Re: 4GL Error

See also "Revenge of the CoreDumps" in DejaNews . . . . been there, done that. That was a major pain for me when we upgraded to 7.20.ud6. We were on 7.20.ud2; they recommended an upgrade to fix it. The problem was still there after we upgraded to ud6. I also found a problem within the REPORT function where 4gl didn't like doing a calculation on a decimal within a PRINT COLUMN. After I made those corrections, my coredumping went away. Informix call 789609 touched a bit on this issue. John Carlson Informix DBA WHSmith USA Tony Flaherty wrote: > > Hp-Ux 10.20 > > I've just upgraded our development machine from tools 6.05 to 7.20.UD8 and > engine 7.24.UC5 to 7.30.UC7. > > I've recompiled my code by executing all of the makefiles (C Compiled) No > code changes have been made. This is probably our heaviest used program and > its been running fine under 6.05 for years. > > Now I'm getting the following message generated from an If statement which > compares a FLOAT variable with a literal > > /usr/lib/dld.sl: Unresolved symbol: isnan (code) from > /usr/informix/lib/tools/lib4gsh720.sl > > If I move the contents of the FLOAT to a DECIMAL(10,2) before the IF stmnt > and then test the DECIMAL instead, the code works, it does bomb with the > same error later on. I can't find any reference to this On Informix's web > site. > > I have a call outstanding with Informix but I thought I'd ask the question > while I was waiting. > > I can't reproduce this using a simple program written and compiled from > within I4gl. The offending code works fine compiled using R4GL > > I'm not a C programmer alas, but this looks like a bug in the informix > library to me. > > -- > --------------------------------------- > Tony Flaherty aef@mfs.misys.co.uk > Analyst Programmer > Misys Financial Systems > All statements and opinions are my own, > Misys don't pay me enough to have opinions > on their behalf > > .