Re: RDS

On Fri, 5 Feb 1999, Rekaish Bhardwaj wrote: > Has anybody ever had problems with ESQLC and RDS. Not personally, but that's because I know what problems to avoid. I've seen plenty of other people run into problems, though. > Is there anything different I need to do when creating a customised > runner using filename.ec as opposed to filename.c No, and Yes. No; you compile with the correct version of ESQL/C and link the code. Yes; you have to know which is the correct version of ESQL/C. You don't say which version of I4GL-RDS you are using, but fortunately that doesn't matter very much because your options are limited: I4GL-RDS ESQL/C 4.12..4.20 4.12 6.00..6.05 6.00 7.20 7.24 The hard part is getting a 4.12 ESQL/C for your 4.x I4GL-RDS, or the 6.00 ESQL/C for your 6.x I4GL-RDS. Neither of those versions of ESQL/C is orderable in their own right. The relevant version of ESQL/C is part of the I4GL (c-code) system, so the correct way to compile the ESQL/C code for inclusion in the p-code runner is to compile it with the c4gl script, not some random ESQL/C. Note that if you try to compile with 5.x ESQL/C and link the result into a 4.x p-code runner, it generally won't work with a bunch of missing symbols, including _iqlocate_cursor(). If you try to compile with a 7.x ESQL/C and link the result into a 6.x p-code runner, you might, or might not, get compile time errors -- but you could expect runtime errors. Don't use a ClientSDK (9.x ESQL/C); it might even link, but will be unreliable at runtime. If you're lucky, you simply get no data returned; if you're unlucky, you get spurious results. Yours, Jonathan Leffler (jleffler@informix.com) #include <wish/I/was/skiing.h> Guardian of DBD::Informix v0.60 (v0.61_02) -- http://www.perl.com/CPAN Informix IDN for D4GL & Linux -- http://www.informix.com/idn