I’ve been writing (and blogging) for about ten years now. I’ve written lots of things over the years and I have learned how much better my writing becomes when I remove words upon editing. So much improvement can be found that I often feel it might have been best to have not written anything at all in the first place.

Over the weekend I came across a quote that gave me pause. It is attributed to Antoine de Saint-Exupéry, and has to do with writing as well:

“Perfection is achieved, not when there is nothing more to add, but when there is nothing left to take away.”

After reading that quote I started to think about data. I often see people trying to cram more data “just in case” in might be needed later. I will also find databases that are so badly abused that entire teams of people agree that “a re-write is needed”. They will discuss at great length what to do about sunsetting one system as they build another system with EVEN MORE crap than what the first one had. It can be surreal to experience, too. Some people are so attached to the old system you would think it was a family member or pet. They fight to keep someone from pulling the plug while others talk about what snacks they need to order for the wake.

Before you go about trying to shove more things into a perfectly good database do yourself a favor: stop and think. Are you shoving those extra columns onto that table so you can run a new report? Consider the benefits if you have your own reporting database, maybe even your own cube, before you consider just blindly shoving more and more stuff into the current system. Many times I see this as a result of “scope creep”, which is like the herpes of any good database design project…you know it’s there, you put some ointment on it, and the blemish goes away for a while but it never really goes away, the bad stuff is still with you, just lurking under the surface.

Think more about what you can take away from your current systems.

Think how much better things would be if you could separate out the transaction parts of your system from the reporting components.

Think about how much better things might be with less, and not with more.

“Il semble que la perfection soit atteinte non quand il n’y a plus rien à ajouter, mais quand il n’y a plus rien à retrancher.”

Yesterday I showed you some real world SQL joins and their associated diagrams. I also left as an exercise for the reader this question regarding a standard three table join:

“explain why I could need 127 different diagrams”

I had a few people guess the answer on Twitter and G+ as well as some emails. So I decided to post the answer here today in case you wanted to understand it a bit more.

Let’s consider the seven regions that we have in the following diagram: 

Now, consider the fact that no matter what type of join we want to execute (INNER, CROSS, LEFT, RIGHT, FULL OUTER) the end result is that we are either going to return data from one of those seven regions or not return data. In my diagram the green shaded regions represent areas for which we will return data and white is for regions where no data is returned.

In a nutshell: they are bits. On or off.

Most people still stumble at what needs to be done next, so I offer them a more familiar example. Consider you have a coin to flip, and also consider that after each flip we will either have a heads or a tails (yes, we will leave out the other possibilities such as landing on an edge, or disappearing into a wormhole, or having it swallowed by a marmot). If I flip that coin three times then this becomes the entire list of possible outcomes (H = result of flip is heads, T = results of flip is tails):

HHH
HHT
HTH
HTT
THH
THT
TTH
TTT

Eight possible outcomes, right? Most people see this example and say “yeah, I know that, but how does that help me with the above diagram?”

It is because each region in the diagram is just like a coin flip. Either we are getting data or not. Having seven regions is like having seven coin flips.

When I flip a coin three times I end up with 2^3 = 8 possible outcomes.

When I flip a coin seven times I end up with 2^7 = 128 possible outcomes. But one of those outcomes would be for when all regions are not selected, which is a fairly trivial example that we wouldn’t put on a diagram, so we are left with 127 total diagrams that we would need to draw.

That being said, many of the diagrams are repeats of each other. The join syntax for returning one table and the intersection of two other tables is the same; we just swap A with B with C and we can consolidate the need for a lot of the 127 diagrams.

But it was still far too many diagrams for me to think about wanting to draw and show the syntax for. Also: it isn’t practical, because the examples I gave yesterday are what we see in reality. Who wants to see textbook examples of joins anyway?

Just old math geeks like myself, I suppose.

I resisted the urge to take part in the meme for a while, but recently I had a handful of folks convince me to just put something together, so here you go.

It’s the night before Valentine’s Day and I know what you are thinking: “What would Tim Berners-Lee be doing right about now?”

I’ll tell you what he is doing. He is linking data. After all, that is what drove him to create HTML over 20 years ago. No, he didn’t invent the web. He wasn’t the first to dream up the idea of linking together documents and information between computers and systems.

But what he does do is to see the potential in things. And for the past few years he has been seeing the potential in connecting open data sources from around the web.

His ultimate goal is simple enough: to connect all data being created, everywhere, in real-time.

Piece of cake.

Big data is not new. It is not a passing fancy. Linked data  is, quite simply, the Egg McMuffin of future opportunities.

In Seattle this week, so naturally coffee is on the menu and I felt the need to share the following information:

• Caffeine is “technically” lethal, but you would have to drink an insane amount to have a 50% chance of death
• There is no reason to ever stop drinking coffee because it is pure awesome
• Caffeine helps improve your memory
• Caffeine can help fight cardiovascular disease, allow me to consume more bacon

Also, apparently the drink of choice used to be beer and gin, but then people started to consume more coffee and tea. So what will be the next drink of choice? I’m hoping it’s wine.

If so, you’re old. Sorry to be the one to tell you. 

Absolutely love this video, looking forward to seeing this technology used in my lifetime for flying cars or…better yet… hover-boards.

Here we are, at the end of 2011, and I suddenly realized I hadn’t done one of those obligatory “crap I did last year” posts like everyone else likes to do. So I decided to write about all the things that didn’t happen to me with the thought that in a year from now I can write “wasn’t consumed by the Mayan fires”. Here’s hoping, right?

Anyway, as we all look forward to 2012, here are my reflections on 2011. You’re welcome.

1. I didn’t die in a weather-related incident

My little corner of the world saw more than it’s share of extreme weather. In the span of ten months my children witnessed two blizzards that dumped over 18 inches of snow at once (causing one of our schools to have a partially collapsed roof and forcing students to finish the year in a neighboring town), a tornado (where we lost power for over two days as major transmissions lines were severed and if you’ve never seen a tornado travel down a river you should watch this), an earthquake (due to our founding fathers rolling over in their graves, apparently), a hurricane (where the eye came right over us), and a freak early snowstorm where we lost power for five days and it also canceled Halloween (no make up date has been announced yet).

That’s a LOT of extreme weather in a short amount of time. Fortunately for us we are all safe and warm right now, our house is relatively undamaged, and life continues as normal. I can’t say the same for the people that live within a few miles of here. And I won’t forget finding a piece of yellow tinted insulation in my yard this summer while the sky was colored orange-green and thinking “hey, ours is colored pink…I wonder where this came from?”

2. I was not awarded an Oracle MVP

I don’t even know if Oracle has this, the closest thing I know of is the Java Champions and while I do think I drink a good amount of coffee I can’t imagine how much it must take to become a Champion.

3. I was not present for the final launch of the Space Shuttle

No, I was not at the STS-135 launch. But I had a good reason why…because I had already seen STS-134, thanks to Stephanie Schierholz (@schierholz) and NASA. You may recall I was invited to the NASA Tweetup for STS-133 back in 2010. That launch was delayed by about 100 days, and I never made it to the actual launch. The week of STS-134 going up I was told about a tweet from Steph that those of us that missed STS-133 were to contact her immediately. Turns out they had some room for us in the twent (that’s short for ‘Tweetup Tent’) if I could get down to Kennedy Space Center in, oh…two days. I cashed in some miles, got my but down there…and the launch was scrubbed.

But just 17 days later I was back down there to watch Endeavour take her final trip into space. Also, I was able to have a most unique photo opportunity thanks to my friend Michael Steineke, who happened to be down there doing some actual photo work for NASA that day.

4. I did not quit my bacon addiction

Of course not, if anything my powers grew stronger these past 12 months. The number of bacon-themed items people give me is astounding. For Christmas this year I was given a string of bacon lights! I love all of the items I get…bacon peanut brittle…bacon t-shirts…bacon soap…bacon toothpicks…bacon sea salt…bacon lube…and even just a package of bacon delivered on Christmas morning. It’s all good.

5. I didn’t quit Twitter, or Facebook, when G+ arrived

And neither did anyone else. All G+ did was give us yet another place to spread some social networking love. I think I’ve hit my limit of social networks to maintain, I believe I have enough. I use each for a unique purpose and can’t imagine I need any more. I would like the ability to merge my G+ accounts though, here’s hoping Google can find a way to get that done in 2012 for me.

6. I did not travel to Brazil

Despite being personally invited to attend SQL Saturday #100 in Sao Paulo, I did not attend. Here’s hoping I do make it to Brazil one of these days. Good weather, good food, good drinks, and that girl from Ipanema.

Look! It's something deep friend in bacon! IT MUST BE GOOD!

7. I was not asked to star in my own reality show

Which seems odd because apparently everyone in America gets to be on their own show, otherwise we wouldn’t have ‘Jersey Shore’, or the Kardashians, or Nancy Grace. Even my sister-in-law (Heather M.) was on a show this year.

8. Still waiting on those funds from Nigeria

I have no idea what the holdup is. I gave my account numbers, just as Mr. Yan asked. I haven’t heard back from him, but I have heard from many others who have made similar requests for my bank information. Which seems completely normal because I am in the habit of just giving those numbers to anyone who needs them. Like when I get a coffee at Starbucks. I could hand them my card, but yelling out bank routing numbers is more fun. Sometimes they just give me that coffee for free.

9. I did not get Hootsuite to pay me $5.99 a month

I can’t stand Hootsuite. To be specific, I can’t stand being treated poorly. Then again, it’s not as if I was a customer of theirs anyway, so I guess it’s OK for them to pretend I don’t have half a brain. I’m leaving that post as “featured” as a reminder to everyone how to NOT do customer service through the use of social media. As if we needed any more evidence of that, right?

10. Did not win the chance to be launched into orbit

Has anyone else noticed that there are no details about coming back to Earth? I only ever see talk about the launch, not the recovery. Anyway, I didn’t win the contest which makes me part of the 99%, I believe.

11. Was not escorted from a plane for playing Words With Friends after the door was closed

Considering I flew over 75,000 miles last year, I’m a little surprised I didn’t have some kind of incident with regards to the TSA or FAA regulations. I have no idea how many miles Alec Baldwin flies, but he sure loves to play Scrabble. As for the TSA, I managed to get through the year with only a handful of “Freedom Pats” being administered, and an equal number of exposures to the radiation from the backscatter machine that does next to nothing to enhance security. But hey, don’t take my word for it, just listen to one of the voices behind the creation of the TSA, John Mica. I was one of the 31,956 people that signed the petition to abolish the TSA, and you can read the response I was given. Hey, at least they wrote back, right?

They don’t. Not one thing. I don’t buy SQL Server, I get it for free. My clients and customers buy it, so maybe they care. You might think that would mean I care, but I don’t. I have no interest in listening to folks at MS tell me that they raised prices but are *still* cheaper than Oracle. I just don’t. They eliminated Datacenter edition? Whatever.

Here’s what I care about: standards.

I want people using database tools and products to have a similar experience. It’s one of the things I can’t stand about NULLs, there are no standards applied to third party vendors that make default connections to data sources, which leads to no end of confusion for end users. It sucks, and I’m mad as hell.

Here’s an example for you. Apparently there is a difference in the default auto-commit settings between Oracle and MSSQL. Can someone explain why that would be the case? You know what, I don’t care why, it doesn’t matter. What I want is for such things to be standard. If I have to be the SQL Server admin and I get an instance of Oracle shoved down my pants then I should have a fighting chance at supporting it, and vice-versa. So why don’t we have some standards regarding transactions, logging, recovery, etc.? Is that too much to ask?

Most people aren’t going to care about these licensing costs, regardless of platform. What they care about is ease of use and standards. They care about known expectations. The difference in a few dollars per core means nothing to the average DBA. They care about performance, administrative overhead, and ease of use. They care about standards and configurations and being able to recover data that is erased by accident.

I’m sure the CFOs and CIOs in the world quibble over the price of the technology, but I am equally sure that those CFOs and CIOs are so far removed from doing actual work that they don’t appreciate the little things, such as an auto-commit setting, and how much money it will cost them to get things back on track. What money? Well, how about the hours lost for your admin (that you decided didn’t need any extra training) that needs to recover from an end-user mistake? Care to add those hours up? Probably not, that is the type of thing to get swept under the rug, especially when you want to show your boss how well everything is going.

If it is really too much to ask for some type of standards, then can I at least get some frozen yogurt?

In his book Outliers: The Story of Success, Malcolm Gladwell mentions about how most people need roughly 10,000 hours of training at something before they are an expert. Despite my disagreeing with many of the book’s assertions, that was one statement I did not take issue with. I know that it takes dedication and hard work to achieve a goal. I’ve seen it, time and time again. The earliest known examples I was able to witness were (no surprise) related to my experiences as a basketball coach.

Many years ago I met a man named Rob Fodor while we were working at the same summer camp together as coaches. Rob had been a standout player in high school and college and was even drafted by his hometown Detroit Pistons. One thing you noticed right away about Rob was that he could shoot the basketball really well. The first time I saw him give a talk to the whole camp about shooting he spent the first ten minutes going through drills with just one hand (because if you can’t shoot with one, you can’t shoot with two, the second hand is just for balance anyway). So Rob would be out there warming up (and showing off) with one hand shots from all over the court and using both hands. He kept backing up until he would make half court shots with just one hand…either hand…

Part of Rob’s talk was about how long it took for him to become a good shooter. Not just good, actually, but so good he would be drafted by a pro team. That was his goal, he wanted to play professional basketball. And in order to be the best, Rob knew what needed to be done.

He needed to practice.

Rob would shoot 2,000 shots a day. He did 200 shots from 10 different spots on the floor. Know how long it takes to do 2,000 shots? Hours, that’s how long. But he did it, because he had a goal. But Rob knew something else, too.

See, Rob knew that there was some other kid that was shooting 2,000 shots a day as well. And that on some Friday night, that kid would go to the movies, and Rob would go do his 2,000 shots, and in Rob’s words:

“I KNEW I WAS BEATING HIM!”

And that gave Rob the incentive to do the same thing, over and over. His hard work would put him at a certain level of ability, and instead of going to the movies he would go practice and would then have the edge on those that chose to take the night off.

So why am I telling you this story?

If you want to be good at SQL Server, you gotta put in the hours. You can’t just go take an exam and expect that will make you a DBA. There is no substitute for experience. Right now, you and another database professional are of equal skill. Tonight they are going out for a bacon martini (or two). You are going to stay home and build some VMs in order to install the CTP3 of SQL2012  just so you can try out some new features.

Come tomorrow, guess who will have the edge on the other?

Trust me, the hard work will pay off in the long run, it always does.