Isabelle Mode: Known Problems and Bugs 
======================================

. Sometimes duplicated ruletable screens appear in Emacs 19.
  This manifests itself as a failure for the ruletable buffer
  to be properly highlighted - if you move the X-window,
  you discover there's another one underneath that *is*
  correctly highlighted.  Delete the top one!
  (And if anyone can suggest a fix for this, I'd be most
   grateful - see below).

. Rule tables occasionally have spurious blank lines, caused by
  generation using auto-fill-mode.  (No ill effects).

. Proof state: process filtering can get confused if you
  submit several commands at once to the Isabelle buffer
  (for example, by pasting an interactive proof).
  Hit SPC in the proofstate buffer to update it properly.


None of these is major.  If you find other problems, please let me
know:  send mail to da@dcs.ed.ac.uk





==============================================================================
==============================================================================


Developers' Working Bug List:
============================

------------------------------------------------------------------------------

 - Output redirection
 In case the output redirection function values get corrupted
 and output gets lost from the Isabelle buffer, use
 M-x isabelle <LOGIC-NAME> again to reset the hooks.
 (NB:- not possible if the buffer has been renamed)

This doesn't work!  (try deleting output buffer).
Probably not worth fixing...

------------------------------------------------------------------------------

Emacs 18 - loss of start-up messages - sporadic.
Emacs 19 - ditto, when key pressed before startup completed.

------------------------------------------------------------------------------

It's possible to get the interfacing with sml-mode mixed up.  e.g., if
sml-mode is loaded first (or second?), then key redef doesn't properly
work, and TAB doesn't exit sml-mode.

We need a consistent and stable version of sml-mode to interface with
(and bundle?).

------------------------------------------------------------------------------

. Sometimes duplicated ruletable screens appear in Emacs 19.
  Is this is because the lisp data structure holding frames
  isn't updated quickly enough, so successive calls to
  get-screen-for-buffer-noselect can create duplicate
  screens?  It certainly appears to have something to do with
  the speed of the code.  Can we block somewhere?
  Temoporary patch to screen.el inserts a startup delay.
  But something's a bit odd...

------------------------------------------------------------------------------



   
 
