Re: doc elisp intro cross reference fixes

[Luc Teirlinck]

Stefan Monnier wrote:

   > I did misunderstand the intent.  Nevertheless, there are problems.
   > First of all, there is the possibility of existing code being broken.

   I think calling such cases "broken" is alarmist.

Well, maybe "broken" was not the right word, it is definitely not
going to make Emacs crash, but something getting in a completely
inappropriate group is still potentially confusing to the Custom user,
and looks sloppy.

   > Secondly, before your change, adding a defgroup to a file was
   > harmless, but now one has to be extremely careful with that.

   For someone who used the feature, it's pretty obvious.
   So the use of "extremely" again sounds very alarmist.

Not all files are edited by one person.

This feature would have to be documented in the Elisp manual.  Do you
really recommend that it would be considered OK to use this feature
anywhere whatsoever, say in simple.el?

Somehow, with your feature, we have to make sure that everybody knows
that one can not just add defgroups to files, nor rearrange the order
of defgroups or defcustoms, without checking the effect on defcustoms
without groups.  They can not just move a defgroup to a different file
without checking both files for such defcustoms.  They can not move a
defcustom to another file, without checking that it has a :group.
Of course, one has to be careful about all sorts of stuff when editing
files, but these are _new_ things to worry about.

So where do you recommend use of this feature?  If people use it in
their private files, not part of Emacs, then I agree that there is no
no problem. 

Sincerely,

Luc.