My channel is dominated by backend devs who find frontend more challenging while codenight is the opposite

I personally think that Backend development is much more challenging if you work on complex projects that need to support millions or billions of users, implement database sharding, implement complex algorithms and work on a system with a vast number of components in production. I think both of them have complexities to tackle, and I find the backend more satisfying to work on.

I think the main thing that makes front-end development hard to navigate is the constant introduction of new tools and frameworks that promise to be better than the other, and keeping up is a non-negotiable if you specialize in front-end. Of course, backend also has its fair share of those, but I don't think it is evolving as quickly as frontend. And also, I think frontend development is more challenging than backend development when working with CRUD apps that have all the common features we usually implement.

Today, I received an email informing me that I was shortlisted for an interview with a company in Latvia, to which I applied through LinkedIn. I was excited for a bit, and I searched the company on LinkedIn. To my disappointment, there were multiple people who posted that they worked 6 months for free without receiving a single payment from the company. They just go around "hiring" people and making them work 40 hours a week with a promise of pay that would never arrive. Make sure to research the client/company that you want to work with. At least make sure there isn't any negative review of them like this.

This is actually a really great guide. I'm glad I found it. Shoutout to @genenetise01

𝗧𝗵𝗲 𝗖𝗼𝗺𝗽𝗹𝗲𝘁𝗲 𝗚𝘂𝗶𝗱𝗲 𝘁𝗼 𝗕𝗶𝗴 𝗧𝗲𝗰𝗵 by Linda 👋

Link to the docs

Truer words were never spoken

If you guys have wondered where you could find open source alternatives for Amazon S3, here is one I have tried:
https://min.io

It allows you to store objects, get download urls with a custom expiration time and to top it all of, it has a web UI where you can see the uploaded objects. It also has a docker image available on the docker hub. Check it out.

We also got 100🔥

Coding is not Enough

For someone to be successful as a Software Engineer, the current market demands other things in addition to coding. Most people focus on their coding skills when they try to grow in their career as a Software Engineer. Although coding is a great skill to have, in today's market, it is not the most important one. The AI models that exist today can write good code as long as they are properly guided in the right direction. The following skills are great to focus on moving forward.

1. Architecture

Architecting systems for performance, scalability, security, and user experience is a very important skill to develop. Anytime we think of building a project, we should first think of the architecture. Planning should take most of our time rather than coding. We should not start coding unless we know exactly what we are going to do, at least at a high level.

2. Tools and Technologies

There are many tools and technologies out there that make our lives easier as Software Engineers. We should read about these tools and understand them. Understanding the tools and technologies we use would help us make decisions about which tool or technology aligns more with our goals whenever we plan our projects.

3. Communication and Collaboration

As Software Engineers, we have to communicate and collaborate with clients, managers, designers, and cross-functional team members. For this communication to be effective, we have to improve our skills of speaking and writing clearly in a way that is unambiguous. We should learn to explain technical things to non-technical people in a way they can understand. We should learn to write clear documentation that is very easy for our team members to understand. We should also learn to ask questions whenever we don't understand something. We should avoid assumptions and communicate clearly.

I would love to hear your thoughts on whether there is anything I missed or if there is anything you want to add.

came across this service which provides thousands of models just for free with unified single API just like the vercel ai-sdk. downside: only 1 concurrent request, but no restriction on usage, and tbh fast inference given that you got unlimited requests. you can access 141k models (including gpt 4.1 and many multimodals) on the free tier. they got both python and js sdk. cool stuff to experiment with

models: https://bytez.com/models
docs: https://docs.bytez.com/model-api/docs/welcome
#resources

https://www.linkedin.com/posts/a2sv_a2sv-education-generation7-activity-7388960222482079744-6DrE?utm_source=share&utm_medium=member_desktop&rcm=ACoAAD7wiqoB3AP7D1_JY_SDhbFUN4lWRrmeKDI

For anyone who is interested A2SV has opened the G7 applications phase

If you want a simple LLM framework that allows you to build AI workflows and agents, here is one I recently found:
https://github.com/The-Pocket/PocketFlow

The framework itself is literally just 100 lines of code. And the repo contains links to a great tutorial video and a link to their docs.

It even has a .cursor_rules directory which explains the framework and how to use it to your coding AI agent. I hope you'll find it useful.

So basically, let’s start from what authentication is…

Authentication simply means verifying or proving the identity of a user before giving them access to a resource.

In the real world, it’s like showing your ID card when entering a restricted place like a university. The gatekeeper checks two things:

If the ID is valid and issued by the university.

If it’s not expired meaning it’s still active.

If both are true, you’re allowed to enter.

Now, in the digital world, this same idea applies but digitally. There are many ways to authenticate users, and the two most common ones are Session-based authentication and JWT (JSON Web Token) authentication.

Let’s talk about JWT, since it’s one of the easiest to implement.

In any digital product, there are two main sides:

- The Client (the browser or app that users interact with).

- The Server (where we store all our data — posts, images, user info, etc.).

Let’s imagine a social media website.
When someone wants to post, like, or edit something, we need to know who the user is so others can’t edit or delete someone else’s posts. That’s where authentication comes in.

So, how do we verify a user’s identity?
👉 We use JWT Authentication.

JWT stands for JSON Web Token it’s basically a digital ID in the form of a long string that holds user information (like ID, username, and expiration time).

Here’s how it works step by step:

When a user signs up, they enter an email and password.
We save those details in our database on the server.

When they log in, we check if the email exists and if the password matches.

If it’s valid, we generate a JWT token containing the user’s info (like ID and username).

We then encrypt this token using a secret key and send it to the client.

This makes sure only we (the server) can read and verify it later.

Whenever the client makes a request (like fetching their posts), they send this token along with it.

The server decrypts the token using the secret key and:

- Checks if it’s valid.

- Checks if it’s expired.

If all is good, it uses the user’s ID from the token to fetch and return their specific data.

And that’s the basic idea of JWT Authentication a safe and efficient way to verify users and control access in modern web apps

@sifendev

Through: bot

Whenever you are doubting whether to use pointers or values in Go, default to values. Using values is simpler, and it is easier to switch to pointers in the future. It is also easier to reason about. However, there are some clear signs to indicate when to use one or the other.

Use values when
1. You are working with primitive data types and small structs that do not require modification
2. If you want to prevent data manipulation during compile time

Use pointers when
1. You are working with big structs
2. Your struct requires modification
3. If one of your methods accepts a pointer, use pointers for all others for consistency.

My Upwork Stats for 2025

I am switching my main backend language from Python to Go. Since it has been a while since I touched Go, I decided to refresh my memory of its syntax by building a wc clone from John Crickett's https://codingchallenges.fyi/.

I built the line, word, byte, and character count functionalities and wrote unittests using testify. I have also included the architecture documentation and an easy-to-follow Readme.

Here is my repo: https://github.com/abeni-al7/aben-wc

I would appreciate any feedback on my code and suggestions for improvement.



LinkedIn
X

I just solved today's Advent of Code challenge. The second part took me a lot of time. Been staring and writing out the pattern on excalidraw for so long. My solutions are in my repo:

https://github.com/abeni-al7/advent_of_code