"Light" and "heavy" igrostroy: the experience of Dmitry Degtyarev, COO Inventain

Field of game design today is one of the fastest growing in the world. The average annual growth of the gaming industry sostavljajut 4,8% and it is expected that by 2020 the market value will reach 90 billion U.S. dollars.

Largely this market is "fueled" mobile projects. The number of mobile users increases — this statista.com in 2017 their number will reach 2.32 billion — therefore, increasing the cost of the mobile gaming market.

Not long ago, I became interested in this subject, but it is to the full podcast on game design I far. To begin acquaintance with this industry I decided on the example of one existing and successful projects — with their approaches to working with technology shared with me Dmitry Degtyarev, COO Inventain.

Disclaimer: given the fact that we are talking about the example of certain companies, I decided to put this material in the hub "I PR".

/ foto office

Inventain is a small and mobile company with a narrow professional circle of experts. This allows not to be distracted by routine and to focus on creating products.

Now the company is actively developing two areas of activity: game and start-up direction. In the first case we are talking about gaming products, and the second on apps for people with gamification elements (without it is often difficult to just show the user something new and to give him "visual jolt").

One of the main problems of many services is the complexity in development. Gamification allows us to represent the process of development of the service as the game calculate where the user is required effort, what is its purpose, what can we do to reward him for his labors. Similar methods are used for customer retention: when it needs to return to the service, for what purpose, as he finds out about it and so on.

In addition, made a substantial bet on the artistic component and the user experience from the viewpoint of UI convenience and understanding, where the person uses the product (for example, starts the game in public transport). All this cannot exist "in the air" — the work of designers and designers must be implemented on the optimal technology stack, but in the end, it is necessary to publish the finished product and "deliver" it to the audience.

the

Development and infrastructure


The company Inventain application development is conducted with the use of the Swift programming language and the Unity engine. Unity 3D is used to create game projects and was selected due to its simplicity, functionality and affordability. Also at the moment of choice, the price was more loyal than the Unreal Engine. Game engine Unity is far more popular amongst developers than any other software. For example, 34% of the top 1000 free mobile games developed in Unity.

As for Swift, he was selected for "non-game" projects. The popularity of Swift as Unity 3D, is also growing — in 2017 he first entered the ten most used programming languages, according to the Tiobe index. In this case, Swift has replaced Objective C because of its functional benefits (short syntax and performance).

One of the most important approaches, which ensure the "portability" and "ease" of development, is to use ready-made tools. Before you take on any difficult task, the team conducts analysis of existing solutions (including surcharge).

Cheaper and faster to get something ready than himself to spend the time and effort, reinventing the wheel. The ultimate goal for the team is the product, path should choose the best
— Dmitry Degtyarev, COO Inventain

Unity has its own Asset Store platform, which includes thousands of ready-made solutions and resources. For Swift you can use open solutions which are on sites like Github. In this case, the implementation should not generate in the format of "one on one" any public decision can take into account all the features of your product, they need to be modified for a particular application.

In addition, you should pay attention to the materials that go on the topic in the relevant blogs. An example is code to implement linear regression. Such notes can be used to search a fresh look at a particular problem, and then return to implement their own solutions.

Another good solution is to create a private pool of practices that are used from project to project. You can start with a simple wiki.

If to speak about testing, the scale a company like Inventain (and similar projects), it makes no sense to create a whole unit. Here, only one expert on QA, which will ask the vector. The realization of this task should be given to developers who will try to achieve a balance between code coverage tests. Priority — the care of time-consuming routine actions of a tester and eliminate potential mistakes due to the automation. Thus, a huge number of scenarios and possible errors in the "far corners" of the application will be always under control.

The desire not to spend the effort on routine work affects the organization's infrastructure — it cloud (AWS), although the vector to the cloud was not chosen immediately. The cloud is much easier to manage. In the case АWS don't need any specialized data infrastructure — enough to see the very complete and well structured documentation. It turns out that all I can manage engineers and don't need system administrators. In addition, with the cloud it is much easier to adapt to the real load and add (remove) servers dynamically, not overpaying for a simple iron. All of this translates into real benefits in cost and flexibility pay only for what you in fact use.

the

"Heavy" should only be the specifics


To keep pace with the evolution of modern technologies and use in their products solutions, which are gaining popularity, you need to spend the time and effort. This fee for the opportunity to compete with the best products and teams in the industry.

In this case we are talking about the accumulation of expertise in areas such as machine learning and motion capture. If we talk about specific examples, the team Inventain can boast that they taught the neural network to recognize human faces, and a similar three-dimensional avatar. In the context of machine learning and neural networks is the implementation of the transfer of the picture styles to other objects plus the ability to cast on behalf of users of the virtual mask and keep them in motion.

Such "customization" is impossible solely due to the search for open projects and "quick" solutions, but it is the "core" of the product and justifies a significant time investment.
Article based on information from habrahabr.ru

Популярные сообщения из этого блога

Approval of WSUS updates: import, export, copy

Kaspersky Security Center — the fight for automation

The Hilbert curve vs. Z-order