
# Complexity comes at a cost, and for me the AWS juice is no longer worth the squeeze.

In 2010 or so, I was working at a small company that had it's important web apps on private VPS hosting.

At the time AWS was the new hotness, so I fired up an EC2 instance, started dabbling with RDS and before long I had a delicious cloud soup brewin'.

It's been a flexible and useful platform over the years, but ultimately I've found that for me personally it's way overkill. It's using a landmine instead of a mousetrap, and the destruction (cost) it leaves behind is no longer worth it.

Back in the day VPS providers were the challenge, or even worse being on shared hosting where you didn't have any control over the environment. Install an apache module? Ha, you have no power here!

And then I got hit with the JAM. JAMStack really changed the trajectory of my career, though I've come to realize that this term is loaded and was a bit of a marketing tool by Netlify. I really fell in love with Next.js, and the architecture of how it works has in my my mind replaced the need for things like WordPress and Drupal. It's not that these are bad tools per se, but the old model of database driven CMS sytems has been superceded by the javascript and markup driven static platforms.

They are easier to host, they are easier to work on, and the nature of the companies that provide ecosystems for them (Netlify, Vercel) are in the business of making your website FAST. Why spend time piping the output of some elaborate Terraform process into a CDN when you can just hook a repo up to Vercel and just go. CI/CD and feature branches out of the box.

Why was I maintaining the AWS infrastructure for so long?

A website I hosted for a friend. He and I lived together many moons ago and I helped him put together a small website for a book and story he created. He actually got an e-book published on Amazon for it, and I hacked up a little WordPress instance for hosting his content.

I spent probbably thousands of dollars in hosting, and countless time and energy spent avoiding this problem all because I felt bad I didn't have the time to save this relic of the past. I have backups of the filesystem and database but the real problem is I just don't want to deal with WordPress, or any of that ball of mud any time soon.

There are new tools that are less complicated than AWS is, because new patterns and toolsets have emerged. Modern web development complexity has bloomed to an insane level, and it's time for developers to take the wheel back from Jesus and steer towards more reasonable tooling that mere mortals can afford.

I'm turning off AWS because the type of businesses it attracts are not the kinds I want to work with. It's not about the platform, it's about the types of cultures that choose this platform and how they treat their employees, vendors and customers.

AWS isn't a bad platform (probbably the gold standard actually) and the people who work with it and on it are not inherantly bad people. Except for Jeff. Cultures that embrace AWS expect a wide range of understanding of an ever evolving list of products that I frankly cannot keep up with anymore. In my experience they have been less willing to "show you the way" and expect to throw a link of documentation at you and just be done with it.

I have emotional, technical, and philosophical reasons for leaving AWS behind. Mostly there are just better tools out there, simpler, more effective solutions. I don't care if these solutions themselves run on AWS, that's not my problem.

I want tools made for me, and I found some awesome ones.

Time to clean up some old messes.

Bye, AWS 👋

Mat Gilbert

 Complexity comes at a cost, and for me the AWS juice is no longer worth the squeeze

matgilbert.com

aws

Bye Bye AWS 👋

categories

links

social