Currently in Maneage, we give all the version and names of all software built within Maneage. We also report the low-level CPU properties like the endian-ness and its address size.

But we don't report some other critical information like the name of the OS, the kernel version, and the C library name and version: the components we don't install within Maneage.

Because we don't install these components, it can be argued that (with the CPU information), reporting these critical components in the PDF, are more important the reporting our own installed components: because anyone can go into the project's Git history and see the version of AWK that was used for example. But there is no way they can know what OS we used, what kernel it had and what C library everything was built with.

Fortunately extracting these information components is easy and they can be included in the generic statement about software versions. But we can also define separate LaTeX macros to let users include this in the paper without the full software environment (which might not be possible due to its volume in some journals).