Karl Brodowsky's IT-Blog – IT Sky Consulting GmbH

UQL

SQL has been the „standard“ for databases for decades now and it has been improved a lot. But it has its limitations, because it enforces the relational data model.

But now Oracle has decided to replace SQL by UQL, „universal query language“. We live in modern times and of course the standard for UQL is defined, but the engine that parses and executes UQL will not be implemented. Instead queries are passed to the cloud and Oracle’s AI engine understands the UQL queries and returns the internal low level commands to access the data accordingly to the local database engine.

The new database product „Oracle U1“ works with UQL. It supports both transactions and cisactions. Data is structured in an object model. Objects belong to one or more classes, inheritance and aggregation is fully supported and it is also possible to store standalone functions and methods for classes, which are written in PL/UQL.

With Oracle U1 a new pricing model will be introduced. Licenses for Oracle U1 will be free (free like „free beer“, not like „free software“) for every kind of usage and Oracle will generate money only through the fees of AI usage for parsing the UQL queries.

For users of functional and object oriented programming languages there will be ways to integrate UQL and PL/UQL and since Oracle U1 can directly express the object hierarchy of the application domain model, it will be much easier to integrate and avoid the disruptions caused by tools like hibernate that exist due to the discrepancies of the relational and the object model.

When upgrading the software, there will be no need to migrate the database. AI will figure out how to translate stored objects that matched the domain model of a previous version of the software to the current domain model.

Also there will be UQL integration into ReactJS and Angular, so modern applications can avoid writing a „backend“ and just access the database directly from the Angular or React application running in the browser. Extra performance can be gained in such a setup by putting the database into the Oracle Cloud. For delicate data models it will be offered to store the data in data centers in a specific country and to encrypt all data in the cloud database. This will revolutionize security architectures.

Christmas 2025

З Рiздвом Христовим − Merry Christmas − Frohe Weihnachten

Available for new Projects 2026

I am available for new projects starting 2026.

I can support you for SW-architecture, SW-development, Linux, databases and security.

See IT Sky Consulting to see for more details on what I can offer.

Available for new Projects

I am available for new projects starting in 2026.

I can support you for SW-architecture, SW-development, Linux, databases and security.

See IT Sky Consulting to see for more details on what I can offer.

Devoxx 2025

In 2025 I visited the conference Devoxx in Antwerp from 2025-10-06 to 2025-10-10. As always, the first two days were „deep dive“, with longer talks and fewer participants, followed by two intense days with many talks, mostly about 50 min long, followed by the Friday with only 3 talks and much fewer participants. A lot of people see to skip the last day. I used to think it was because of Belgian beer, because there were some evening events with beer on Thursday, but I think it is just the desire to go home for the weekend. I recommend staying in Belgium for the weekend after the conference, at least once or twice.

Once again AI was an important topic during the whole conference. The euphoria of the previous years was still present and the organizer Stephan had done a lot of great development for the Devoxx-Web-Application using AI. Also the intro movie before the talks was great and this would not have been possible without AI.

Here are some photos of my impressions of the Devoxx:

Some Conference Talks

Here I have aggregated some conference talks that I have held in the past:

Python

Python is actually a few years older than Java, but in the last couple of years it has been tremendously successful and it is considered „modern“ today.

Python comes in two major versions Python 2 and Python 3. It took more than a decade for Python 3 to surpass Python 2, but it looks like that has finally happend a few years ago. The Python community was fortunately strong enough to support both versions at a high level for such a long time. But, please, use Python 3, if you can, preferrably a recent version.

Actually I recommend learning Python, because it is good to know it in this time. I would not call it a favorite language. To my impression Ruby has similar capabilities, but is a much more beautiful language. Perl is still a bit easier for „small tasks“, especially one liners using regular expressions. Raku tried to be a better replacement for Perl and has some great concepts, but didn’t really pick up momentum. Clojure is a phantastic language, but it forces two challenges on the ones who try to learn it. You need to embrace the Lisp-Syntax and the functional paradigm at once.

Then again there are statically typed languages, like Rust, Java, Kotlin, Scala, C#, F# or Swift, that tend to eliminate some errors at compile time and and that tend run faster, especially in conjunction with full native compilation like using llvm or Graal VM.

I do not like the syntax with „significant whitespace“, but since I use indention to emphasize structure anyway, I would not call that a show stopper.

So what is cool about Python?

I think that it is relatively easy to learn for someone who already knows several programming languages. It is a multi paradigm language which support object orientented programming, procedural programming and functional programming and it follows pretty much common sense of what can be find in other programming languages.

The most important advantage comes from the fact that Python is heavily used in certain areas and that there are a lot of libraries available, especially in these areas. It has actually bein among the most popular languages on Tiobe Index and it is actually Number 1 today (2025-04-10) and has been „number 1“ or at least „almost number 1“ since 2021.

I see the following areas:

Scripting language for system engineering and related tasks (competing and replacing Perl and Ruby and to a minor extend Bash)
Artificial Intelligence
Data Science
Scientific Computing
Usage as embedded language
Google is known to heavily rely on Python (among other languages)
Many other organizations list Python among the „recommended programming languages“, but do not recommend using Perl or Ruby
and of course a lot of programming as a general purpose language

In the end of the day the results that can be achieved count and a lot of cool software has been done in Python.
I will use Python in the future and explore where it can be useful.

New Input Device

We have seen keyboards, mouses, touchscreens, touchpads, voice recognition and even foot pedals as inpunt devices.

But apparently this is all outdated technology. The newest generation of computers will come with some kind of rubber balloon which is squeezed in different ways. It has been proven that this can replace both keyboard and mouse and that after some practice it is much faster than a keyboard. The nice thing is that it also generates its own energy and does not need to be charged or connected with a cable.

The newest Gartner reports suggest that IT departments will save billions by applying this technology and bypassing the bottleneck keyboard+mouse.