Technological Dinosaurs or Social Dinosaurs?

You're going to have to bear with me on this one because my thoughts aren't fully formed.

As I'm sure I've mentioned, I'm a little elderly (and you thought I was going to talk about radios). As such, I've seen the death of a few technologies. I may not have shared this widely, but my uncle ran a newspaper for a time in a small town in Missouri. They were still using hot type (literally a bath of lead turned into letters). I got to help type-set a newspaper using this when I was a kid. My uncle then went into the printing business, cold type. My introduction to computers was partly doing publishing & printing, but no hot or cold type, just PageMaker through desktop publishing. I talked to the people doing hot type who moaned about cold type and how it wasn't as good. Then I heard the cold type people moaning about desktop publishing and how it wasn't as good. I watched these technologies shrink and the jobs, mostly, die. Heck, who does desktop publishing now?

These were, to me, technological dinosaurs. Their time had come and gone. So, I've spent a career, trying hard to watch to see if I'm becoming a dinosaur. I've had a career in databases, most of it focused purely on relational data engines, and most of that SQL Server. However, lately, I've been working in PostgreSQL as well. Not because SQL Server is dying, far from it. Just because, it was pretty clear, a lot of work was happening in that space, and, I have a fear of becoming a technological dinosaur.

Now in the last few years, we've seen machine learning become, well, boring. It's pretty thoroughly embedded and becoming more so. In addition, we've seen the large language models coming along (LLMs, aka AI). There's been quite a bit of FUD (fear, uncertainty & doubt) about AI, but it's looking more and more like it's just going to be one more tool in the toolbox (although, I still have my doubts about how AI affects peoples learning & career paths, different discussion).

No, what's really changing the landscape, and has me concerned is something else entirely. Fabric. And DataBricks. And SnowFlake. Redshift. Data Lakes. All that stuff. And no, not because I see that replacing relational data stores. Nope. Not at all. Instead, what we're seeing is that no one seems to care about the relational data stores. Oh, yeah, we're still building them. OLTP (OnLine Transaction Processing for you kids out there) is mighty and strong and ongoing. Yet... No one cares about it. There even seems to be a very large indifference to it.

Why? What happened? Was there a major technology shift that we just didn't see?

Funny enough, no. Technology is pretty much transparent. OLTP is still one of the major collectors of data. That then feeds into all these magic analytical engines and the AI. So why doesn't anyone care about OLTP any more? Well, while I, and I suspect a lot of you, have been focused on watching technology to understand if I'm working on hot type, cold type, what have you, in order to protect my career, a social shift has been occurring.

See, more and more, people are just assuming that OLTP will just work. We've got Object Relational Mapping (ORM) tools to handle the code. Structures just work. If you're on the cloud, backups are handled. High availability is handled. Everything is just kind of taken care of. So, yeah, we're still using OLTP, relational databases, all the stuff I've spent 30 years learning about and supporting. It's all there. But, socially, we've stopped worrying about it. In theory (and I think this is where things get weird), it's a solved problem. All you have to do is go to a higher service tier  and all your problems are solved. Our challenges in this brave new world are all in the analytical space. Collecting data is easy. Using data is hard. This is why you see more, and more, and yet still more, Fabric/DataBricks/SnowFlake/RedShift being talked about and less and less traditional data management, monitoring, query tuning, and all the rest.

So, we've hit the technological dinosaur stage then?

No.

This is my problem. People are paying more than they should in order to make their poorly built, badly coded, horribly maintained, relational databases work. It looks like something that's not a problem, yet, it's still there. What's happened is we've socially moved on. Our attitudes have changed. The tech is in place, in operation, stinking on toast, but no one seems to worry about it any more.

And now, I'm back to wondering, am I a dinosaur, or am I the little boy pointing out that the emperor seems to be showing quite a few of his naughty bits?

Sorry for the rambling nature of this one. As I say, I'm still cogitating on this one. What do you think?

Grant Fritchey

Join the debate, and respond to the editorial on the forums

Learn how Redgate’s Foundry drives AI innovation in database management - from intelligent monitoring and ML-based automation, to smarter SQL optimization.

With our latest release, we are introducing callback scripts for state-based deployments, which lets you execute custom logic before or after your state-based deployments.

Kids have always played with and talked to stuffed animals. But now their toys can talk back, thanks to a wave of companies that are fitting children’s playthings with...

Teaching developers to work effectively with AI means building habits that keep critical thinking active while leveraging AI’s speed. But teaching these habits isn’t straightforward. Instructors and team leads...

Central bank says market concentration hasn't been this extreme in 50 years.

This article dives into a fun (and interesting!) strategy for widening fixed-width columns in SQL Server, to reduce downtime, risk, and runtime at the time when a column’s data type needs to be changed.

Synchronous replicas in SQL Server Availability Groups promise no data loss, but they don’t promise zero delay; under heavy load they can still fall behind. This article shows how to measure and track that hidden replication delay using SQL Server performance counters, so you can see how well your system keeps up during IO‑intensive operations and plan maintenance more safely.

 Why Developers Shouldn’t Have sysadmin access...

There are moments in technology when the ground sh...

Introduction These are my SQL Server Diagnostic In...

THE VIDEO THE SYNOPSIS In this video, I take you through some of the limitations of window functions in SQL Server, including the lack of intervals and oddities around...

Question 1: Pick the Problematic Plan: someone han...

 DBAs should never run SSMS under their everyday Windows account If you open SSMS under the same Windows identity you use for email, browsing, and chat, you bring all that... The...

Parts 1, 2 and 3 got you to the (SQL) engine room. Now we use community-trusted tools to find what’s going on, fix it safely, and hopefully keep it... The...

 Truncate Table Pitfalls Truncating a table can be gloriously fast—and spectacularly dangerous when used carelessly. If you want the speed without the face-palm moments, here’s a practical, interview-ready guide to... The...

In my previous post, I showed how I collect index information across all replicas to get the whole picture. In this follow-up, I add more context.

If there’s one thing every Scooby-Doo episode teaches us, it’s that solving mysteries is a team skill. The gang doesn’t stumble onto answers by luck (well… except when Scooby...

How the value filter behavior works in SUMMARIZECO...

…and right now, the same goes for AI in your dat...

While every managed database service has high points and low points, there are three things that make RDS for SQL Server shine in comparison to Azure SQL Managed Instance:...

“Oh, hi, don’t mind us – we’re just having...

Error handling is an important part of scripting a...

Patrick Gruenauer decides to continue: Error handl...

Authors: Shahrzad Azizzadeh, Kaustubh Khandwe, Bahar Biller, and Paul Venditti On large-scale solar farms, power loss is the silent drain on profits. Unoptimized panels chip away at efficiency, causing...

Decluttering a visual is an important first step�...

I watched Bob Ward demo vector search in SQL Server at the NYC PASS road show in late August, and I have a confession that I’m not proud of....

Financial institutions operate in a highly regulat...

For decades, enterprises have approached data mana...

Recently, we worked on a production PostgreSQL dat...

Whether you’re a seasoned DBA or just exploring database tools, DBeaver offers a powerful, cross-platform GUI for interacting with PostgreSQL and many other databases. As a continuation of the... The...

The Sinking of the RMS Titanic in 1912 has shocked and entertained for over a century.  I will admit to being shocked to find a GitHub Repo with a...

As we discussed, Lock Manager manages heavyweight ...

Let's talk about relation-level locks and various ...

Introduction Every year, we get a new PostgreSQL major release, each bringing performance improvements, new features, and extended compatibility. While minor version upgrades are straightforward and...

The long-running debate around how Power BI calcul...

If there is one thing everyone needs, but doesn’...

I was reading LinkedIn posts the other day when I saw this blog about what was apparently an interview question about some forms of a COUNT aggregate function This...

As more sites require IDs for user age verificatio...

Texas requires sensitive info just to check weathe...

Scattered LAPSUS$ Hunters gave Salesforce until Friday to pay or else.

John Clarke, Michel H. Devoret, and John Martinis built an electrical circuit-based oscillator on a microchip.

Yes, it's another potential qubit. But it's also some very cool physics.

With the new investments, Stoke has raised nearly $1 billion.

Workers with ropes could make the moai "walk" in z...

In the ancient times of the late 90s, Bert worked ...

Four motors, 800 V, more than 1,000 horsepower, and it's all made in-house.

The honeymoon is over, and macOS 26 Tahoe broke th...