So I started exploring a bit deeper just to understand the idea.

First I tried to understand how CDN works. I learned how content is cached in different locations and how systems like AWS CloudFront serve users from the nearest location. That helped me understand why modern apps try to move logic closer to the user.

Then I started reading about serverless architecture. The idea itself is quite interesting. In a normal backend server, something is always running. But in serverless the function runs only when a request comes.

If nobody is using it, nothing is running. When a request comes, a small environment is created and the function runs. That is why sometimes the first request can take a little time because the environment needs to start. The concept of scale to zero was something I found really interesting.

After that I explored a bit about edge computing. I read about how Cloudflare Workers run code near the user using the V8 engine. Instead of going to a single central server, the logic can run at edge locations.

While reading all this I also came across how platforms try to optimize these systems. For example how instances are reused, how environments stay lightweight, and how cold start delays are reduced.

I also learned about how Vercel tries to improve this with things like fluid compute where instances can be reused more efficiently instead of always creating a completely new environment.

Another interesting thing I noticed is how important serverless databases are becoming today. If your compute is serverless but the database is not designed for that architecture, it can create problems. That is why many modern databases are also trying to support serverless scaling and distributed access.

Earlier I used to think frameworks like Next.js are doing some magic. But after reading a bit about CDN, serverless compute, edge functions and infrastructure optimization, things started making more sense.

I am still learning and exploring these topics.

Because of this, I spent less time building basic UI components from scratch and could focus more on improving UX, responsiveness, and overall design quality using CSS. Combining shadcn with AI development worked really well— it reduced errors, helped maintain a stable UI, and made customization easier. Overall, it made frontend development more efficient and enjoyable...for me

honestly deployment teach me lot. still initially use auto deploy platforms but still configure diff domain server to connect each other pain..

i already know cors little bit so that not big issue. but browser security things wow too much. cookies sharing issue, same site config, third party cookie problem frontend cant read api cookie. spend many hours just understanding what happening.

also learn cookie security part like secure true use for https only before never think about this but production time understand why important.

learn how same site works, when use lax strict none, how subdomain help make same site better. also cors deeper understanding came when real production.

sometimes frustrating but after fixing one issue that feeling different. like small adrenaline rush after long debugging...

Docker solved one of my biggest pain points. I use Linux, especially openSUSE. A few months ago, I even made a post about PostgreSQL installation issues because many packages don’t properly support openSUSE. It was frustrating.
But with one Docker command, I can run PostgreSQL or almost any database without worrying about OS-level issues. That single thing made me love Docker even more.
For me, learning Docker didn’t feel very difficult because I am already familiar with virtual machines. Containers feel similar to VMs, but they are much more lightweight, resource-efficient, and practical for daily development.
Still learning, but honestly enjoying the process a lot.

While learning i understand how code flow works, how to write cleaner structure, and how to think while building a feature.
I also explored packages like React Hook Form for forms, learned better ways to handle state, how to keep code neat, and how bundlers like Vite make project faster. Even with Tailwind, I learned how to use it in a more proper and professional way.
This one month was not just learning React, but really understanding how it works and why it’s so powerful.....

What I understood is TypeScript sits on top of JavaScript and gives extra safety. Things like static checking and types help catch errors earlier,..something I missed in plain JS. While learning TS I also got a better idea of some JS flaws and why types matter.

I learned things like tuple, interface, union types, enums and generics. Some concepts were new, some just made JS patterns clearer. It already feels like writing cleaner code and thinking more about structure.

Next I will try to slowly use TypeScript in my existing JS projects and see how it improves code quality and my confidence. .....

From tutorials and videos I come to know that database design is also very important part in development. .....Before I was only thinking about coding part, but slowly I understand that how data store and how data flow inside system is also very important...........

I am not learning to become database engineer, but as developer I should know at least basic idea. That is where I saw about er diagram. I knew About NoSQL database ...from before so I tried to make er diagram in that way. recently I also start learning sql in near future i will definetly try on sql diagram also.....

I searched in internet, watched videos, practiced small small examples. Then I made two ER diagram by myself. One for banking system and one for learning management system. While making this I understand more clear how data move between different part of system. This give me some confidence and also improve my development thinking.

learning and exploring....

On the same day, I decided to try setting up SSH again, as I had forgotten how to do it from scratch. I followed the usual steps but encountered a "broken pipe" issue on my local machine. I tried everything, searched online, and even consulted AI tools like GPT, Claude, and Perplexity, but nothing worked. After an hour of frustration, I discovered that the issue wasn't with SSH or the remote machine—it was my external Warp terminal causing the problem. Once I switched back to my system's default terminal, everything worked perfectly without any errors. 😂

So I start exploring… MySQL, PostgreSQL, SQLite. Finally I choose PostgreSQL.

First days was setup headache I use Linux (openSUSE) so install was not smooth like Windows. pgAdmin setup, MySQL Workbench try, database connect… But when it worked, I feel like small win.

Then I found one gem VS Code extension “MySQL by Database Client”. Name is MySQL, but it connect both SQL & NoSQL databases. Do everything inside VS Code… no need extra apps. For me, best tool so far.

Learning now SELECT, WHERE, ORDER BY, GROUP BY. Also different functions, Aggregation, Joins (inner, left, right, full) and many more. Joins was little mind bending, but slowly it start making sense.

Also touching ORM like Prisma

Still learning…

First I added the PostgreSQL repo and managed to run the server using the terminal. Then I needed pgAdmin4 for GUI, but there’s no official package for openSUSE. After searching for a long time I found it on Flatpak and installed it.

When I tried to register my local server in pgAdmin, I got this error:

FATAL: Ident authentication failed for user "postgres"

For hours I couldn’t understand what this meant. Then I found that openSUSE by default uses peer or ident authentication. That means your Linux username must match the PostgreSQL username.

I was logged in as sourav, but trying to connect as postgres, so it kept failing.

The fix I found was to edit the pg_hba.conf file and change the authentication method to md5, restart PostgreSQL, and set a password for the postgres user. After that, everything worked smoothly.

This small struggle taught me so much about how Linux handles permissions and how database authentication really works.

Right now I’m learning things like how Node.js works, restApi,MVC patterns all about Controllers routes,express, middlewares,view,service ,env files,authorization, session, cookies, SQL, NoSQL database, aggregation, file upload with multer and cloudinary, JWT, bcrypt, npm packages all these slowly slowly.

It’s been around 1 month. Honestly, I’m not making my own projects yet but I do definitely, just following instructor videos and trying to understand and implement properly.

I first made the mistake of installing Linux on the same partition as Windows. That created a lot of problems later. Always install Linux in a different partition. Don’t touch the one where Windows is already installed.

BitLocker was another big issue. If the system breaks or something changes in the boot setup, Windows suddenly asks for the BitLocker recovery key. If it’s not saved, you’ll get stuck like I did. I had to log in to my Microsoft account and find it. It’s better to turn BitLocker off or at least keep the key saved before starting.

Fast Boot caused multiple issues. It keeps the NTFS drive in a half-on state, so Linux can’t write to it only read. I had to turn Fast Boot off from Windows power settings. In some cases, Fast Boot also messed with the GRUB menu it skipped Linux and directly booted into Windows. Disabling Fast Boot fixed that too.

Secure Boot didn’t cause problems in my case modern Linux systems handle it well.

One of the most frustrating things was trying to create a Windows bootable USB from a Linux system. There are flashing tools made for Linux I remember using one called WoeUSB but they didn’t work properly. The USB either didn’t boot or gave errors. It’s not that creating a bootable drive is hard, but for a new user, it’s really confusing. I ended up using Rufus, which works perfectly but the problem is, Rufus only runs on Windows. So unless you have a second device with Windows, you get stuck. It sounds simple, but for someone trying dual boot for the first time, this part wastes a lot of time.

During installation, I didn’t create partitions manually. I just cleared some free space from Windows, and Linux took care of the rest. Also, no need to create a second EFI both Windows and Linux can use the same one.

After installation, my GRUB boot menu didn’t show Windows properly. To fix that, I installed a tool called Grub Customizer in Linux. It helped me reorder and clean up the boot menu easily without touching any complex files.