@JamesLupolt As far as I recall, no there's no easy/direct way. I'll try it again later to see if I'm forgetting anything, I don't immediately recall the details of what I did when writing that. Maybe I just mapped it based on group cardinality.

Gosh that was over four years ago.

Thanks. Yes, that was all I could come up with. Look at group cardinality and process of elimination. Seems fine for relatively simple plans.

It does seem like a long time ago

First the 90s seemed like a long time ago. Then 2007. Now 2012. What's next?

@JamesLupolt In the end it will be as a Midsomer Murders character said, "Yesterday is another country."

[has started answering Azure Data Factory questions over on stackoverflow]

Only the basic ones mind.

How do people feel about ADF questions being over there rather than here?

MySQL [optimiser] 8 gets CTEs: mysqlserverteam.com/… Welcome to the 20th century MySQL

@wBob I would guess it's the old ask it there for a dev point of view or here for a data professional point of view. Both valid.

@Phil 8 ?

Is the next version going to be named 8? Or 5.8 ?

The optimizer is a separate download?

N/m, I guess it's a build of MySQL with the new optimizer

I don't suppose there are any Teradata experts around?

@ypercubeᵀᴹ so @Zane is our expert, because he mentioned the product more often than the others (1 vs 0) :D

@dezso Referring to it as their environment :)

But that'll do

@TomV I am so incredibly lucky I haven't mentioned [redacted] and [redacted], ever

@dezso more than 10 times actually. It must have bitten him

@CadeRoux works with Teradata, based on questions he has asked

A few heapers in the top teradata tag users dba.stackexchange.com/tags/teradata/topusers

@wBob CadeRoux used to be frequent in the Heap, some years ago

I'm currently in characterset hell. /cry

The review queue needs a little love

@Phil love given

not to the to-be-closed posts (except one)

@dezso TERRORDATA

@swasheck ha-ha-ha

frickin' hilarious

@ypercubeᵀᴹ thanks!

I love it when the Heap comes together : )

@PaulVargas Sure. Normalization is, basically speaking, about getting rid of non-simple (or multivalued, non-atomic) columns and ensuring that all non-key columns depend exclusively on the keys (primary and alternate [the latter commonly implemented via UNIQUE constraints]) of a given table and, also, making sure that such non-key columns never depend on each other.

@PaulVargas In this regard, I recommend A Relational Model for Large Shared Data Banks, by Dr. E. F. Codd, in which he introduces his theoretical framework that includes normalization.

@PaulVargas Regarding the terminology he used in the referred paper, it's helpful to read some of the concepts via SQL wording, e.g., keep in mind that a relation is a table, a domain is a column (along with its declared type) and an n-tuple is a row.

@PaulVargas As for the logical data modeling (or database design, or database architecture) part, I suggest studying Integrated Definition for Information Modeling (IDEF1X), which is a technique that was established as a standard by the United States National Institute of Standards and Technology (NIST).

@PaulVargas It is based on the early theoretical material authored by E. F. Codd; on the Entity-Relationship view of data, developed by Dr. P. P. Chen; and also on the Logical Database Design Technique, created by Robert G. Brown. I hope you find these concepts and resources of value.

jeez 1750, you really know your stuff

@Lamak I've been studying myself these materials lately, and continue finding new things that can help in practical situations.

ah, yes, I wasn't being sarcastic. You do really know your stuff

@Lamak I took it as a sincere comment, thank you. My point is that, from my own experience, it takes some time to get the most out of these tools (I'm still working on it).

@MDCCL yeah, you are doing fine

@PaulVargas The objective of relational normalization is to avoid data modification problems introduced by data duplication, which leads to information inconsistencies. For example, a database that stores multiple versions of a certain fact becomes an unreliable resource in the decision making process (which version is the one that reflects the truth?) and, from a technical point of view, such a system is harder to maintain (e.g., requires implementing "update synchronization").

grumpy me on that question

I wonder what the actual query and plan is if the tuning advisor suggest 25 indexes, and at the same time I dread seeing them

Not even Dynamics produces such beasts

@TomV the "what am I doing wrong?" ?

@ypercubeᵀᴹ He forgot about the "I inherited this" excuse