End of May 2019​
​

A very common term in tech is tech debt. It is used in many contexts. But many times as a scarecrow. But what is it really? Tech debt is usually shortcuts that were taken at some point. Things that are not in optimal state. As I understand it, the expression is used when the engineers have some house cleaning to do.

The problem is, when you talk to leadership about tech debt, in their minds, it’s something good. It is a financial tool. It’s a loan. Loans are good, get cash now, pay back installments with an interest rate. Why should you pay it back? Why pay all of it back?

​
​

We took some shortcuts. We did have some non-critical components in a non-optimal state when we launched. Some bits needed more work when we launched fr-par and again later on. We took non-critical shortcuts on block storage as well. This was part of the reason behind the velocity.

Tech is an ouroboros. It eats itself. There will always be things to do. Things to fix. Things to tidy up. There will always be tech debt. Like in the in real life world there will always be debt. Some things cannot be achieved without debt.

Once you accept that, the question changes. How can I leverage the tech debt in the right way? My answer is look at the term sheet.

Some debt had low upkeep, low interest, low repayments. We didn’t need to fix it. Others had high interest, we were already working on it. Some, we knew in the black of our minds that they would require a lot of work, later on, when we reached a certain stage. We encouraged customers to not put more than 500 000 objects in a single bucket, but that debt was going to be recalled at some point.

Patching a little used software with a LD_PRELOAD? Not so much. Only having the command line interface to insert IP ranges, meh, didn’t need to be touched. Installments were a drop in the sea.

I didn’t consciously use those mechanisms at the time. I didn’t consciously calculate each loan grade or their terms. Whether interest rate, options or recallability. But I did make decisions using a similar mechanism intuitively. Prioritizing some things and letting others gather dust.

The finance analysis makes more sense to non-tech people. I now try to use it. Tech debt, like debt, is a tool. How you wield it and how much you leverage it matters.

Once you sort out your debt, what do you plan next?

1. ​https://openclipart.org/detail/184584/scarecrow-in-the-wind​
2. ​https://commons.wikimedia.org/wiki/File:Ouroboros-simple.svg​

If you have missed it, you can read the previous episode here​

To pair with :

• Here's to them - Rawtekk
• The plague (La peste) by Albert Camus