Re: Segmentation fault. Informix SE and Linux.

In article <m3hfakcs4o.fsf@yakisoba.forte-intl.com>, Ronald Cole <ronald@forte-intl.com> writes >"Art S. Kagel" <kagel@bloomberg.net> writes: >> The 4GL you downloaded was compiled with the older GCC, GNU and Linux >> recently adopted the EGCS version of GCC as the official version of GCC for >> Intel Pentium platforms. The generated code and libraries are incompatible >> with the older code compiled into the 4GL runner. Get out the RH CD and >> install the RPMs for the older GCC compatibility. Then recompile your >> customizations and link using that gcc, linker, and libraries. There are >> instructions included in the RPM, look around in docs. Both development >> systems will be co-resident and you will be able to switch as needed, I just >> don't remember how right now. > >In /usr/informix/bin, change c4gl, cfgldb, cace, cfglgo, cperf, and >esql as follows: > >CC=${CC:-cc} > to >CC=${CC:-i386-glibc20-linux-gcc} > NO! Don't edit scripts, set the environment variable INFORMIXC=i386-glibc20-linux-gcc Something like that? >> This is not Informix's fault, except for not monitoring Linux development >> more closely, the change over happened during testing. > >Yes, it was Informix's fault. They documented the list of features >that they deliberately chose not to port to Linux. But what about not >being PAM aware? I found the problems within minutes of installing >the product. So, was QA aware of the problems during testing, or was >QA just as clueless as the porting engineer? There is no excuse for a >half-assed port with problems that cripple it's use in a production >environment. Informix just isn't serious about Linux. If they don't >know how to port their product to Linux, then they shouldn't be >charging for their "efforts". Instead of "you get what you pay for" >(as with Oracle's and Sybase's Linux efforts) it's more like "caveat >emptor" here. They know I got "sucker" stamped on my forehead because >I bought IDS-7.30 on Linux "twice". Check my posts, I've been waiting >for over a year for the shadow password fix... > -- David Williams