> I think we agreed BLOB SUB_TYPE 0 is LONGVARBINARY not BINARY.
>
> In fact DatabaseMetaData.getTypeInfo reports since some months ago
> LONGVARBINARY for BLOB SUB_TYPE 0.
>
> I think the implementation is exactly the same.

Ups. You are right. Sorry.

> >Section 6 (end of it)
> >
> >"Interclient uses a protocol in which the character values are
> >received trimmed, JayBird doesn't." -
> >
> >Are you 100% sure in this (have you checked)? I remember there was

a

> >discussion about this issue, and we agreed to keep space padding

code

> >on the client, because it might happen that server will send a
> >trimmed string.
>
> I'll review it again.

Forgot to mention that most likely discussion was in firebird-devel
list. But I do not remember exactly.

Thanks!
Roman