Cogito ergo caffeinatus

Maketh Not Thine Own Site

Why, oh why did I do this?

I love code. You write words and things happen. Its essentially magic.

Then there is learning how to make this website. This involved sacrifices to Eldritch Beings.

Lets do this story style.

Being able to understand something means being aware of its limits. GenAI is fun to read about, but since March, I have been constantly taking them through their paces.

Part of that, was building my personal website.

The last time I looked at website code intently was before CSS existed.

Overly ambitious isnt the half of it.

Mistake 2: Listen to an Expert

On May 31st, I reached out to a friend and asked him ” If you were to start out today, what would you choose to learn?.”

His suggestions:

With that sage advice, I imperiously instructed CGPT to build me a website, with all the right design cues. Oh and use Astro.

Obviously Chat GPT generated code failed.

I had zero idea why.

Well, it must be a google search away at least? Right?

Mistake 3: Break projects into manageable parts

There are, far, far too many things that go into getting you to see this text.

  1. There is the site itself.
    1. Design - Colours, Fonts, Images, (done well)
    2. Content
    3. Accessibility
    4. Responsive design (Make it look good on any screen)
  2. The tools to manage or add the content. (unless you write in bare HTML)
    • Just way too many things to list.
  3. The platforms needed to host it.
  4. Make sure all of these work together.

Just figuring out, what you need to figure out, is your first task.

You can’t just know about these tools though. You have to find out how they play with each other.

Mistake 4: I’m sure I can figure this out.

There is a CSS YouTube tutorial vide, that talks about “tutorial hell”. That is how many tutorials there are for Front End and web design work.

Mistake 1: Not getting someone else to do it.

I enjoyed this. entire process. I learnt a ton. I used so many fun things. It involved design, code, content, and figuring things out.

There are just faster ways to do this, with less technical debt.

As of 26th Jun. I have learned just enough to make spectacular mistakes.