Re: Warn about invalid node names

[Eli Zaretskii]

> Date: Mon, 12 Aug 2013 20:56:38 +0200
> From: Patrice Dumas <address@hidden>
> 
> On Mon, Aug 12, 2013 at 12:07:33PM -0400, Glenn Morris wrote:
> > 
> > Karl Berry wrote (on Wed, 7 Aug 2013 at 23:12 GMT):
> > 
> > >     Surely these people can live with the warnings, or use --no-warn?
> > > 
> > > No, people absolutely hate warnings, and absolutely hate using --no-warn.
> > 
> > Really? This position makes no sense to me.
> > Why wouldn't you want to be warned about something that could be a
> > bug in your document?
> 
> When we did some pre-releases of texinfo 5, there were a lot of new
> warnings, and there were a lot of complaining about that (maybe offlist,
> I don't remember exactly).

Those were warnings about questionable Texinfo practices that might
not be 100% in line with the language definition, but which produce
perfectly valid output.  Thus, those warnings could be considered
unhelpful, at least from some perspective.

Here we are talking about issues that produce bad Info output, one
that causes Info readers to barf.  That's an entirely different ball
game.  I can hardly see how people could be opposed to warnings that,
if disregarded, will yield invalid output.

> The main issue with that specific warning is that it is only an issue
> for the info output, and we try to avoid format-specific warnings.

The warning is about the Texinfo source.  Even if the user does not
produce Info at the moment, the Texinfo sources are invalid and should
be fixed.  There's nothing format-specific here.