Re: [Psgml-devel] Re: Key sequence C-c C-f C-e uses invalid prefix characters

[Robert J. Chassell]

   > Please understand, no one wants to provide a broken file to someone
   > else; that is not what the discussion is about.

   No, but a lot of people who don't grok the fullness of SGML believe
   that supplying a file with pointy brackets is sufficient.

Good point.  It is the confusion as to what is the `deliverable'.

   ... I would prefer to see psgml-mode eventually becoming the
   default for .sgm[l] and .xml files. By all means keep sgml-mode as
   Bob suggests for those occasions when it's needed, but rename it
   ...

Yes.  Just as I use Mail mode for email, Texinfo mode for Texinfo, I
expect to use SGML mode for SGML and XML mode for XML.

Clearly, the mode should handle proper `deliverables' well and degrade
gracefully to handle files with point brackets as well as the current
SGML mode.  

Hardly anyone, when editing an XML file he just received, will want
Emacs to tell him that the mode into which Emacs automatically put him
is wrong and that to edit the file comfortably he should shift to a
different mode.  Such a message will come across as a bug in Emacs.
But that bug is what seems to be suggested by the phrase `keep ... for
those occasions when it's needed'.

Instead, the mode that Emacs determines automatically for the file
should handle the problem.  Of course, Emacs may give messages saying
that the file looks corrupted or that the assemblage is incomplete, as
in `No DTD found for this file!  Will attempt to continue anyway!
Using very limited, default DTD that may not be appropriate for this
file.'

Does the current PSGML mode gracefully handle a file that has only
point brackets and nothing else?  Does it do as well as the current
SGML mode for these occasions?  If so, it sounds to me that PSGML mode
provides all the features that the current SGML mode provides and some
additional.

On the other hand, if it does not yet well handle a file that has
pointy brackets and not the full syntax, then the mode fails to
degrade gracefully.

-- 
    Robert J. Chassell                         Rattlesnake Enterprises
    http://www.rattlesnake.com                  GnuPG Key ID: 004B4AC8
    http://www.teak.cc                                     address@hidden