On 03/07/14 01:38, Dimitry Sibiryakov wrote: > 06.03.2014 17:39, Alex Peshkoff wrote: >> Yes, as soon as we stop to call BLR-oriented functions to implement >> SQLDA-oriented API calls converting Sqlda->BLR->IMessageMetadata instead >> Sqlda->IMessageMetadata directly makes absolutely no sense. > Funny things I see in current sources: sqlda is converted into BLR, BLR > is converted > into IMessageMetadata, in remote module IMessageMetadata is converted into > BLR, BLR is > converted into RMessage. Did I miss something? >
Are you afraid we have not enough steps? :) I can explain why BLR is present between IMessageMetadata and RMessage. BLR format is ideal for travelling over the wire - it's HW independent byte stream. And it's not correct that it's intermediate format between IMessageMetadata and RMessage - BLR is sent over the wire by RMessage. Therefore in remote client we have the following order of things: IRequest conians BLR created (typically) by GPRE or GBAK. IStatement contains IMessageMetadata and we build BLR code based on it. Later no matter where did that BLR came from we build RMessage based on it, which transfers that BLR to the server. What about BLR between SQLDA and IMessageMetadata - this step can be definitely removed. ------------------------------------------------------------------------------ Subversion Kills Productivity. Get off Subversion & Make the Move to Perforce. With Perforce, you get hassle-free workflows. Merge that actually works. Faster operations. Version large binaries. Built-in WAN optimization and the freedom to use Git, Perforce or both. Make the move to Perforce. http://pubads.g.doubleclick.net/gampad/clk?id=122218951&iu=/4140/ostg.clktrk Firebird-Devel mailing list, web interface at https://lists.sourceforge.net/lists/listinfo/firebird-devel
