Re: [clisp-list] (DRIBBLE) in debugger

[Sam S. <sd...@gn...>]

> * In message <87b...@bi...>
> * On the subject of "[clisp-list] (DRIBBLE) in debugger"
> * Sent on 01 Dec 2002 00:14:07 +0100
> * Honorable Edi Weitz <ed...@ag...> writes:
>
>   edi@bird:/tmp > clisp
>   [1]> (dribble "foo")
>   #<OUTPUT BUFFERED FILE-STREAM CHARACTER #P"foo" @1>
>   [2]> (/ 1 0)
> 
>   *** - division by zero
>   1. Break [3]> (dribble)
>   #<CLOSED OUTPUT BUFFERED FILE-STREAM CHARACTER #P"foo" @1>
>   1. Break [3]> :a
> 
> 
>   *** - The value of SYSTEM::*DRIBBLE-STREAM* is not a stream: NIL
>   1. Break [5]> (quit)
> 
>   *** - The value of SYSTEM::*DRIBBLE-STREAM* is not a stream: NIL
>   1. Break [6]>

this is because debug loop binds i/o stream vars which interferes with
dribble "bindings".
basically, starting and stopping dribble should be on the same debug
loop level.

> I don't think this is a bug but I'd say it is a bit
> uncomfortable. CMUCL shows the same behaviour, LispWorks - on
> executing (DRIBBLE) - will automatically take you back to the top

this is not hard to go - just do (UNWIND-TO-DRIVER T) at the beginning
of DRIBBLE.
personally, I would find this behavior inconvenient.

> level while AllegroCL just carries on without problems - I think this
> is done by setting EXCL::*DRIBBLE-ACTIVE* to NIL.
> 
> I prefer ACL's behaviour and think it would be worthwhile to add it to
> CLISP.

Instead of the current (quite complicated) implementation, we could
just check SYS::*DRIBBLE-STREAM* whenever a stream action is done on a
strmtype_terminal and act appropriately on SYS::*DRIBBLE-STREAM*.
This will not be too hard to implement.
The drawback is some (probably negligible) slowdown of terminal i/o
(which should not be too important since people generally do not use
terminal to do much i/o).

Bruno, what do you think?

Edi, are you prepared to check out the CVS head and test this change?

-- 
Sam Steingold (http://www.podval.org/~sds) running RedHat8 GNU/Linux
<http://www.camera.org> <http://www.iris.org.il> <http://www.memri.org/>
<http://www.mideasttruth.com/> <http://www.palestine-central.com/links.html>
The difference between genius and stupidity is that genius has its limits.