Selectcoumns

[Lou_Reed]

You I guess that that have not considered that. I would like to know, however, when the tables has blanks in it what changes?



Respectfully,

Lou Reed

[Lou_Reed]

Someone on this forum said that they think the table in question does not even qualify as a database. I would be interested in hearing the arguments so I can make them to the sponsor.

It has been a while since I made these arguments myself, so I am looking to gather as many arguments as possible. I do not mind being subtle if that is what it takes.

I just need to set the sponsor right. I agree it is not database and they must know this. Hopefully, they are open to calling it a heuristic or something else then I world feel better.

Respectfully,

Lou Reed

[June7]

Whether a single large table or a lot of single column tables, may be a database but not a relational database.

These numerous 1-column tables have no relationship to each other. Since the original table has a lot of empty cells then you will get 1-column tables that have a lot of empty records.

If each row of the original worksheet has data to describe an entity, then how does user expect each record of single field tables to continue to describe the entity? Each record is no longer associated with the entity.

[John_G]

when the tables has blanks in it what changes?

Nothing, really. It just means that the resulting one-field table will have the corresponding number of blanks. Your sponsors have to realize though that the positions of the blanks in the one-field tables can not in any way be reliably used to determine where the blanks are in the original large table.

Someone on this forum said that they think the table in question does not even qualify as a database.

No table by itself could be considered a database. A table by itself is just a bucket of "stuff". It's what you do with that table that makes a database - queries, forms, reports that allow you analyze the data. You could have a database that has only one large table in it, but the chances of it being properly normalized would be minimal I think, which would make working with it difficult. Generally speaking, if you had only one large table you would use it as a starting point to develop a proper relational structure.