centaur mailing list / bug report

[bosch@loria.crin.fr (Guido BOSCH)]

Thanks a lot to Vincent Prunet for the prompt answer to my first
contact with centaur@mirsa.inria.fr. I didn't actually know that this
is an international mailing list, so I beg your pardon for the mailing
in french. (Btw., could you put on that list to let me see what happens in
the Centaur user community)

I've appreciated a lot that there is a way to define the printed
representation of an object in LeLisp via the function
{<object-class>}:prin, but this invokes an other question that could
maybe discussed here: 

Why does the VTP formalism hasn't been implemented entirely on the
basis of LeLisp objects? 

This would not only make it possible to define a short printed
representation (to facilitate the program development) but also to
build pretty printers (I ignore how PPML is actually implemented, but
to do this in an object oriented way using the print function of an
object seems a good way to do so). Any comments?


And here is a bug report: 

Consider the following function to parse and attach a tree to
a given  ctedit, say current-ctedit:

(defun my-parse (file ctedit formalism)
  (let ((parsed-tree ({parser}:tree file ({name}:formalism formalism) ".")))
    (send 'set-variable ctedit parsed-tree)
    't))

Probably this is not the right way to do this (I suppose I should
consult "ct-env.ll" to get a better solution), but anyway, here is
what happens when I call my function:

;; Fol stands for First Order Logic
? (my-parse "test1.Fol" current-ctedit 'Fol) 
= t

Looks not to bad to me, but now I select from the menu bar the
"Display" menu and choose first the "Redisplay" and then the "Redraw"
items.  This makes appear the following error messages in the centaur
window:

? ** sun4 : erreur de la machine : erreur de bus
= ()
? ** sun4 : erreur de la machine : violation de segment
= ()

After that, my ctedid becomes more or less unusable (its behavior is
not predictable, so I will not try to give a description of what
happened exactly).

Looks like a bug to me, doesn't it? 


---------------------------
Guido BOSCH, CRIN/INRIA-Lorraine
Centre de Recherche en Informatique de Nancy (CRIN)
Institut National de Recherche en Informatique et en Automatique (INRIA)
Campus scientifique, B.P. 239            
54506 Vandoeuvre-les-Nancy CEDEX       	
Tel.: (+33) 83.91.20.00 ext. 3058      	
Fax.: (+33) 83.41.30.79                	
email: bosch@loria.crin.fr