emorris1000
Registered User.
- Local time
- Today, 04:42
- Joined
- Feb 22, 2011
- Messages
- 125
If (desire to skip backstory) Then Goto 15
Else
Read on
I have been working on this database off and on for ~1 year. The first number of months were just trying to figure out how to programattically upload the data (~250-500k records in ~1k files with inconsistent formats.) But I got that, then spent the next 6 months periodically messing with it or completely forgetting it for a couple of months.
Then I realized that the whole thing had become a tangled mess of garbage filled with a bunch of prototype tools that I had developed while trying to learn how to use Access. I have a programming background, sort of, but had almost no database experience. So I had like 40 queries, terrible naming conventions, redundant forms, and too many temp tables to admit to in public. But there was at least some value in it, I learned quite a bit about things.
I scrapped the entire front end and started again. Cleaned the naming conventions and designed properly. Within 2 months I had something that looked good. 2 primary forms and 4 queries gave me access to the data I wanted in a strong variety of ways. Context menus lead to non-primary form. All data exports to Excel are automated etc etc.
15
I finally launched my alpha executable at work today for some people to play around with, and the people using it like it, a lot. But there's something in my gut that makes me wonder if I have McGuyvered this too much.
Every form is unbound. I used to try and use bound forms but then I decided that I liked combo boxes and drill-down listboxes more than subform/datasheets, and I couldn't figure out how to do that with bound forms so I unbound every form. Also I still don't entirely get bound forms.
There are ~15 tables and technically none have defined relationships. They used to, its just that when I manually split the database it killed the relationships. Everything still worked and I didn't even notice the relationships were gone until a couple of hours ago. Everything still works fine because they are all joined in queries as needed.
I think I have done a good job of removing excess/repeated fields, but I still don't really get what normal form it is, if it even is one.
There's a couple of key fields that exist in almost every table. The relationship structure (if it existed) would be a bit of a spoked wheel with a couple of central hub tables from which a bunch of different sub tables sprang from. But there are some really wierd relationships in there that I am pretty sure are highly innapropriate and would give a proper DB designer a coronary.
Anyways, long story short is that I have no idea what I am doing when it comes the theory of design. I got it to work, but I have no idea if it's going to explode in my face. Part of the problem is that I simply can't find good examples of this type of usage. Basically it is a prototype LIMS system that manages research project data from a couple of batch chemical reactors and the analytical data of the products and the ingredients, as well as sensory data from the reactor.
---------
To the question then. MacGyver or MacGruber? Clearly you can teach yourself to use Access and then do so in non-traditional ways, but what are the mistakes that can lead you to total MacGruber situations?
The only real problems I have right now have to do with proper implementation of a backend on a server, and I am pretty sure I can figure those out (move the front end to the workstation etc etc.). So things are running good, but I have this small voice in the back of my head that keeps saying "yeah, it's working fine now, but this is gonna blow up eventually."
Else
Read on
I have been working on this database off and on for ~1 year. The first number of months were just trying to figure out how to programattically upload the data (~250-500k records in ~1k files with inconsistent formats.) But I got that, then spent the next 6 months periodically messing with it or completely forgetting it for a couple of months.
Then I realized that the whole thing had become a tangled mess of garbage filled with a bunch of prototype tools that I had developed while trying to learn how to use Access. I have a programming background, sort of, but had almost no database experience. So I had like 40 queries, terrible naming conventions, redundant forms, and too many temp tables to admit to in public. But there was at least some value in it, I learned quite a bit about things.
I scrapped the entire front end and started again. Cleaned the naming conventions and designed properly. Within 2 months I had something that looked good. 2 primary forms and 4 queries gave me access to the data I wanted in a strong variety of ways. Context menus lead to non-primary form. All data exports to Excel are automated etc etc.
15
I finally launched my alpha executable at work today for some people to play around with, and the people using it like it, a lot. But there's something in my gut that makes me wonder if I have McGuyvered this too much.
Every form is unbound. I used to try and use bound forms but then I decided that I liked combo boxes and drill-down listboxes more than subform/datasheets, and I couldn't figure out how to do that with bound forms so I unbound every form. Also I still don't entirely get bound forms.
There are ~15 tables and technically none have defined relationships. They used to, its just that when I manually split the database it killed the relationships. Everything still worked and I didn't even notice the relationships were gone until a couple of hours ago. Everything still works fine because they are all joined in queries as needed.
I think I have done a good job of removing excess/repeated fields, but I still don't really get what normal form it is, if it even is one.
There's a couple of key fields that exist in almost every table. The relationship structure (if it existed) would be a bit of a spoked wheel with a couple of central hub tables from which a bunch of different sub tables sprang from. But there are some really wierd relationships in there that I am pretty sure are highly innapropriate and would give a proper DB designer a coronary.
Anyways, long story short is that I have no idea what I am doing when it comes the theory of design. I got it to work, but I have no idea if it's going to explode in my face. Part of the problem is that I simply can't find good examples of this type of usage. Basically it is a prototype LIMS system that manages research project data from a couple of batch chemical reactors and the analytical data of the products and the ingredients, as well as sensory data from the reactor.
---------
To the question then. MacGyver or MacGruber? Clearly you can teach yourself to use Access and then do so in non-traditional ways, but what are the mistakes that can lead you to total MacGruber situations?
The only real problems I have right now have to do with proper implementation of a backend on a server, and I am pretty sure I can figure those out (move the front end to the workstation etc etc.). So things are running good, but I have this small voice in the back of my head that keeps saying "yeah, it's working fine now, but this is gonna blow up eventually."