Re: Run away 4GL's

On Thu, 25 Feb 1999, Art S. Kagel wrote: >I have a vague memory about a particular R4GL pcode runner that had >been incorrectly coded to poll the keyboard for input rather than using >a blocking read. Hmmm...maybe. Whether it was polling (which I doubt) or simply ignoring the return status from read(2) (which I distinctly remember being completely flabberghasted by when I found it) is open to negotiation. There have been bugs in the code at times in the past, but they should have been fixed by now. >If that memory is correct could 4.20 be that version? If I remember it >was quickly caught and a new release made available that did not have the >problem. This goes back about 5-6 years. No; 4.20 was released last year (1998) or even late the year before (1997). So, it would have been a 4.1x version that had the problem if it occurred as long ago as 5 years ago. I don't recall which version numbers were quoted in the original question, but it could be fixed in a point release (UC2, UC3, etc). Either way, with the truss evidence showing that the program is looping on a read(), the problem is fairly clearly a bug which has been fixed in some code versions at some times in the past, and which is not fixed in this particular code in this particular version. As such, it should be reported through normal Informix Tech Support channels for resolution. No-one in this news group is likely to be able to provide a fix via the news group, and the information given warrants asking Tech Support about it. 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