My venerable Alpha 800 was retired long ago, and I've been relying on resources at work to answer the regular queries that I get about VMS and the code examples.
Now being retired however, it's well past due that I have a VMS machine at home again, and with the advent of the community license for x86_64, I've been able to get a virtual machine up and running OpenVMS 9.2-1 and updated the examples code.
Over the last week or so, I've been testing the examples on x86, and I'm happy to report that after some minor tweaks for architecture specific code, all but two pass their DTM regression tests (more on those in a moment).
Obviously, I no longer have VAX, Alpha, or Itanium systems to test. The status of anything specific to those architectures should not have changed, and if you experience an unexpected result running the example code, drop me a line and I'll be happy to work with you to correct the error as long as you can test for me.
Now, the two problem children, both associated with the maths library: ots$powc and ots$powcj.
It appears that there's an issue with the way the C compiler handles double complex numbers presently, and both of these examples operate on this data type. I've confirmed this on the forums, but am yet to see a resolution. Hopefully something will be coming soon, as although this would probably only affect a very small number of applications, scientific apps using things like this tend to be very important.
Posted at August 9, 2023 11:54 AM