Emacspeak X and non-X keyboard issues

["T. V. Raman" <raman@xxxxxxxxxxx>]

strange, ...

as I read your  message I was about to say set terminal type to
eterm in the term that comes up  when you do M-x term.

but the fact tha you're seeing this with slime, and also the fact
that m-x term does set term ype correctly  on its own maesme
believe that his ha nothing to do with terminal type.

Try disabling the advice on the command invoked by backkspace for
a start, we can debug it from there. 

>>>>> "Tim" == Tim Cross <tcross@xxxxxxxxxxx> writes:
    Tim> Hi Raman, here is a test for your memory......
    Tim> 
    Tim> Back in about 1998, when I was a new user, I had some
    Tim> problems with emacspeak and the handling of some
    Tim> characters under X that caused an error. I'm not sure if
    Tim> you remember it or not. At the time, my familiarity with
    Tim> emacs, elisp and emacspeak was too limited to know how
    Tim> to fix this - or even find how to investigate it
    Tim> effectively.  However, I'm now thinking my skill may
    Tim> have improved enough to track this down and with luck,
    Tim> find a fix.
    Tim> 
    Tim> An example is when you run M-x term. Running this under
    Tim> a non X windows environment has no problems. However,
    Tim> when you run it under X, you get an error each time you
    Tim> try to do something which generates one of the special
    Tim> characters, such as backspace or up/down arrow keys. In
    Tim> fact, you get an error as soon as the term mode
    Tim> starts. The output from the debugger is as follows
    Tim> 
    Tim> Debugger entered--Lisp error: (wrong-type-argument
    Tim> number-or-marker-p return)
    Tim> term-emulate-terminal(#<process terminal>
    Tim> "//home/tcross/Mail
\n")
    Tim> 
    Tim> This only occurs under X and only when running
    Tim> emacspeak. It does not occur without X or without
    Tim> emacspeak. It also does not occur is you run under X but
    Tim> within an xterm using the -nw switch.
    Tim> 
    Tim> Anyway, I'm raising this issue again as it is now
    Tim> causing me problems with slime mode. The error is
    Tim> different, but I suspect the cause is the same and
    Tim> again, it only occurs when running slime mode under
    Tim> X. If I run under a console or I run from within an
    Tim> xterminal with the -nw switch, you don't get the
    Tim> error. The error I get with slime (from the debugger) is
    Tim> as follows -
    Tim> 
    Tim> Debugger entered--Lisp error: (wrong-type-argument
    Tim> number-or-marker-p backspace)
    Tim> emacspeak-self-insert-command(1)
    Tim> call-interactively(emacspeak-self-insert-command)
    Tim> 
    Tim> The slime package needs an emacspeak module to allow it
    Tim> decent voice support. I recently started doing this and
    Tim> then on the weekend ran into this problem again. It
    Tim> wasn't an issue last time I used slime, but both emacs
    Tim> 22 and slime have had an upgrade since then, plus I have
    Tim> had to switch from using cmucl to sbcl. This means
    Tim> unfortunately, I'm not sure exactly which piece of the
    Tim> equation has triggered this problem, but as it seems the
    Tim> same as the term problem, fixing one will likely either
    Tim> fix the other or provide guidance on how to do it.
    Tim> 
    Tim> What I'm hoping for is any guidance you can give to help
    Tim> narrow down my search space a bit. From looking at
    Tim> dribble files and view lossage, it seems that under
    Tim> console and xterm, a backspace is represented with the
    Tim> traditional ^H - or more specifically, the ascii
    Tim> number. However, under X Windows, it seems to be
    Tim> represented by the symbol <backspace>.  The same seems
    Tim> to hold for other "special" characters like return, up,
    Tim> down, left and right. I'm not up on the key mapping
    Tim> differences between X and non-X environments, but do
    Tim> vaguely remember reading something about this some time
    Tim> ago.
    Tim> 
    Tim> Any advice on possible things to look for or where the
    Tim> fix should be would be appreciated. For example, I
    Tim> suspect it is somehow tied in with comint mode as the
    Tim> problem only seems to occur whith modes that are
    Tim> interacting with a comint process. If my SWAG
    Tim> (Scientific Wild Arse Guess) is correct, to obtain
    Tim> consistency, we possibly may need some sort of mapping
    Tim> which will provide consistency in what functions see.
    Tim> 
    Tim> For the rest of the list, if any of you run emacspeak
    Tim> under X, but not within a terminal emulator like xterm,
    Tim> I'd be intrested if you experience the same problem when
    Tim> running M-x term. To test, set debug on error to true
    Tim> and then just start term. You should get an error
    Tim> straight after being asked if term should use bash (or
    Tim> whatever default shell your system has). Hitting return
    Tim> will generate the error. If you don't get an error,
    Tim> enter a letter into term and hit backspace and see if
    Tim> that generates an error. If you cannot generate the
    Tim> error, I definitely want to hear from you as this may
    Tim> help identify what is different between my systems and
    Tim> yours. If you do get the error, it wold be good to hear
    Tim> as that means its less likely to be something specific
    Tim> to my installations (i'm pretty confident it isn't a
    Tim> local config problem as I've been able to reproduce it
    Tim> on different systems and under specific circumstances
    Tim> etc.
    Tim> 
    Tim> Tim^��^��ݲ��jq&j)kjf��ǫ���zՖZm��ޙ�,�椱�޷�ڙ�����,��6����jf���&j)\�c�>m��ޙ�,�椱�޷�ڙ�������fiקj���%z�������B��?�*�~����^��zf��:.�˛���m��kay�r�^jIb��+r��쨺��v������i�)y�%��lzwfj)m��r�^jJު笵�/j���n�+aj˛��-������r��z��zZ