operating-system-release

[Eli Zaretskii]

I've looked at the code added for the OS release support, and there
are a couple of design issues there that I don't understand:

First, why is there a need for the C-level accessor function
get_operating_system_release?  The C code could simply look at the
value of the variable Voperating_system_release.

Second, why is the value of Voperating_system_release set to nil if
the underlying library functionality (<sys/utsname.h> and the function
uname) is unavailable, but get_operating_system_release returns an
empty string in that case?  Isn't it better to have the default value
of Voperating_system_release be an empty string as well (or perhaps
something like "<UNKNOWN>")?

Finally, and most importantly, if we add the call to uname, why not
store all of the info it returns in some Lisp data structure, not just
the OS release?

(I came up with these questions because I wanted to add the missing
documentation about this new feature to both NEWS and lispref/os.texi.
I will wait with adding the docs until the design issues are
resolved.)