RE: EDS 2 questions
Robert Craig wrote:
>
>
> jondolar wrote:
> > 1- Lets suppose I have a database with 100 records but each record has
> > very big data sequences say 100,000 elements. Would EDS try to load
> > entire records for an append or an insert or maitain its disk blocks
> > allocation as per database.e and load only a portion of the data?
>
> Using the functions in database.e, you have to read or write
> an entire record (actually, the data or the key).
> There is no way to directly update just a small portion of
> one record on disk.
>
> If you read the data for a record, append to it,
> and write it back, EDS will try to store it in the same place
> on disk if there's enough space, but otherwise EDS
> will look for a big enough empty space in the .edb file.
> As a last resort, EDS will store the data at the end of
> the .edb file, thereby making the file bigger.
>
> > 2- How many concurrent open tables does EDS support?
>
> There is always one "current" database and one "current"
> table. There are functions for choosing a new current
> database and new current table. A database could have
> millions of tables. If you want to copy data from one
> table to another, you have to call a routine to explicitly
> switch back and forth between the two tables.
>
> Regards,
> Rob Craig
> Rapid Deployment Software
> http://www.RapidEuphoria.com
>
>
Thanks Rob for the straight answer. Do you intend, eventually, to buffer
reading/writing to disk in some future versions making EDS a true
diskbase DBMS?
Thanks again
Serge Lavigne
|
Not Categorized, Please Help
|
|
