Re: Emacs as word processor / Text Properties

["Pascal J. Bourguignon" <pjb@informatimago.com>]

To me, text properties, and faces as text properties, have always felt
like a kludge.


At best, they may be used to transport information from the modes to the
display engine, like the faces, but otherwise (apart for quick & dirty
jobs), it seems to me that in general modes sh/would have more pertinent
lisp data structures, and use them to set the face in the buffer to
represent that data structure.

I know that emacs puts a lot of importance on the buffer data structure
(a sequence of characters with attributes), which is all right for a
text editor, but when you write _applications_ in emacs, it is much too
limiting.  And a word processor is an application, with more complex
data structures than a sequence of characters.

So indeed, when copy-pasting text in a word processor, depending on the
range of text, ie. depending on the structure that is copied: simple
sequence of characters, punctuated sequence of words, paragraphs,
sections, chapters or whole document, you may or may not want to bring
along the styles.  But most often you will want to bring along the
structure.   


eg. probably, if your selection crosses element boundaries, for the characters
that have only one "tag", you won't want to bring the style and faces,
but for characters that are inside elements copied in whole, you will  
want to apply the style for those elements in the target place.  In both
cases, the faces as copied text properties  are useless.

-- 
__Pascal Bourguignon__
http://www.informatimago.com/