Java Development

[simon_hain]

there is a limit, but i don't know if it is linked to handles or only the objects used. you can use object grouping to lower the used handles. reading and writing from/to flash is always a speed bump, i don't think the lookup speed of a hashtable is relevant in most usecases. would need some tests with a profiler i guess.

[Orzech]

Is there any significant security gain when using ControlledAccess object instead of Hashtable with AES-encrypted values, for instance?

Regards.

[simon_hain]

a controlled access object cannot be retrieved, while an encrypted hashtable could be read and attacked on a desktop PC. with AES not being vulnerable, given a good key length, i doubt it is a security flaw not to use controlled access. we'll do it anyhow, better to have one extra layer then one too few.

[Orzech]

Thank you very much, Simon. That's the answer I needed.

Regards.

[tibi95]

Using BB 8900

I've made a little tool, and I created 1000 x 2 PersistentObjects with a Vector with a byte[] inside of random data (15K length).

The 1000 Vectors (with 15K of random data) were saved in 2 different PersistentObjects (under different key) - the same byte[] data was saved in two places.

After creation, I've checked all 1000 pair of saved data taken from the two different PersistentObjects, by testing the values of the two byte[], byte by byte.

All things were ok (BB 8900 seems to have a lot of memory for storage: it was used ~ 30M of saved data)

So, you seem to create a lot of PersistentObjects (I've created 2000).