Hot Tech on Tech Pron

[Renegade]

You guys are going to love this...

MS is... well... I don't want to be totally rude here, so have a read at my blog -- expect rudeness.

IP Porn – Hot Tech-On-Tech Action!

It's just obscene!

MS on Google action. Totally NSFW~!

[Stoic Joker]

Good grief generating a report from data pulled out of two different sources is a big creative whoopee why exactly?? I did that for the office back when I started there (like 10 years ago), just because I (was too lazy) didn't think the 8 hour manual excel data massaging method they were using was a good use of my time.

...Can I sue both of them?

[Renegade]

I'm a data purist.

I think in terms of the 6th normal form, and calm myself down to the 5th normal form, then work in the 4th normal form as I try to tame myself down to work in the real world for the 3rd normal form, and if absolutely needed, as shown by empirical evidence in speed tests, etc., denormalize to the 2nd normal form if ABSOLUTELY necessary.

I'm not capable of talking on this topic without flying off into a profanity and obscenity laced rant.

It's just so completely... well... I think my position is clear enough by now.

[Renegade]

Good grief generating a report from data pulled out of two different sources is a big creative whoopee why exactly?? I did that for the office back when I started there (like 10 years ago), just because I (was too lazy) didn't think the 8 hour manual excel data massaging method they were using was a good use of my time.

...Can I sue both of them?

I wish someone would... This is just completely idiotic. Not as stupid as the round corners suit with Apple & Samsung, but damn near that stupid as to make no significant difference.

[rgdot]

So mashups themselves are patented too? 99% of web 2.0 is a defendant I guess

[Renegade]

So mashups themselves are patented too? 99% of web 2.0 is a defendant I guess

Yep.

The really f'd-up thing is that EVERYONE in the world who studied algebra or non-linear mathematics for the past, oh, few hundred years or so?, already knows all this stuff.

It might a bit newer than "rounded corners", but it is still in the same general ball park.

[Renegade]

A^B
C^D

Now... (A^B)^(C^D)

( ^ used as the union operator, usually an upside down U)

It's that simple... No more than that.

[40hz]

<moved>

Thx SJ!

[Stoic Joker]

Er... I think you're a thread off...but interesting post none the less.

[Renegade]

^^ I think 40hz was trying to force this into the basement, where we could beat the shit out of each other over 5th/3rd normal form.

[40hz]

^Nope! 

I draw the line (usually) at 3NF. Never once saw an advantage to taking it beyond that in any database I ever designed due to how the engines I used (FoxBase/Paradox/R:BASE) actually worked rather than how Jedi Master Ted Codd said they should.

(P.S. 40hz does not go down into the basement anymore. So that's a good place to hide from him if you ever feel the need! )

[Renegade]

^^ I won't disagree in principle...

However... I don't think a DB should be designed in 3NF. Starting in 5NF is the best approach. After that, you can denormalize down to 3NF, and 2NF if absolutely necessary. But the 5NF principle is the best starting point. From there, you clearly understand everything, and the 3NF is just a subclass of 5NF, and easier to deal with. But not understanding the 5NF for a DB is just, well, I don't like it.

[40hz]

^^ I won't disagree in principle...

However... I don't think a DB should be designed in 3NF. Starting in 5NF is the best approach. After that, you can denormalize down to 3NF, and 2NF if absolutely necessary. But the 5NF principle is the best starting point. From there, you clearly understand everything, and the 3NF is just a subclass of 5NF, and easier to deal with. But not understanding the 5NF for a DB is just, well, I don't like it.

Understood. And that's why you're a coder and I'm in systems.

[Renegade]

^^ I won't disagree in principle...

However... I don't think a DB should be designed in 3NF. Starting in 5NF is the best approach. After that, you can denormalize down to 3NF, and 2NF if absolutely necessary. But the 5NF principle is the best starting point. From there, you clearly understand everything, and the 3NF is just a subclass of 5NF, and easier to deal with. But not understanding the 5NF for a DB is just, well, I don't like it.

Understood. And that's why you're a coder and I'm in systems.

A while back I was doing optimization for a MySQL database. It was quite fun! You really need to dig down pretty deep to find out WHY/HOW things work the way they do, and then work your design based on that. Sticking to XNF theory just isn't useful as everything comes down to implementation.

I still think I enjoy the purity of abstracting data into a normal form more than doing actual optimization though. They're not the same thing by any stretch. Apples and green tea ice cream I suppose.