Re: [Firebird-devel] Delayed statement execution and transaction

2014-03-17 Thread Alex Peshkoff
On 03/16/14 19:41, Dimitry Sibiryakov wrote: > Hello, All. > > Currently when execution of a statement in why.cpp is delayed from > execute to fetch, > the transaction provided to execute call is saved and later used as is, > without any check. > Nevertheless, it is possible that a user b

Re: [Firebird-devel] Delayed statement execution and transaction

2014-03-16 Thread Dimitry Sibiryakov
16.03.2014 17:49, Dmitry Yemanov wrote: > What is a fetch transaction? Txn handle was never passed through the > fetch calls. Oops, my fault. -- WBR, SD. -- Learn Graph Databases - Download FREE O'Reilly Book "Gra

Re: [Firebird-devel] Delayed statement execution and transaction

2014-03-16 Thread Dmitry Yemanov
16.03.2014 19:41, Dimitry Sibiryakov wrote: > > Currently when execution of a statement in why.cpp is delayed from execute to > fetch, > the transaction provided to execute call is saved and later used as is, > without any check. > Nevertheless, it is possible that a user between execute and fet

[Firebird-devel] Delayed statement execution and transaction

2014-03-16 Thread Dimitry Sibiryakov
Hello, All. Currently when execution of a statement in why.cpp is delayed from execute to fetch, the transaction provided to execute call is saved and later used as is, without any check. Nevertheless, it is possible that a user between execute and fetch do something stupid like commit