Decaying Technology

A few weeks ago I received an email about a price update to my cPanel software.

Thank you for being a valued cPanel customer.

We want to inform you of an upcoming price adjustment that will take effect for all cPanel accounts and services starting  January 1, 2025. New pricing will also apply to any orders placed after December 16, 2024.

I don't really want to use cPanel anymore, but there is this fatigue about change compared to the effort of moving off a solution. Hell, I used to pay $170/year for a cPanel license and now I pay over $50/month and been doing so for years.

Now I want to finally get rid of this server (cPanel box), but its been a multi-year problem. To roughly summarize some of them:

• Moving friends/family off the server who aren't very technical to their own solution.
• Moving sites so old they only run on PHP 5.6.
• Moving sites when I don't control the DNS and unable to reach the respective party.
• Moving a site that is successful with a 40gb database with 0 downtime.

Now every year during price hikes I've just accepted it on the account of fatigue and 0 time. Now with cPanel upping my price even further because of an End of Life (EOL) operating system - my needle is being pushed further to resolving this.

At times I just want things to work and work for near forever and I know that's a selfish belief. I'm happy at least with Linode that continues to just migrate my Linode with 0 effort on my part between new virtualization and data-centers. This is a stark comparison to some of the big cloud providers who deprecate something as fast as they introduce it thus breaking something I depended on.

It reminds me when TOR broke all v2 URLs forever because they were insecure. For all the solutions/sites that hadn't migrated - they were lost forever as the amount of nodes running v2 compatible relays fell.

It reminds me of this one work exchange we had. We had a piece of purchased software we used and as updates came out - they announced they would no longer support version x of the underlying code it ran on. We knew about this, but the software still worked so it was a tough sell to the client until something actually broke.

Years went on and the client had an issue with the software. We triaged and determined we finally had a forced reason to upgrade to the new version of this piece of paid software. However, we took so long that this wasn't a simple little upgrade as tons of pieces under the hood needed to move with it.

So I always wonder whether in work or hobby - do you do the little amount of tech debt cleanup every so often or push it off until it becomes absolutely required in a huge mess of related items?

I'm torn because if you look at the Node ecosystem that stuff moves very quickly - I remember just a few years ago using "gulp", then it was "webpack" and now its "vite". If you look at the open source space of software its a never ending battle of monetizing free software (Redis, Elasticsearch, Terraform) which pits licensing vs forking into a constant battle.

You follow your favorite OS and one day they decided to end it entirely (CentOS) and tons of forks emerge. Where do you go? What do you do?

All I know is I have a greater respect for the "maintenance" cost of anything digital. This stuff never stops moving and if your software/project/thing doesn't - it'll just stop working one day.