 12-18 month app development cycle; now there's a zero-time development cycle that's days or even hours. There used to be centralized teams; now there are different teams with different (polyglot) preferences. Monolithic applications have given way to smaller components (for microservices). Rather than just building another app, we're creating engaging interfaces for mobile and chatbots. Many developers are also embracing API-first design.

This has all caused new application development demands. Developers want to write polyglot microservices, deploy containerized applications (with Docker), do DevOps with cloud-ready tools like Gradle, Maven, and Git. DevOps team want elastic scaling, smart performance monitoring, and they want to deliver applications quickly.

Zavery described a demo titled "The Art of Buying a Car." He showed slides with screenshots of an  app that can give you the history of a car on your phone using Facebook Messenger as the UI for the app. It allows you to upload a picture of a license plate and get the history of that car in a few seconds. The chatbot then prompts you to see if you want to buy it, then shows similar cars in your area based on your zip code. Zavery coined this as "API-first, mobile always development."

To talk about API-first development, Zavery invited Jakup Nesetril on stage. Nesetril is vice president, production development API service, Oracle Cloud Platform. He used to work at Apiary before Oracle acquired them.

Nesetril's appearance was mostly to talk about how to do API-First - both design and development.

He mentioned that 300,000 developers are currently using Apiary. Apiary allows you describe your API with a flavor of Markdown. You write documentation, and you can run commands (e.g.) against the mock server. You can push the code directly to GitHub, and there's a bi-directional sync. You can edit files on apiary.io or via GitHub.

This demo was used to announce Oracle offers full API Lifecycle Management. You can analyze, design, consume, govern, secure, and monetize your APIs, from start to end.

To talk about Microservices: Build, Deploy, and Iterate, Anand Kothari was brought to the stage. Kothari is a senior director, product management, Cloud Native Apps, Oracle Cloud Platform.

The first thing you'll want to do with microservices is to have a complete development lifecycle with CI/CD. You'll need to manage and track agile development processes. You'll want to improve team collaboration and have source code management. He said that Oracle Cloud offers all of these features, and it contains some nifty add-ons like Slack integration and one-click deployment to any container.

Oracle Developer Cloud Service is a polyglot runtime with automated container management. It provides an elastic and highly available container runtime, designed for modern, lightweight, polyglot apps. This allows you to leverage multiple development skillsets. Its built-in container orchestration and scheduling looks pretty slick too.

Kothari then did a demo of Oracle Developer Cloud Service. The UI looked a lot like GitHub for source control. It has an Agile tab with a backlog, active sprints, reports, swimlanes and other project management features. The Build tab offered continuous integration and looked a lot like a reskinning of Hudson. The Deployment tab shows deployment history and production apps statistics (e.g., memory usage). He also showed how you can use this interface to bind apps to services (e.g., databases).

One feature he showed was how the system prompted for an update to Java 9. Choosing "yes" does a rolling upgrade with no downtime.

This demo was used to announce that Oracle offers Container Native Application Development.

Much of Oracle's Cloud Native offering comes from Wercker, an acquired company that has a container pipelines cloud service. Not only does it offer CI/CD, but it also provides container deployment to Kubernetes and Registry so you can have managed Kubernetes and Registry on Oracle Cloud's infrastructure. Oracle recently joined the Cloud Native Computing Foundation. Kothari also mentioned microservices and serverless functions, a future application development platform.

Next up was a talk by Arun Goel to talk about Cloud Security: Protect and Monitor. Goel is a director, Product Management, Identity and CASB, Oracle Cloud Platform, the security component of Oracle’s cloud offering.

CASB shows risk events and risky users based on a score. You can see where a user is trying login from (e.g., all over the world).

Suhas Uliyar took the stage next to talk about Next-Gen Engagement: Rise of the Bots.

Uliyar went slideless to demo bots as a conversational UI. The demo app returned to the theme presented earlier - purchasing of a used car. He used Facebook Messenger to talk to the bot. He uploaded an image, and an AI program processed the image. It figured out the model, year, and odometer and returned that information, along with a picture of the car, to the chat window.

The bot then prompted "Do you want to see more about the car history?" If you selected "Yes," it gives previous owners, known recalls, and accident history. Next, it prompts "Are you thinking of purchasing it?" If yes, "Do you want to hear my price recommendation?" If yes, it detects your zip code and says there are similar cars in your area. "Do you want to see them?"

Bot builder is part of the Oracle Cloud Platform. To develop a bot, you there are three different parts:

• Part 1: configure and train MLU engine
• Part 2: conversational design of dialog
• Part 3: integration (front-end channels, e.g., Messenger) and integrating backend

You start this process by adding an intent. There were two intents in this example: Greeting and Fetch License Plate number.

Then you do Entity extraction. It tries to identify things like "Tesla" and "tomorrow." They're basically data types. Others include address, currency, date, duration, email, number, and phone number. As a developer, you don't have to code that. This is automatically taken care of by the MLU engine.

Uliyar showed how this is all driven by a YAML-based scripting language. You add a service that ties into your microservice or function call. The final part is configuring the channels. You can do this with webhooks. There's a lot of differences between channels.

At the end of his demo, Uliyar announced a few things:

• Conversational AI Platform. Where you can enhance experiences with conversational interaction.
• Blockchain Cloud Service
• Big Data Platform Enhancements

He also showed some stats about Oracle Cloud's Momentum.

To close the keynote, the founder of the DevOps Movement, Patrick Debois CEO, Zender.tv was invited to the stage. His talk was titled "Serverless to ServiceFull, How the mindset of DevOps is evolving." You can find his slides on SlideShare.

His talk mainly revolved around that many companies are "ServiceFull." They buy existing services instead of building their own. Netflix does it with AWS. Slack does it with Google. Serverless is the next step from virtual machines and containers. You deploy one function at a time. AWS Lambda, Azure, Oracle now has one.

Oracle now has its own serverless offering as well in the Fn project. Fn is a container native serverless platform that you can run anywhere. It’s Apache-licensed, written in Go, and was open sourced on stage during the Java Keynote at JavaOne.

Summary and Thoughts

Oracle Code and the Oracle Developer Community suggest that Oracle recognizes that developers are the new Kingmakers.

This keynote was enjoyable to watch because Oracle demoed some impressive apps and products. Granted, many of them came from companies they acquired, but for a company that has historically struggled to develop good products for developers in-house, purchasing other companies actually established credibility.

Oracle cloud has some compelling looking offerings, but is coming late to the market and may struggle against more established competitors such as Cloud Foundry and AWS.