Hi folks.
I'm going over a laboratory application that I started a few years ago, trying to do a better job with some of the table structure that I set up, and improve the application. Right now, I'm working on the very first thing I did when I started building this application. This part has users pick a sieve stack setting for a size test, and record the grams of material weighed from each sieve in the test. This stack is based on the top and bottom size of the sample. In the original, I used a field for top and a field for bottom. In the new structure, it's together. There are over 100 different sieve stack combinations, and the number of sieves used in a combination for a test varies from 3 to 8. The way I was told to set it up is not in a normalized table structure, but it has worked fine for us to date.
I've gone in and broken the tables up into what I believe to be a normalized structure. The funny thing is, the size of the db is bigger with the tables put into a normalized structure than it is with the orignial, non-normal structure.
This seems weird to me. I've done this with other sets of non-normal tables in the application, and saw a subsequent shrink in db size because of the optimization of the table structure. Now I'm wondering if I've done this right, so I'm asking if some of you experienced hands can look at this and tell me if I've gone wrong in my attempt to normalize the table structure.
I appreciate any input in this issue, as it has me scratching my head a bit. I've posted db1, which is non-normal, and db2, which is normalized (I think). I put a few records in these, so you can get the gist of the application use. File size with these few records is negligible, but the application has over 18,000 records in it, and the file size difference is over 1MB, which was unexpected.
Thank you.
I'm going over a laboratory application that I started a few years ago, trying to do a better job with some of the table structure that I set up, and improve the application. Right now, I'm working on the very first thing I did when I started building this application. This part has users pick a sieve stack setting for a size test, and record the grams of material weighed from each sieve in the test. This stack is based on the top and bottom size of the sample. In the original, I used a field for top and a field for bottom. In the new structure, it's together. There are over 100 different sieve stack combinations, and the number of sieves used in a combination for a test varies from 3 to 8. The way I was told to set it up is not in a normalized table structure, but it has worked fine for us to date.
I've gone in and broken the tables up into what I believe to be a normalized structure. The funny thing is, the size of the db is bigger with the tables put into a normalized structure than it is with the orignial, non-normal structure.
This seems weird to me. I've done this with other sets of non-normal tables in the application, and saw a subsequent shrink in db size because of the optimization of the table structure. Now I'm wondering if I've done this right, so I'm asking if some of you experienced hands can look at this and tell me if I've gone wrong in my attempt to normalize the table structure.
I appreciate any input in this issue, as it has me scratching my head a bit. I've posted db1, which is non-normal, and db2, which is normalized (I think). I put a few records in these, so you can get the gist of the application use. File size with these few records is negligible, but the application has over 18,000 records in it, and the file size difference is over 1MB, which was unexpected.
Thank you.
