Re: Maximum size of 4gi-program

informix@ucs.att.com (informix@ucs.att.com) wrote: > > > >Hello Informixers, > > > >we are develloping with I-RDS 4.10.UC2 and have develloped a program > >which has now a size of ~1,1 MB. This program runs on different platforms, > >but not on a SCO-box. On this system we get only a coredump. We have > >discovered that the main function is not reached. If we run the program > >out of the debugger, it runs!!! Now we have made the program smaller > >(its now ~950 kB) and it runs!!! Has anybody else made similiar > >experiences??? Is there a limitation for the size of a 4gi-program??? > > > > Dirch, > > > There is a known problem with executing .4gi files which approach > the size you mentioned (varies on differing platforms but is amazingly > close). The symptom is that the RDS cores and the debugger also will > generate an error message if you make your .4gi large enough. > I don't know if this was fixed in the latest tool release. > You should contact Informix Tech Support. > > > > Bill Williams > > +-----------------------------------------------------------------+ > | Disclaimer: My views, opinions, etc., are not those of AT&T or | > | or ATSI, Inc, nor should they be. | > +-----------------------------------------------------------------+ > > There is a runner option which might solve your problem. Try: fglgo -d mybigprog.4gi I hope I remembered that correctly. Another solution is the compiler version of 4gl which doesn't seem to have any limits. I have created a program which is over 2Mb as a .4gi file which doesn't run, but the 4ge file runs fine and loads within a second on an HP9000. Another way to solve the problem (though painfully) is to decrease the number of strings in the program and move them to a common area, best a text file. Version 4.13 of Informix does better than 4.10 with the size problem, but version 6 is worse.