Re: Predicting / computing node links

On Sun, Aug 09, 2026 at 01:03:12AM +0200, Denis 'GNUtoo' Carikli wrote:
> But then I have strange things.
> 
> This texinfo in our manual:
> > @node          Non-standard GNU/Linux configurations
> > @subsubsection Non-standard GNU/Linux configurations
> 
> Gives me that in the HTML output file:
> > gnuboot.html#Non_002dstandard-GNU_002fLinux-configurations
> 
> And This:
> 
> > @node          Load Operating System (can be fully encrypted)
> > @subsubsection Load Operating System (can be fully encrypted)
> 
> Gives me this in the HTML output file:
> > gnuboot.html#Load-Operating-System-_0028can-be-fully-encrypted_0029
> 
> So my question is if there is some (informal?) specification for how the
> node links are generated, or if I can make them predictable somehow, to
> be able to make sure to point to a manual node.

Yes, absolutely.  See:

https://www.gnu.org/software/texinfo/manual/texinfo/html_node/HTML-Xref.html

and especially:

https://www.gnu.org/software/texinfo/manual/texinfo/html_node/HTML-Xref-Node-Name-Expansion.html


> And if possible the question also applies across a wide range of
> texinfo versions (to not have to use a very specific texinfo version
> through Guix, and enable anyone to also build our website without Guix,
> on the distribution of their choice).
> 
> Denis.

The links should be stable across Texinfo versions.  According to the
Texinfo manual, the algorithm for node links dates to 2003-2004.