Please Don’t Use HTMX

Radi-dev — Mon, 05 Jan 2026 14:43:50 +0000

This is not a post about hating HTMX.

HTMX is actually very good.

This is a post about why I don’t use it yet, even though people keep telling me to.

The title is a little tricky on purpose.

A better title would be:

Please Stop Telling Me to Use HTMX, or
Why I’m Not Using HTMX (Even Though It’s Cool)

What People Say About HTMX

When people talk about HTMX, they usually say things like:

“You don’t need a frontend framework anymore”
“Just write HTML on the server”
“Less JavaScript is better”
“It’s simpler and cleaner”

And honestly, that sounds great.

If you are tired of complicated tools, HTMX feels like a breath of fresh air.

So I tried it.

The First Problem: Seeing Changes Is Slow

When I use modern frontend tools, this is what happens:

I save my file → the screen updates right away

That feels nice.

With Go + HTMX, it usually feels like this:

I save my file
→ I stop the server
→ I start the server again
→ I refresh the browser
→ now I see the change

That’s a lot of steps for one tiny change.

Yes, there are tools that can help:

tools that restart the server for you
tools that refresh the browser for you

But you still have to set all that up yourself.

Modern frontend tools already do this for you.

The Tailwind Problem (Why My Styles Don’t Show Up)

While styling with Tailwind, I prefer to build only the classes I actually use in the specific project to avoid packaging the full CSS bundle.

But Tailwind only knows about styles that already exist at the time it runs. Meaning that after the initial build, further changes down the line may be prone to containing undefined CSS classes.

With HTMX:

HTML is split into many small pieces
some parts only appear sometimes
new styles are added later

So this happens a lot:

“Why doesn’t my new style work?”

The answer is usually:

Tailwind didn’t see that style
you forgot to rebuild the CSS
or the style only appears sometimes

You can fix this — but you must:

run Tailwind in watch mode
configure it carefully
remember special rules

That’s more thinking than people admit.

HTML Starts Simple… Then Gets Complicated

HTMX makes simple things very nice.

A small button looks clean and easy.

But real apps are not small.

Real apps have:

loading states
error messages
permissions
conditions
reused pieces

So your HTML slowly becomes:

longer
harder to read
full of logic

You didn’t remove complexity.

You just moved it into HTML.

That’s not bad — it’s just something to know.

“The Old Web” Wasn’t Always Better

People say HTMX feels like the “old web”.

But the old web:

refreshed the whole page a lot
felt slow
still used JavaScript when things got hard

HTMX isn’t the old web.

It’s something new that mixes old ideas with new ones.

And mixing things always comes with tradeoffs.

People Recommend HTMX Too Quickly

One thing that bothers me is this:

You say:

“I’m struggling with this problem”

Someone replies:

“Just use HTMX”

No questions.

No context.

No thinking.

HTMX is not magic.
It’s a tool.

And tools are good only when they fit the job.

Why I Don’t Use HTMX (Yet)

I don’t avoid HTMX because it’s bad.

I avoid it for now because:

I like seeing changes instantly
restarting things breaks my flow
HTML gets messy faster than expected
styling takes more care than advertised

That doesn’t mean HTMX is wrong.

It means it’s not always the best choice.

When HTMX Is Actually Great

HTMX is amazing for:

forms
dashboards
admin pages
internal tools
simple business apps

If that’s what you’re building, you might love it.

If you’re building:

very interactive pages
fancy animations
fast-changing designs

You might not.

And that’s okay.

The Real Point

This post is not saying:

“Don’t use HTMX.”

It’s saying:

“Please stop acting like HTMX is the answer to everything.”

I’ll probably use again HTMX one day.

Just not because someone told me I should.

Building Open Source for 100 days

Radi-dev — Thu, 27 Jul 2023 06:02:53 +0000

I'm overwhelmed, I just made an announcement on Twitter about my intention to embark on the 100 days of code challenge starting 1st August. I intend to build fully functional open source projects during the course of the challenge .

Why?
Why not?

I found myself feeling demotivated of recent due to high levels of unproductivity, then I stumbled upon a post talking about the 100 days of code challenge. Honestly, this is not the first time I'm learning about the challenge, but it hit different this time around because I thought "why not?"; My social media presence is very low. In fact, my total online presence is low; My Github activity is below inspiring. This, I thought, is the best time to start. See how far I go.

https://twitter.com/radi_dev/status/1684422490678759429?s=61&t=XKS8vD5JNdx3m6lFOsrGIQ

DEV Community: Radi-dev