Development

Every time when the essential question arises, whether to upgrade the cards in the server room or not, I look through similar articles and watch such videos.

Channel with the aforementioned video is very underestimated, but the author does not deal with ML. In general, when analyzing comparisons of accelerators for ML, several things usually catch your eye:

• The authors usually take into account only the "adequacy" for the market of new cards in the United States;
• The ratings are far from the people and are made on very standard networks (which is probably good overall) without details;
• The popular mantra to train more and more gigantic models makes its own adjustments to the comparison;

The answer to the question "which card is better?" is not rocket science: Cards of the 20* series didn't get much popularity, while the 1080 Ti from Avito (Russian craigslist) still are very attractive (and, oddly enough, don't get cheaper, probably for this reason).

All this is fine and dandy and the standard benchmarks are unlikely to lie too much, but recently I learned about the existence of Multi-Instance-GPU technology for A100 video cards and native support for TF32 for Ampere devices and I got the idea to share my experience of the real testing cards on the Ampere architecture (3090 and A100). In this short note, I will try to answer the questions:

• Is the upgrade to Ampere worth it? (spoiler for the impatient — yes);
• Are the A100 worth the money (spoiler — in general — no);
• Are there any cases when the A100 is still interesting (spoiler — yes);
• Is MIG technology useful (spoiler — yes, but for inference and for very specific cases for training);

Privet, comrads!

In this article i’ll show how easy it is to setup Spring Java app with Kafka message brocker. We will use docker containers for kafka zookeeper/brocker apps and configure plaintext authorization for access from both local and external net.

Link to final project on github can be picked up at the end of the article.

There probably is no way one who stores some crucial data (and well, in particular, using SQL databases) can possibly dodge from thoughts of building some kind of safe cluster, distant guardian to protect consistency and availability at all times. Even if the main server with your precious database gets knocked out deadly - the show must go on, right? This basically means the database must still be available and data be up-to-date with the one on the failed server.

As you might have noticed, there are dozens of ways to go and Patroni is just one of them. There is plenty of articles providing a more or less detailed comparison of the options available, so I assume I'm free to skip the part of luring you into Patroni's side. Let's start off from the point where among others you are already leaning towards Patroni and are willing to try that out in a more or less real-case setup.

I am not a DevOps engineer originally so when the need for the high-availability cluster arose and I went on I would catch every single bump on the road. Hope this tutorial will help you out to get the job done with ease! If you don't want any more explanations, jump right in. Otherwise, you might want to read some more notes on the setup I went on with.

Introduction and Context

So you’ve just finished developing the next mobile gaming smash hit. You’ve already planned to launch your game worldwide and win the hearts of gamers everywhere. But what you haven’t done yet is think about how you’re going to make any money from your success and hard work. Of course, you’ve imagined becoming an instant millionaire, but you just don’t know how to get there.  

Fear not, intrepid developer! Instead of just rehashing the tired headlines and beating the same old drum, we’ve tailored our Ad Monetization write-up for developers who are gearing up to publish their game or app in China. We’ve scoured the internet, grabbed insight from ad giants like Google, Smartyads, and PocketGamer.biz and combined it with our own hands-on knowledge of what works in China—knowledge that we have gained from the past  5 years in the industry.

We need to briefly explore how monetization functions as a utility within the context of game design. This is especially true when examining how developers in China have made meaningful design decisions that have both shaped and been shaped by user habits. Unpacking the utility of Chinese game monetization will help developers understand why Ads Monetization or Hybrid Monetization strategies have become so popular, especially in China.

Moving from monetization design, we’ll examine some of the best practices regarding Ads Monetization. Finally,  we’ll turn to specific case studies in the Chinese market that show how powerful Ads Monetization, specifically Rewarded Videos, can be in a well-balanced monetization strategy. 

Designing Monetization as a Utility

Example with JSONDecoder

struct Test<T>: Codable where T: Codable {
    enum CodingKeys: String, CodingKey {
        case value
    }
    
    let value: T
    let info: String
}

extension Test {
    init(from decoder: Decoder) throws {
        let container = try decoder.container(keyedBy: CodingKeys.self)
        self.value = try container.decode(T.self, forKey: .value)
        self.info = "Default init(from decoder:)"
    }
}

extension Test where T == String {
    init(from decoder: Decoder) throws {
        let container = try decoder.container(keyedBy: CodingKeys.self)
        self.value = try container.decode(T.self, forKey: .value)
        self.info = "Custom init(from decoder:)"
    }
}

let data = #"{"value":"Hello, World!"}"#.data(using: .utf8)!
let object = try? JSONDecoder().decode(Test<String>.self, from: data)
print(object.debugDescription)

Optional(
    Test<String>(
        value: "Hello, World!", 
        info: "Default init(from decoder:)"
    )
)

Recently Epic Games’ Fortnite was removed from the Apple App Store and Google Play Store. The main reason being Epics bypassing of in-app purchases instead of using the officially sanctioned system for both platforms. While it is still possible for you to download Fortnite directly, this large scale case brings to light the duopoly of Apple and Google in the mobile market.

For most developers, these two stores account for almost all of their revenue and userbase. While Epic Games will be fine to go without, for the time being, what about the other 99% of developers who rely on these two stores for distributing and monetizing their apps. In this article, we’ll provide some of the alternative stores available for both developers and consumers for finding or distributing apps.

For one reason or another, you may have found yourself wondering, ‘where are some other places I can go to find and download new apps and games for my mobile device’? Or ‘are there any other marketplaces for me to share my apps’?

Major players in the Chinese app market are joining forces to take on the almighty Google Play store. Xiaomi, Oppo and Vivo are reported to launch the Global Developer Service Alliance (GDSA), a platform allowing Android developers to publish their apps in the partnering stores from one upload.

The GDSA is expected to launch in nine countries—including India, Indonesia, Malaysia, Russia, Spain, Thailand, the Philippines, and Vietnam—although paid app support may vary across the regions. Canalys’ Nicole Peng explains the wide reach of this alliance:

By forming this alliance each company will be looking to leverage the others’ advantages in different regions, with Xiaomi’s strong user base in India, Vivo and Oppo in Southeast Asia, and Huawei in Europe. 

The Chinese Mobile RPG: the Genre of Giants.

Chinese RPGs make up 56% of the top 500 grossing iOS games in China. According to data gathered in 2019 by Game Refinery’s Joel Julkunen. Let’s take a look at the genre and some important points for game developers and RPG enthusiasts to consider Chinese Mobile RPGs.

RPG or role-playing game is a game where the player plays as a character, often the main character, in a fictional game world.

This article presents generalized approaches for using value converters into writing of XAML code.

>> Read in Russian