Smart thinking on that reply.
The day could have turned out a lot worse if you would have immediately replied "Ah, the one with...."
Posts by Erik van Roon
Good point by Connor to add tiebreakers to the order by clause.
I will add: do NOT use default RANGE BETWEEN but please explicitly use ROWS BETWEEN !
www.kibeha.dk/2013/02/rows...
Get Oracle Partition High Values as VARCHAR2 types instead of LONG in Oracle 19c and earlier (new blog).
There's also some data 23/26ai data dictionary internals!
tanelpoder.com/posts/oracle...
Colours are so overated
When you feel insulted, but don't have very strong counter arguments 😢
Hmm. Disappointing.
I thought it might be to activate the cesspool, implying it can be deactivated as well.
I've had customers where that would've come in handy 😁
Technically... it's a solvent not a solution.
Alcoholic beverages however.....
You really need a personal assistant.
No, really!
Remember how they told us they needed the second amendment to be able to protect themselves against a tyrannical government?
Yeah, turns out it wasn't that.
Assertions for data integrity on a less-than-ideal data model.
www.kibeha.dk/2026/03/asse...
Or "How I could have used assertions 25 years ago when I worked with Concorde XAL"...
#OrclDB #SQL #Assertions #OracleACE
😮
7? No way. Everything added after 5 is just hype.
Ah. Check! Missed that. Thanks.
Well, actually...
After reading the blog post (again) I decided to watch the video too.
Glad I did. The underscore parameter isn't mentioned in the blog post.
(I fear that parameter is going to be set in many places 😢)
I'm pretty happy with this, as I have run into people who did not solve the problem by using proper datatypes but by a rock solid rule: "don't ever use searched case".
That won't work anymore.
OTOH, they might just change the rule to "26 broke case expressions, don't ever use case" 😢
I get the feeling there are people in this world you like even less than me 😮
Aha. So two identical records in the resultset.
Of course.
Thanks.
Looks good.
But why "rownum=1"? I would expect only 1 of the queries to return only 1 row. Even if it's 'just to be sure', it implies that more than 1 row *could* be returned, in which case the query needs an order by somewhere to ensure a consistent resultset.
Or am I missing something?
If you exclude the redacted parts from your calculations, you're left with a little over a workday, though.
📣 26ai is now available for download, x86-64 bit Linux distributions.
www.oracle.com/database/tec...
www.youtube.com/watch?v=dNvI...
Finally...
It's here at last!
Oracle AI Database 26ai (23.26.1) Enterprise Edition for Linux x86-64 on premises is GA
Once again, there's a host of #SQL updates in this RU. Here are a few to get you started:
You have a weird definition of the word "friend"
I can hardly wait to get my hands on this.....
"It doesn't matter if AI spoke to an Agent which spoke to your Uncle Mike whose used his iPad to inscribe COBOL onto a parchment that was attached to a carrier pigeon...It is still your code"
connor-mcdonald.com/2026/01/04/s...
Also, without exception, while creating that MVP the definition of "Minimal" shifts into something completely different than I was ever taught.
Best Christmas 2025 present ever! 🙌
The new #SQLcl release brings a tiny but mighty new feature! Goodbye pesky little line numbers! 🧹
A shoutout to the SQLcl Dev team to get this in! 👍
You're trying to tell me, when I retire I'm still expected to waste my time on other people?
In that case I'd better continue to keep working. At least I'll get paid for that nonsense.