Alex Peshkov wrote:

> On Saturday 12 April 2008 00:42, Vlad Khorsun wrote:
>
>> For GTT with transaction scope (ON COMMIT DELETE ROWS) i see no
>> problem to allow modifications in read-only transactions. But i see a
>> problem with attachment scope GTT's as we need in this case to not treat
>> read-only transactions as pre-committed to preserve isolation between them.
>>
>
> In read-only transactions we should enable only modification of transaction
> scope GTT.

Not talking about implementation detail, I think all GTT should be
writable in readonly *databases*. They store connection and transaction
states (like RDB$SET_CONTEXT in SESSION and TRANSACTION).

It will not be difficult to see one inserting data on a SESSION GTT in a
ON CONNECT trigger.

--
Adriano dos Santos Fernandes
http://www.firebirdsql.org