This article is the most comprehensive description of signals and slots in QML compared to all previous articles on this site.

In this article, I will try to explain the following when working with Qt/QML + Qt/C++:

• ways to declare signals and slots, also called methods in the C ++ class, which will be registered in the QML layer
• ways to connect to signals of classes declared in C ++ as context
• work with Q_PROPERTY, which also requires signals and slots
• ways to connect signals and slots in QML
• etc.

To speed up the site, in addition to optimizing database queries, you can use caching.

Django allows you to cache:

• individual view , both Class Based View , and ordinary functions view
• whole templates or parts of these templates
• QuerySet
• as well as properties of model objects using cached_property

I was interested in the ability to cache individual properties of model objects for heavy computing or heavy database queries.
The cached_property decorator has such a functional, but the drawback for me was that caching occurred only for the lifetime of the object.
Whereas I need caching for a longer period of time than the existence of an object when requesting a page. And also I needed to cache properties depending on the input arguments. This decorator on the site caches the number of likes and dislikes, as well as information about whether the current user liked a particular content object.

Thus the decorator model_cached_property was written

This will be the first article on the use of a set of established utilities that are present in the core of the EVILEG site. This part of the site’s program code is in open source, so I want to talk about what functionality is present in this battery and how to use it.

I'll start by using Google reCAPTCHA , and what is required to quickly implement this functionality.

Introducing the new release of EVILEG-CORE 0.0.5.

Код, на котором основана эта статья, находится в стадии разработки, с различными коммитами на рассмотрении в теме «Аллокатор». Обратите внимание, что в коде используются различные функции C++17.

Несколько месяцев назад разработчики Qt Company работали над исправлением того, как QHostInfo отправляет результаты вызывающей стороне, что привело к небольшой оптимизации того, как выделяется память для аргументов в QMetaCallEvent. Разработчики используют объекты этого типа в соединениях сигнал/слот с очередями, и некоторое время, проведенное в этом коде, напомнило, что в Qt есть схема размещения довольно маленьких, недолговечных объектов в куче, а также в производительности критического пути кода.

Поэтому они стали задаваться вопросом, можно ли разместить экземпляры QEvent из выделенного пула памяти, используя специальный аллокатор.

Recently I have devoted a lot of time to website optimization and now I would like to talk about it.
This article will explain the use of the select_related and prefetch_related methods in QuerySet, as well as their differences. I will also try to explain why Django is considered slow, and why this is still not the case. Of course, Django is slower in many ways than the same Flask, but at the same time in most projects the problem is not in Django itself, but rather in the absence of optimizing database queries.

Therefore, let's optimize the EVILEG website forum page . And the Django Silk battery will help us with this, which serves to measure the number of queries to the database, as well as measure their duration.

В третьей части серии статей о графике Qt рассмотрим, как обрабатываются шейдеры в Qt Quick в версии Qt 5.14 при переключении графа сцены на рендеринг через QRhi и Qt Rendering Hardware Interface (аппаратный интерфейс рендеринга Qt). Охватим обработку шейдеров, прежде чем углубляться в сам RHI, потому что приложения Qt Quick, использующие элементы ShaderEffect или нестандартные материалы (custom materials), должны сами предоставлять код фрагмента и/или вершинного шейдера, и поэтому они должны знать (а в Qt 6 перейти на) новый подход к обработке шейдеров.

Our project uses a very large number of resources, both ordinary icons and svg images. At some point, the project stopped assembling on developers' PCs with a small memory capacity of 16 GB.

The problem turned out to be that when compiling resource files, all available memory is uncontrollably consumed, and if it is not enough, then the assembly fails.

After about 5 months of refinement of the site, I release a more elaborate version of the evileg_core package.

This release includes both new functions and utilities, and the old functionality has been finalized. For example, Markdown field behavior has been improved. Initially, EMarkdownField drastically reduced content rendering performance, but the problem was identified and resolved.