DB size

[jda]

>>Sorry to follow up so late, but the inordinate size of "small" db's 
>>is going to be a big problem for me. With my main app I distribute 
>>hundreds of dbs containing perhaps a total of 1K of actual data 
>>that is used for controlling certain functions. When they were 200 
>>KB (V1.x) it wasn't so bad. But now an empty file is clocking in at 
>>1.4 MB! Storing as UTF-16 has nothing to do with this enormous 
>>increase, of course (I did have to increase the segment size from 
>>512 to 1024 in v2.x, too, or I would get crashes, and that accounts 
>>for some of it). I simply can't use hundreds of MB for these files. 
>>I certainly hope there is a way to get the size back to 1.x levels.
>>
>>Jon
>
>I am curious why you would use hundreds of dbs with tiny bits of 
>data each. Seems that combining them into a larger db be reasonable. 
>Or if this is just control data, use non-db files.
>

It's control data, mostly. The main problem for me is that this is 
the approach I took with V1.x, and it was satisfactory. Now my code 
is riddled with calls to these dbs, and I have thousands of users who 
have modified them to suit their own needs. So for me to move to 
another method (e.g. saving as XML) is going to require a ton of work 
and present compatibility problems when my users migrate to V2.

If I had it to do all over again I'd use another approach, and in 
fact I may have to bite the bullet and do so. But it still seems to 
me that an empty db should not require 1.4 MB. If optimizations are 
the problem, that can be dealt with and we will all benefit. If this 
is inherent to the new db structures, then...

Jon