Development

Greetings, dear readers! In this article, we will take a detailed look at all aspects of searching for information about a person by photo, paying special attention to three main areas. We will discuss the capabilities of search engines, which many underestimate, although they can provide surprisingly accurate data when used correctly. We will analyze specialized online services based on neural networks and complex facial recognition algorithms. Special attention will be paid to Telegram bots, which combine the convenience of mobile use with powerful search technologies. These tools will help you obtain data about a person, including their name, phone number, and other useful information.

You will learn not only about the technical capabilities of each method but also about the practical nuances of their application. We will explain how to achieve maximum accuracy in data search and what photo parameters affect the result.

When OpenAI first showed Sora, the reaction was instantaneous: "that's it, the era of video production is over." But that was just a demonstration. The technology itself remained inaccessible to a wide audience — until the appearance of Sora 2, an updated version where any user can create up to 100 videos a day for free.

In short, Sora 2 is not just a "text-to-video generator." It's a full-fledged multi-agent system where several OpenAI models work under the hood: from text analysis and scene staging to motion and light synthesis. Unlike early versions of Runway or Pika, the quality here is almost cinematic: Sora understands context, follows the frame's composition, and maintains visual integrity between scenes.

Сервис по генерации музыки Sferoom AI - эдакий прототип и двойник Suno. Испытав его, я остаюсь при своём - выбираю Suno.

Тут не выведение на чистую воду и не обличение, а исследование того, насколько удобно Suno вмонтирован в Sferoom

Почему я сравниваю именно эти две подписки? Потому что под капотом у Sferoom всё тот же знакомый нелицензионный движок Suno v4.5+, предоставляемый через посредство kie.ai - нелицензионного поставщика API Suno.

В феврале 2026 выбранный мной дистрибьютор песен и музыки Sferoom добавил на свой сайт возможность генерировать свои треки с помощью встроенного ИИ. То есть сейчас на одном русскоязычном Интернет-сайте можно и сделать песню, и подготовить публикацию на стриминги, и собрать статистику прослушиваний.

Мог ли я проигнорировать этот факт? Конечно, нет.

Должен ли я докопаться до самой сути? Конечно, да!

В статье разбираем функции ИИ-генератора песен и музыки Sferoom AI и считаем плюсы и минусы.

Greetings, dear readers! We continue to test Telegram bots that can effectively help us with the task of finding target data in open sources, or simply put, in OSINT. In this article, we will look at more worthy candidates to become permanent search tools without leaving the well-known messenger.

Disclaimer: All data provided in this article is taken from open sources. It does not call for action and is provided for informational purposes only, and for studying the mechanisms of the technologies used.

In my previous Russian-language article on Machine Learning as Alchemy, I discussed the possibility of discovering novel solutions without relying on GPUs or expensive computing clusters. In this article, I will share my experiments with continual learning and the compositionality of thought using micro-neural networks, and explain what the philosopher Lev Vygotsky has to do with it all.

I've been practicing Zettelkasten for the past five years and still haven't found anything better than Niklas Luhmann's method. The problem is that Obsidian doesn't support it out of the box, so I had to write a plugin for organizing notes as close to Luhmann's original method as possible.

I spent a couple of weeks digging through Luhmann's original archive before I understood how it actually works. Structurally, the archive resembles a table of contents, but with the difference that a note can be inserted at any point, adding nested chapters.

You’ve probably had this happen too: a service runs smoothly, keeps users happy with its stability and performance, and your monitoring stays reassuringly green. Then, the next moment — boom, it’s gone. You panic, dive into the error logs, and find either a vague segfault or nothing at all. What to do is unclear, and production needs saving, so you bring it back up — and everything works just like before. You try to investigate what happened, but over time you switch to other tasks, and the incident fades into the background or gets forgotten entirely.

That’s all fine when you’re on your own. But once you have many customers, sooner or later you start feeling that something isn’t right, and that you need to dig into these spikes of entropy to find the root cause of such incidents.

This article describes our year-long investigation. You’ll learn why PostgreSQL (and any other application) can crash because of a bug in the Linux kernel, what XFS has to do with it, and why memory reclamation might not be as helpful as you thought.

The Mandelbrot set. And it's a program! I made it in g++, a freely distributable C++ compiler. Read it! Very interesting. Using OpenMP, you do parallel programming at the multithreading level. And I decided - this would be a completely different level of quality! I implemented honest supersampling (antialiasing) - with 8x8 antialiasing (64 passes per pixel!!!) That is, not 1920 by 1080 pixels, but 8x8 more! 15360 by 8640 pixels! And then these 64 passes reduce by one pixel, but smoothly - and no longer 8-bit, but 24-bit TrueColor!

A utility for supporting Xray on Keenetic routers — Xkeen.
All the code is written in pure shell and is open source on GitHub.

Builds Xray for your Keenetic on Entware.
GeoIP and GeoSite from AntiFilter, AntiZapret, and v2fly are available.

Automatically updates Xray, GeoIP, and GeoSite at a specified time.

What is a sequence diagram? What does it consist of? Where and how to use it?

Sequence diagrams in simple terms.

And there are also interesting features here that you might not have known about.

The Chinese neural network DeepSeek has recently gained popularity. Today, we'll figure out how to work with its API, how to pay for it from Russia, and whether there are any difficulties.

A Telegram Mini App is a web application built into the messenger that allows users to access a service without installing separate programs. This allows for a convenient service integrated with the messenger's ecosystem, whether it's an online store, a booking system, etc.

We have written a general overview of this application, its capabilities, the first steps to creating one, and existing examples.

In this article, we will look at how you can customize Windows, adjusting the appearance and functionality of the operating system to your preferences. This can be done using both built-in tools and third-party programs.

When customizing Windows, it's important to remember that most third-party programs make changes to system files and the registry. This can lead to malfunctions, conflicts between utilities, and problems after operating system updates.

List of utilities that will be covered in this article: PowerToys, WinAero Tweaker, ExplorerPatcher, Rainmeter, TranslucentTB, EarTrumpet.

By the way, if you are interested in IT technologies and useful tools, check out my Telegram channel. I regularly post fresh news, life hacks, and other useful things there. And now, to the heart of the article.

Image generation by neural networks has become a 'regular button' alongside familiar design tools. Today, you can create an image from a description in Russian, right in your browser, often without registration and, importantly, for free. Such a 'free image generator' is useful not only for designers: entrepreneurs create product cards and hero banners, SMM specialists create ad creatives and stories, journalists and bloggers create illustrations for their materials, and developers create interface prototypes and game mockups.

Why has this topic become so popular?

A brief and clear description of the RAG (Retrieval Augmented Generation) approach for working with large language models.

> As of April 10, 2024,
> three months later,
> this article has been blocked by Roskomnadzor (RKN)
> within the territory of the Russian Federation.
> It has also been removed from the web archive archive.org.
> The article on Habr remains accessible from IP addresses in other countries.
> Yes, now to read about VPNs, you need a VPN.

Against the backdrop of last year’s escalation of censorship in Russia, the articles by MiraclePTR were a breath of freedom for many Russian-speaking IT folks. I want to open the door to free information a bit wider and invite “non-techies” (“dummies”) who want to spin up a personal proxy server to bypass censorship but feel lost in the flood of information or got stopped by a confusing technical error.

In this article I’ve described a universal solution that provides transparent access to the global internet bypassing censorship, uses cutting-edge traffic obfuscation, doesn’t depend on the will of a single corporation, and most importantly has ample “safety margin” against interference from censors.

This article is aimed at “dummies” unfamiliar with the subject area. However, people “in the know” may also find something useful (for example, a slightly simpler setup for proxying via CloudFlare without having to run nginx on a VPS).

If you still don’t have a personal proxy to bypass censorship—this is your sign.

I constantly have multiple Chrome windows open, each with a pile of tabs. One window has a dozen slow analytics queries that I'll check "any minute now." Another has my research on $lookup and $unwind in MongoDB. The third one — with the most tabs — has local school enrollment rules, because life.

Tab-saving extensions have existed for ages, but reviews regularly complain about data loss — so why not build my own, designed specifically for backup? Every Chrome user already has a Google account, usually with sync enabled.

Here's how I used it to make Tab Saver extension that backs up to your Google Account without any signups and with zero configuration needed. It's pretty straightforward - you can easily use it as well.

Much attention is currently focused on the size of neural networks and the gigawatts of power consumed by data centers. However, the future lies not only in giant clusters but also in tiny chips embedded directly into the sensing elements of hardware. When a neural network is placed directly inside a sensor chip, it must be exceptionally efficient.

Through experimentation, I have successfully built a neural network architecture with 380 parameters (with potential for further reduction), capable of operating in conditions considered unsuitable for conventional algorithms.

Asynchronous I/O, ML-based query plan optimization, and built-in connection pooling are among the key features of the new Postgres Pro Enterprise 18. This release brings together the capabilities of the vanilla PostgreSQL 18 core with Enterprise-grade tools for working with large-scale data. Today we will walk through the technical details, new index scanning strategies, and mechanisms for scaling write workloads.

Do you switch between Linux and Windows in dual-boot? Then you're probably familiar with this problem: you have to reconnect all your Bluetooth devices every time. Headphones, mouse, keyboard, gamepad — everything has to be reconnected.

It's scary to even think about it:
3 devices × 90 seconds × 3 switches per day × 250 days = 56 hours wasted per year.

I spent a month solving this problem and wrote BlueVein — a utility for automatically synchronizing Bluetooth keys between operating systems.