DEV Community: Sukma Wardana

Remove Gibberish Messages in Git

Sukma Wardana — Tue, 13 Feb 2024 12:51:36 +0000

I am a perfectionist.

I want to do everything perfectly, including writing commit messages in Git. Whenever I need to work on new features or fix bugs, I will create a new Git branch.

But, I can't avoid writing crapy and meaningless commit messages, especially when change a piece of code many times just because need to trigger the CI/CD pipelines.

Here is an example. I was working on a new feature of my project. I will create a new branch called feat/shinning. After working on the code, here are the commit logs when I type git log:

commit c62397c29f5a299c814d7a138092fc345448dc05 (HEAD -> installer, origin/installer)
Author: swardana 
Date:   Wed Oct 11 21:22:14 2023 +0700

        change styles

commit 4cbf76038d2d6685dacf0f253341d057ce348f50
Author: swardana 
Date:   Wed Oct 11 21:11:04 2023 +0700

        fix typos

commit 8a5dcd63364263dfda0f2ec38a609c3de53c3c08
Author: swardana 
Date:   Thu Oct 5 21:04:05 2023 +0700

        fix once again

commit 3abc24267d4130a9e9b58c051f406249054a906f (origin/main, main)
Author: swardana 
Date:   Tue Oct 3 22:12:13 2023 +0700

        wrong class

Well, now you know how messy and meaningless my commit messages are. I could still merge the feat/shinning into the main, but I hate to see that on my logs history.

My goal is to combine all the commits into only one meaningful commit message.

There are multiple ways to achieve it. But, what works for me is this.

Step 1: Working with temporary branch

Rather than create a proper branch, I will create a temporary branch and I could put many crapy or silly commit messages there.

$ git checkout -b tmp/shinning

Step 2: Create proper branch

When done with the changes and ready to merge it, I will create a new branch from the commit in the main branch where the temporary branch is initiated.

$ git checkout main
$ git checkout -b feat/shinning ecddf76

Step 3: Merge the temporary branch to the real branch with --squash option

$ git merge --squash tmp/shinning
$ git commit # without -m

When enter the git commit, without -m, an editor should be popup with all the commit logs, and files changed from the tmp/shinning. I will delete everything and write proper commit message.

Step 4: Merge the real branch to main

This is the last and final step, merge the real branch to the main.

$ git checkout main
$ git merge feat/shinning

Finally, it's done. With this, I could avoid any crapy and messy commit messages in my logs history.

This is inspired from the Stack Overflow answer which refer to this Github Wiki.

Absolute Number in C Language

Sukma Wardana — Tue, 30 Jan 2024 09:27:30 +0000

An absolute value in mathematics is defined as the non-negative value of x without regard to its sign. The notation of absolute value is a vertical bar on each side of the variable |x|. So, by theory |-16| will become 16.

Now, how could I bring this into the C programming language?

The easiest way is to use the abs API from the stdlib.

#include <stdio.h>
#include <stdlib.h>

int
main()
{
    printf("%d\n", abs(-16));
    return 0;
}

However, what if there is a situation in which you can't use stdlib or any libraries?

I experienced this before and was helpless. Thus, I decided to employ my Google-fu to learn how to implement absolute value without any libraries.

Then, I found the simple and easy-to-implement solutions.

#include <stdio.h>

int absolute(int x);

int
main()
{
    printf("%d\n", absolute(-20));
    return 0;
}

int
absolute(int x)
{
    return x * ((x>0) - (x<0));
}

The source explained it better and was easy to understand. But what made me surprise is, how it leverages the characteristics of C programming language. Which is, in C we don't have a boolean data type.

In C language, to determine true or false it's using an integer value. Zero represents false and One represents true. So, let's examine how the parentheses value outcome will be:

(x>0): True (1) if x greater than 0, False (0) otherwise.
(x<0): True (1) if x less than 0, False (0) otherwise.

Let's break down when the x is a positive number.

x == 10
result = x  * ((x>0) - (x<0))
result = 10 * (  1   -   0  )
result = 10 * 1
result = 10

Let's break down when the x is a negative number.

x == -16
result =  x  * ((x>0) - (x<0))
result = -16 * (  0   -   1  )
result = -16 * -1
result =  16

And lastly, if the x is a zero.

x == 0
result =  x * ((x>0) - (x<0))
result =  0 * (  0   -   0  )
result =  0 * 0
result =  0

The absolute method will never alter the x when it's a positive or zero number. But, when we set the x as a negative number we negate it by multiplying with -1 to get a positive number from x.

What an interesting solution.

Introduction

Sukma Wardana — Wed, 24 Jan 2024 13:10:05 +0000

Hey!

This is the first. Therefore, it's not about anything in particular–just an introduction and my way to greet my future cyber friends.

This will be primarily about software development ideas. I'm passionate about open source software and pragmatic thinking. I will write solely about my ideas and views on it.

Anyway, welcome. Together, let's see how this works out! :)