DEV Community: Gustavo Novaro

The Importance of Including a Static Analyzer in Your Code

Gustavo Novaro — Thu, 04 Apr 2024 14:04:30 +0000

In software development, code quality is a critical factor for project success. That's why more and more development teams are turning to static analysis tools to ensure their code meets quality standards and is more maintainable in the long run. Two popular tools for this purpose are SonarQube and PhpStan, although there are also important options for Python, such as Flake8 and PyLint.

What is a Static Analyzer?

A static analyzer is a tool that examines source code without the need to execute it. It uses a set of predefined rules to detect potential errors, vulnerabilities, bad practices, and opportunities for improvement in the code. By identifying these issues early in the development cycle, static analyzers help prevent costly errors and improve overall code quality.

SonarQube: Comprehensive Analysis

SonarQube is an open-source platform that provides comprehensive code analysis across multiple programming languages. It uses a variety of metrics to assess code quality, including cyclomatic complexity, code duplication, test coverage, and compliance with coding standards. Additionally, SonarQube offers detailed reports and recommendations for code improvement, making it easier to identify and address issues.

PhpStan: Specific Analysis for PHP

For projects developed in PHP, PhpStan is an invaluable tool. PhpStan performs static analysis of PHP code to identify potential type errors, undefined variable usage, calls to non-existent methods, and other common pitfalls. By providing accurate insights into the health of PHP code, PhpStan helps developers write more robust and secure code.

Static Analyzers for Python

In the Python ecosystem, there are several popular static analysis tools. Flake8 combines multiple tools (including PyFlakes, pycodestyle, and McCabe) to offer comprehensive analysis of Python code, focusing on readability and style. On the other hand, PyLint is a highly configurable tool that detects programming errors, style conventions, and other potential issues in Python code.

Conclusion

Integrating a static analyzer like SonarQube, PhpStan, Flake8, or PyLint into your software development process can have a significant impact on the quality and maintainability of your code. These tools provide immediate feedback on potential code issues, allowing teams to address errors before they become major problems. Ultimately, investing in static analysis is a fundamental practice for ensuring the quality and reliability of the software we develop.

How to setup your own ChatGPT with OpenLLaMA

Gustavo Novaro — Mon, 04 Mar 2024 12:35:24 +0000

Do you want to have your own ChatGPT? Do you care about privacy? You don't want to pay for the premium version? or you want to use uncensored models. Or even train a model with your company documents.

For these reasons it is a good idea to set up your own LLM model server.

This is an example:

What is OpenLLaMA?
LLama 2 is an LLM (Large Language Model Meta AI)
LLama Meta

What do we need?

Install OpenLLAMA, OpenWebUI and Docker

Installing Ollama

Ollama have support for Windows, Linux and Mac
I recommend minimum 16Gb of RAM.
https://ollama.com/download

For linux setup

curl -fsSL https://ollama.com/install.sh | sh

Download models

ollama pull llama2-uncensored

Run model

ollama run llama2-uncensored

Install Open Web UI

Before install OpenWebUI we need to have installed Docker and running.

docker run -d --network=host -v open-webui:/app/backend/data -e OLLAMA_API_BASE_URL=http://127.0.0.1:11434/api --name open-webui --restart always ghcr.io/open-webui/open-webui:main

docker run -d -p 3000:8080 -v ollama:/root/.ollama -v open-webui:/app/backend/data --name open-webui --restart always ghcr.io/open-webui/open-webui:ollama

Beyond UUIDs: The Rise of ULIDs for Enhanced Sorting and Efficiency

Gustavo Novaro — Wed, 17 Jan 2024 16:44:29 +0000

In the contemporary digital landscape, the efficient generation of unique identifiers is crucial, particularly in distributed systems where traditional databases with auto-incrementing keys show their limitations. ULIDs (Universally Unique Lexicographically Sortable Identifiers) shine here, offering a robust and efficient solution.

ULIDs provide a 128-bit structure, split into a 48-bit timestamp recording time in milliseconds and an 80-bit randomness portion. This combination ensures uniqueness and lexicographic sorting, facilitating indexing and data retrieval.

Here’s how it looks:

ulid() // 01ARZ3NDEKTSV4RRFFQ69G5FAV

ULID generates 2 parts:

Timestamp: int of 48 bits with UNIX timestamp in milliseconds
Random: Random string of 80 bits

Conclusion

ULIDs outperform traditional UUIDs in systems where write efficiency and sortability are critical. Their design makes them ideal for modern applications requiring scalability and peak performance.

ULID specification

For implementations and practical examples, the developer community offers a variety of libraries across different programming languages, ensuring that ULIDs can be seamlessly integrated into any new or existing project.

Install k9s in Debian / Ubuntu

Gustavo Novaro — Sun, 14 Jan 2024 20:42:59 +0000

If you don't know K9s, it is a Kubernetes CLI manager, one of my favorite tools to use from the console.

This is an example screenshot.

To install it from Ubuntu, it is best to download the .deb binary from the releases page

Or with wget from the console like this:

wget https://github.com/derailed/k9s/releases/download/v0.31.5/k9s_linux_amd64.deb

For setup run

sudo dpkg -i k9s_linux_amd64.deb

Enjoy!