Regression identified in Service Provider library, new version pushed

[Cantor, Scott]

> So a new version was released without bumping the version number?

The affected library was bumped, exactly as always.

> That seems extra unfortunate.  We’re hitting the bug now as the InCommon
> metadata file is expiring across our servers, and when I looked initially I didn’t
> realize the version had changed because of this.

A yum update would have patched the affected library, again, exactly as always. If there were a bug in libgcc, you wouldn't get the fix by updating applications, this is no different. What you are implying would be incorrect behavior on the part of the packaging.

If you mean Windows, the installer was updated to a new service release version that is labeled. If there's some other mechanism to use there, I'm listening.

And the issue was announced on top of that.

-- Scott