Almir Mustafic’s Post

The Art of Code Reviews: How to Provide Constructive Feedback and Improve Software Quality In this article, we’ll explore what code reviews are, how they work, and share some best practices to help you conduct them effectively. We’ll also look at how AI can assist in code reviews, why constructive feedback matters, and common challenges you might face along the way. https://lnkd.in/gbPBk5n4 #publicissapient #publicisgroupe #codereview #codereviewprocess #codereviewers #sdls #softwaredevelopmentlifecycle

How to ask for Software Feedback

A crucial component of software development are test cases. They aid in ensuring that the program operates as planned and satisfies end-user requirements. Here are some aspects concerning test cases that you should be aware of. #testing #testcase #postoftheday #codeautomation #ai #likeforlike #followforfollow

The term "debugging" originated from actual bugs? In 1947, computer scientist Grace Hopper found an actual moth causing a malfunction in the Harvard Mark II computer. She removed it, taped it into the logbook, and that's how the term "debugging" was born! 🪲 Just like bugs have no place in code, errors shouldn't be a part of your software. Our QA testing services ensure a glitch-free, smooth user experience. Let's debug together and create software excellence! 🌐 🌐devccelator.com ✉info@devccelator.com 📞+46736614989 #Devccelator #qatesting #softwaretesting #MobileTesting #AITesting #tech #B2B #linkedin #innovation #linkedin #technology

In the world of software development, there is a hidden superhero that often doesn't get the credit it deserves. Logging 😎 . Troubleshooting, understanding user behavior, performance detection, and continuous improvements are a few reasons why logging is so crucial. ✅ What constitutes proper logging? 👉 Useful context: It should contain enough context to understand the situation. 👉 Severity levels: It will help prioritize and respond to issues effectively (eg: info, warning, error, etc). 👉 Structured Data: It will make analysis easier. It should not be a chaotic stream of text. 👉 Security: It should not reveal sensitive information. Exposing sensitive data (User or system sensitive) can be risky. Proper logging isn't just about writing messages to a file. It's more about crafting a narrative of your application's life. From deployment to every user interaction, logs empower you to build better, more resilient software. So, next time you hit that "log" command, remember that you're not just writing lines of code. #softwaredevelopment #logging #debugging #coding

Who Should Software Serve? "The ultimate purpose of software is to serve users. But first, that same software has to serve developers. ...As a program evolves, developers will rearrange and rewrite every part. They will integrate the domain objects into the application and with new domain objects. Even years later, maintenance programmers will be changing and extending the code. People have to work with this stuff. But will they want to?" ― Eric Evans, Domain-Driven Design: Tackling Complexity in the Heart of Software https://lnkd.in/gj3Dxcg #ddd #domaindrivendesign #refactoring #extremeprogramming #modernagile #agility #tdd #testdrivendevelopment #legacycode #continuousimprovement #maintainability #softwarequality #codequality #softwaredesign #lean

Why are code quality tools still a big mess? Let's understand it from the buyer's (VP of Engineering) and user's (developer) perspective. Firstly, why do we need them? They identify potential issues, bugs, security vulnerabilities, and violations in coding standards and best practices - thus, they are very important. Can you name a few? Sonar (SonarQube), Coverity, etc. Okay, what is the problem for the users (developers)? These code quality tools have very high false positives and are not actionable. Let's go deeper into a developer's journey today: 1. Understand the impact of the issue - such as if it's not fixed, it will break something, or increase CPU, Memory, or Latency, or introduce critical security vulnerabilities. 2. Find out the blast radius - which means understanding all the dependencies to determine if fixing it will break 10 other things or not. 3. Write the actual fix. 4. Test the fix. 5. Push and deploy. Today, we have to go through these all 5 steps every time for every code quality issue that tools like SonarQube detect, and they are generally in the thousands. It's pure technical debt and very repetitive. What is the problem for the buyer (VP of Engineering/CTO)? The VP of Engineering/CTO sees the issues in a dashboard, where the number of issues never decreases. But why? Because their team is always busy shipping features and no one has time to address these issues. However, when they start to see incidents, they make it an OKR with top priority to fix all issues, which leads to delayed feature delivery. It's a bad rabbit hole, and the worst part is you pay for it every time your code is scanned and new issues are detected but not fixed - it's like paying for insurance but not for remediation. Today, 80% of the issues that these code quality tools detect can be automatically fixed with AI paired with a deep understanding of Abstract Syntax Trees (AST) - this is what we do at CodeAnt AI (YC W24). #codequality #sonar

The latest update for #Sentry includes "Improving INP and FID with production #profiling" and "Low effort image optimization tips". #monitoring #crashreporting #devops https://lnkd.in/dCdw2JF

Do we need tests at all? There is a constant debate in software engineering around testing. For standard applications that have relatively simple business logic, it's okay to NOT spend much time on writing tests. Especially in startups where things change at cosmic speeds. However, with complex enterprise systems things are different. When architecture becomes wild and people in teams change frequently, testing should become the #1 priority. Without proper testing suit, even a single line change requires re-testing everything manually. Yet, that's not even the worst thing. Developers usually are lazy enough to ignore the testing step and those changes can directly land on production. Trust me that's the real disaster and can lead to irreversible consequences. How's the testing culture at your company? Share in the comments below 👇 #testing #integrationtesting #coding #softwareengineering

I don't like "Code Freezes." The most critical time for an online retailer to deploy code is in the middle of the holiday season when customer traffic is at its peak. That's when you lose big money with a bug, and a critical fix can have the biggest impact. Making a code push at this time takes: - Confidence - Discipline - And lots of practice But most importantly, it needs automation. That's why you need to invest in: - A fast pipeline - Infrastructure as Code - Automated testing - Feature Flags - Canary releases - Chaos Engineering - Monitoring That's it. Automate fiercely; That's the only way to move away from Code Freezes.