28-Jul-2006, 08:29 AM
|
 | Guest | | | | | | | | |
| Re: Difference between Designing Relational & Non-Relational Database
Fully normalize if possible.
Partially normalised or de-normalised databases are for
* read-only databases
or
* complex data structures only if the database engine fails.
"Cicada" wrote in message
Reference:: Sikh Philosophy Network http://www.sikhphilosophy.net/information-technology/12444-difference-between-designing-relational-non-relational.html
news:e13QQ6enGHA.4364@TK2MSFTNGP05.phx.gbl...
> I am very well in designing Database in Excel.
> To optimize its speed, in my experience, Table Format usually should not
be
> "Normalized".
> "Use Space to exchange Time" is often important skill.
> Sometimes, I even need to create Index-Table myself to teach Excel to
search
> in order to get better performance.
>
> However, I am new to Database.
> Therefore, I have read books about Database (All talk about relational
> database only, but not the traditional one) and find no problem in its
> difference to Excel now.
> I also understand the several "Theoretical" Normalization steps.
>
> However, I feel very very struggle in how to optimize my 1st & real
Database
> design, and I have no time to "try and error" for this project.
> Database program gives me too much choice in table design.
> I can set the table to fully normalized, or partially normalized, or even
> Not normalized.
> Some people tell me that, in practical work, fully normalized table may
not
> offer the best optimization, and some skills in Excel may also be useful.
> But I cannot further find this kind of information in detail.
>
> Also, another critical point is:
Reference:: Sikh Philosophy Network http://www.sikhphilosophy.net/showthread.php?t=12444
> My adopted development tools is for Pocket PC, which is not as fully
> Relational as MS Access.
> i.e. I must know the traditional database design concept also, to design
> this partial relational database. (e.g. Skills of Data Searching,
matching,
> ......)
> (Is it similar to Excel as I think?)
>
> Could anyone kindly provide me some useful website or advice? Many Thanks!
>
>
 |