Re: (emacs-version) in the Elisp reference

[Eli Zaretskii]

> From: Jean-Christophe Helary <address@hidden>
> Date: Wed, 21 Jun 2017 08:25:23 +0900
> 
> >> (why hardcode values that do not correspond to the current manual?)
> > 
> > Becuase they are just illustrative examples, and nothing would be gained
> > by changing them to always refer to the current version.
> 
> 
> In that case, I beg to differ. They are in the elisp reference manual to 
> illustrate the values of functions or variables. When you read the manual in 
> the Emacs whose version corresponds to the manual (as it is supposed to be 
> the case) and you evaluate them, there are no reasons why they would not 
> reflect the actual numbers of the Emacs version.

This is a very minor issue, to be sure, and so are the gains.

We could use the value of EMACSVER where the full version is
mentioned.  But how do you suggest to change this part:

> @defvar emacs-major-version
> The major version number of Emacs, as an integer.  For Emacs version
> 23.1, the value is 23.
> @end defvar
> 
> @defvar emacs-minor-version
> The minor version number of Emacs, as an integer.  For Emacs version
> 23.1, the value is 1.
> @end defvar

to use EMACSVER?  Texinfo doesn't have text-processing function to
extract the major and minor version numbers from a version string.
Are we going to introduce 2 more variables?  If so, how will we update
them when producing a release?  We do that automatically for EMACSVER.

Given this complexity, I think you can understand Glenn's response.