A credit card applicatoin I'm dealing includes a single table with 170 posts.
I believe, it might be preferable to interrupt this table up into more compact, more logical types, though you might have an problem where numerous tables need to be became a member of should you needed data from posts within the various tables.
I'm wondering exactly what the benefits and drawbacks of those approaches are? What will be the benefit of merely one table with this particular many posts?
Just to return to the written text book a little...is every non-key attribute non-transitively dependant overall primary key? Its difficult to imagine what scenario would result in 170 characteristics inside a fully stabilized table.
You may will easily notice us what entity the table signifies therefore we would get to know.
Errrr yeah, obviously you need to join tables together if you have several. That is what relational databases are for. If you are determining what related to this application--without attempting to be rude--I recommend that you are untrained with this particular job as, out of your question, I presume that the understanding of databases, data normalization, entity-relationship modelling and relational algebra is nothing to none.
This is not the kind of factor that anybody here can solve like a simple answer. This is tantamount to "How do you produce a data model?"
If there is a 170 column table, chances are it is so denormalized it isn't even funny. Without more details I can not make sure, but each time I have seen a table that large, I see such things as 'address1', 'address2', 'address3', etc. While (theoretically) this could improve performance, it typically results in issues sooner or later - like if you have someone having a fourth foo, but you've only got foo1 through foo3. Correctly stabilized databases are usually a lot more resistant against change once the needs change.
The only real benefit of that (170 posts) is poor design and also the disavantage is poor design. Hehe.
You are able to really build better queries with increased tables and make more understanding with similar information. You're absolutely correct inside your second paragraph.
I'd imagine lots of individuals fields need to be
That table might have been created by a non-programmer, probably a company analyst or perhaps an eager project manager. You are able to optimize by breaking this table up into more significant tables in line with the business usage. If you're worrying about performance, discover where your database feels safe with joins, some are enhanced for six approximately. You are able to break up accordingly.
Do you've just got a datagrid that feeds of that one large master table with a lot of filters? I'm curious to understand how this table can be used.