according to anna

starvation

I’ve been realizing lately how lucky I was growing up.
I’ve never had to worry about my weight or watch what I eat.
Lately though…my clothes don’t fit the same. It is incredibly frustrating.
I know this must sound insane…because I am still not a big person…but I am just not used to this.

I guess I always figured that this wouldn’t happen until after having kids.
And I never realized how much I eat when I am bored…or working… or doing anything for that matter.

Then the other day, I saw a commercial for the Special K Challenge.
Now this I can do! Substitute cereal for a meal? Sounds good to me! I love cereal!
BUT DO YOU KNOW HOW SMALL 3/4th CUP OF CEREAL IS? IT’S TINY!

I’m not so sure about this…

a farting female feline

Our kitten Tina is the sweetest little thing…but she doesn’t like to be held very much. She would much rather be playing.
Except when she is tired…then she will curl up on your lap. She will circle your lap several times, purring and kneeding before finally settling down in one spot and falling asleep.

Now this would be all fine and dandy, except that when Tina purrs… she farts! As she is circling your lap, she’ll let out a couple big ones. Pee-you!

goodbye taurus!

Today was the first day that I drove to work…IN MY NEW CAR!!!!!

ERwin or ERlose?

I’ve been using ERwin(4.1) for the last month or so. For those of you who don’t know, ERwin is a data modeling tool, used by IT developers to visually represent database schemas. I remember hearing about the tool back in college, but I haven’t had the chance to work with it until now.

I was very excited to start working with it because I had heard so many good things. In college, it was presented as one tool from only a few that would generate the DDL for you, which could save a significant amount of time.

Granted, the version I am using is an older one, but I have mixed feelings towards the app.Overall it seems like a pretty powerful tool, but it has its drawbacks.

Advantages

• Reverse Engineering - This is great functionality that allows you to point ERwin at a database (or a script) and it will generate the data model for you. You can even pick which types of objects to import (ie, tables, stored procedures, views, etc…). If you are wanting to model a schema in which all rules and relationships are defined at the database level, this is perfect.
• Forward Engineering - If you database is not yet created, you can first model it and then have ERwin generate the DDL. This has a potential to save developers a lot of time and energy.
• Formatting - ERwin allows you to make all PKs a certain color, or all FKs another color. While this sounds silly, it makes reading the model a bit easier.
• Complete Compare - This is a cool feature where you basically do a reverse engineer (off a script or database) and then can compare the differences between your current data model and the reverse engineer. It will list out all the differences so you can see what things look like side by side and then choose which differences to ignore and which ones to apply in your data model.
• Subject Areas - Sometimes with a large schema, reading the model can be more difficult than building the model. ERwin has subject areas which allow you to divide up the model in a way that makes sense to you. You can either choose each table to pull into a subject area, or you can pick a table (or multiple tables) and ERwin will automatically pull in all related tables within a certain level of your chosen table(s)! I was even more excited when I ran a second Complete Compare and saw that if a new relationship is created between 2 tables (and you chose to have a subject area for all tables related to table X), ERwin will pull it into the appropriate subject areas! How cool is that? So now, I can run a comparison and not worry about maintaining the subject areas.

Disadvantages

• Cost - This application costs a pretty penny. The more developers you want to be able to have access to the models, the more you are going to have to spend. Depending on your requirements, you may be able to satisfy your data modeling needs through other ways.
• Clunky Viewing - Viewing a data model gets clunky at times. There is a “go to” functionality, but it didn’t seem to always take me there. At times I found myself lost in the model not knowing if I should scroll left, right, or just keep zooming out.
• Speed - Perhaps it is just the version that I am working on…but when you start working with 100 or more tables, this application is slow! If you click on a subject area, you wait. If you reverse engineer, you wait. If you edit a subject area, you wait! When I start a reverse engineer from a script first thing in the morning, and come 5 o’clock, it is still reading from the file & therefore I have to kill it, how much work can I really get done? It should be noted however, that the application is doing a lot of heavy work when it reverse engineers a database. I had a about 126,600 lines in my script.
• Referential Integrity Assumptions - I found out pretty quickly that ERwin makes some assumptions about how you are going to use the tool. I tried to start my model by doing reverse engineer against a database with no RI. With all the tables in the model, I tried to establish RI within the model. I used the relationship tool, but ERwin expects that your foreign keys will not be present on your tables until you establish the relationships. Everytime I tried to establish a relationship, it tried to copy the PK from the parent table to the child table to create a FK. It wouldn’t allow me to connect two columns that already existed. I can only assume that this was changed with later versions. I can’t imagine it still behaving this way.
• Difficult workflow with no DB RI - In this case the schema I am trying to model is pretty much set in stone. The tricky part is that the referential integrity is at the application layer rather than the database layer. So when I go to try and model out the schema and include those relationships for the sake of understanding the schema, I have to jump through hoops. First, I use the reverse engineering functionality to build a model of the current database with no RI. Then, I do forward engineering to get all of those create statements into one DDL file. Now comes the fun part. Using macros and text files describing the RI kept at the app level, I build DDL that would create RI on the database, even though that DDL will never get run. I combine that DDL with the create table statements spit out by the forward engineering to get one complete script with all tables and RI. Finally, I can then to a Reverse Engineer against that script to get the complete model! This seemed to work until I tried to include the entire schema rather than a subset (of about 100). And that is when I came into the speed issues. With the entire schema, it will run for over 8 hours, which is not exactly something I want to wait on every time something changes.

Bad Day at the Office

Oh, how I can identify with these people!

http://glumbert.com/media/baddayoffice