Board index » delphi » dBase vs Paradox vs Interbase

dBase vs Paradox vs Interbase

I'm about to construct a system in Delphi which will be multi user and
very data intensive.  I come from a Clipper background, and I'm wondering
about performance and stability issues.  Data operations in Clipper
(using DBFs) REALLY slows down when appending, changing a key or deleting
records when the table becomes large (over 1 MB).  I will be performing
massive updates on key fields frequently, and I would like to know how  
Paradox and Interbase perform in comparison to dBase tables.  By the way,
this application will run multithreaded on one PC, so network speed and
SQL benefits are of no consequence.

Another question:  Does anyone know if Delphi32 can have more than one
thread simultaneously accessing the same (single user) Interbase
database?  I would imagine not, but it's worth asking, right?

Thanks for any help.

 

Re:dBase vs Paradox vs Interbase


Quote
In article <30F5C2A9.6...@ix.netcom.com> John Crane <johnn...@ix.netcom.com> writes:
>I'm about to construct a system in Delphi which will be multi user and
>very data intensive.  I come from a Clipper background, and I'm wondering
>about performance and stability issues.  Data operations in Clipper
>(using DBFs) REALLY slows down when appending, changing a key or deleting
>records when the table becomes large (over 1 MB).  I will be performing
>massive updates on key fields frequently, and I would like to know how  
>Paradox and Interbase perform in comparison to dBase tables.  By the way,
>this application will run multithreaded on one PC, so network speed and
>SQL benefits are of no consequence.
>Another question:  Does anyone know if Delphi32 can have more than one
>thread simultaneously accessing the same (single user) Interbase
>database?  I would imagine not, but it's worth asking, right?
>Thanks for any help.

I think the only reliable way to answer this question would be to set up some
benchmark tests representative of what you are trying to do -- perhaps using a
test-data generator program for part of it so that you get a representative
spread of values on your index trees.  Time it and plot the results.  I know
that a lot of us out here would be -intensely- interested in the numbers you
obtain.

/mr/

Other Threads