What’s New #6

Hello, and welcome to our second Deep Algo update of 2017.
This time, we solved a number of long-standing issues with our search engine. Here’s the menu: more relevant search results, presented in a revamped user interface. Oh, and a new homepage!

New homepage
Check out the new home page of the Poker project: instead of the tree of algorithms, we now provide a friendlier starting point for the exploration of the project.
First, we report a few key statistics about the project: which programming languages are used and the volume of code that Deep Algo analyzed. We also provide the date of the last analysis, which is useful for users who run Deep Algo regularly on an active project. We then provide the number of business concepts that were extracted by Deep Algo. The most relevant of those are available as shortcuts to start a search.

The second part of the homepage provides a selection of interesting algorithms to start your explorer’s journey. Pick one from the first column (Most used) to discover a usually small and simple algorithm that is called in many places of the project. For example in the Poker project, “How to create action?” is a simple constructor. It initializes an “Action” object and contains almost no business logic, other than moving some data. If you feel more adventurous, pick an algorithm from the second or third lists (Most business-oriented or Most important concepts). You should be taken to lengthier and more complex algorithms that contain most of the business logic of the application. Still in the Poker project, “How to manage BET, CALL, CHECK, RAISE, and action from client with act?” contains the artificial intelligence of the computer-controlled opponents, the “BasicBot”s.
But what if you already know what you’re looking for? I can see you came with scribbled notes about that bug that popped up in production just yesterday. Don’t worry, you won’t need to study a worn-out treasure map to find it! Indeed, we beefed up Deep Algo’s search engine just for you.

Relevant results and suggestions
Our new Deep Algo release ships with major relevance improvements in the search engine. For example, we often spoke of the Chen score, the algorithm that is used by the Poker game to evaluate a player’s hand. Wondering how much a Queen contributes to the Chen score?
Just search for “chen score queen” and click on the first result.

Our goal with this release was to make sure that when you look for something, it’s the first result that comes out. We’ve done a lot of progress in that direction, but we may still have some adjustments to make in the future, so we’re always glad to hear your feedback.

Now, the second improvement: let’s imagine that you didn’t know about Mr. Chen and his poker formula, and mistyped your query: “chain score queen. Or maybe you haven’t had coffee yet and wrote “chnscor quenn”. In those cases, Deep Algo has your back and can suggest alternative queries to keep you going in the right direction.


As you can see, we replicated the well-known behaviors of common search engines, in order to make our search engine as intuitive and helpful as possible.

Redesigned user interface
Search results are now presented in the same way when you search in the whole project or in a single algorithm. Most importantly, search results are always grouped by algorithm. The parts of the text that match are highlighted, to help you understand why we provide each result and which one to choose. Finally, the logos on the right of each result indicate clearly the kind of IOCA node that was matched by your query.


This release was a big step forward for the search, and it was possible thanks to feedback from our beta users. Now, thanks to the same feedback, we’re focusing on improving our algorithm diagrams, to focus on the visualization of data flows. Exciting times ahead!