[Raion]

Respectfully, nobody here has been able to demonstrate a real-world example where an IRIX-supporting program will fail because uname -r now properly prints POSIX compliant release information.

Just to alleviate concerns:

I tried 3 GNU configure programs. Screen, spell, gmp all configured correctly.

I will be very happy to ensure GCC isn't broken either for build. That's more than I am willing to get into on Christmas Eve, I hope you and others understand.

Inst does not use uname command to determine anything, as patches and performance copilot didn't complain when I went through it.

Demonstrate it, and depending on how widespread the issue is, I'll determine the best approach. I cannot of course account for every edge case. I cannot however, IMHO, hold back progress because of fear of breaking either ancient historical programs, or a FOSS program with a configure/makefile/C/whatever file that checks the version using uname that we can certainly patch.

Without a properly functioning uname -r, many programs and configure scripts cannot determine whether or not the OS has certain compatibility, as I've stated there's 30 minor point releases that make up 6.5. I am following the POSIX specification.

In all seriousness, I appreciate the concern and input. I'm of the belief you need to show that it's a problem before we /assume/ it is a problem. Not a hypothetical example, a real-world one.

Thank you everyone. Let's keep the thread focused on the educational aspects of this post, not on rather or not trying to correct an incorrect aspect of the base system that was never corrected by SGI (per POSIX) is the right thing to do. I'm not masking IRIX's flaws, I'm trying to correct them. I understand not everyone can appreciate that, but I'm seriously doing the best I can.