headers
Evans H Winner III | 21 May 2010 08:11
Picon
Favicon
Gravatar

Getting up and running problem on W32 + sbcl

Hi.  I have successfully gotten ltk's test programs and 
hello programs (from the manual) working on an MS Windows 
XP system using slime and clisp.  On the same system, 
however, I am running into a problem with the Windows port 
of sbcl.  I am not sure if this is expected or not, and I 
am at a loss about how to debug the problem, just getting 
started with it as I am.  The debugger output from slime 
is below (with some Windows ^M linebreaks edited out). 
 The wish shell does appear, and a graphical debugger 
window appears asking if I "wish to invoke the debugger" 
appears.

Interestingly, if I set *break-on-signals* to t, the 
result is that it breaks earlier, telling me that there is 
no package :fly.  I don't know if this is relevant... 
looking at the code it looks like it shouldn't be asking 
about that unless swank isn't loaded -- which it is. 
 Anyway, perhaps the nature of the problem will be obvious 
to one of you.  (lisp-implementation-version) => 1.0.37. 
 It was 1.0.32, but I tried upgrading to the latest before 
asking here.  Thanks for any pointers.  --Evans

   (FORMATTER grid ~a -row ~a -column ~a~ <at> [ -columnspan 
~a~]~ <at> [ -ipadx ~a~]~
              ~ <at> [ -ipady ~a~]~ <at> [ -padx ~a~]~ <at> [ -pady 
~a~]~ <at> [ -rowspan ~a~]~
              ~ <at> [ -sticky ~(~a~)~])
Compile-time error:
   (in macroexpansion of (FORMATTER "grid ~a -row ~a 
-column ~a~ <at> [ -columnspan ~a~]~ <at> [ -ipadx ~a~]~
(Continue reading)

Permalink | Reply | 2 comments |
headers
Peter Herth | 21 May 2010 10:43
Picon
Favicon

Re: Getting up and running problem on W32 + sbcl

Hi Evans,

while the error doesn't tell me much, you shouldn't get complaints
about fly. As I am
about to release a new ltk version, perhaps first try that one, you
can find it at:

http://ltk.rplay.net/svn/branches/ltk/repl/ltk.lisp

and tell me about any problems you encounter with that version.

Regards,
Peter

On Fri, May 21, 2010 at 8:11 AM, Evans H Winner III <thorne <at> unm.edu> wrote:
> Hi.  I have successfully gotten ltk's test programs and
> hello programs (from the manual) working on an MS Windows
> XP system using slime and clisp.  On the same system,
> however, I am running into a problem with the Windows port
> of sbcl.  I am not sure if this is expected or not, and I
> am at a loss about how to debug the problem, just getting
> started with it as I am.  The debugger output from slime
> is below (with some Windows ^M linebreaks edited out).
>  The wish shell does appear, and a graphical debugger
> window appears asking if I "wish to invoke the debugger"
> appears.
>
> Interestingly, if I set *break-on-signals* to t, the
> result is that it breaks earlier, telling me that there is
> no package :fly.  I don't know if this is relevant...
(Continue reading)

Permalink | Reply | 1 comment |
headers
Evans H Winner III | 22 May 2010 08:52
Picon
Favicon
Gravatar

Re: Getting up and running problem on W32 + sbcl

On Fri, 21 May 2010 10:43:29 +0200
  Peter Herth <herth <at> peter-herth.de> wrote:
> Hi Evans,
> 
> while the error doesn't tell me much, you shouldn't get 
> complaints
> about fly. As I am
> about to release a new ltk version, perhaps first try 
> that one, you
> can find it at:
> 
> http://ltk.rplay.net/svn/branches/ltk/repl/ltk.lisp

Thank you.  I tried that file and got the same result.

However, I realized that I had not tried it outside of
slime, just using a command line.  I find that when there
are mysterious problems in Lisp, about 60% of the time if
I take slime out of the equation then things magically
work (and if they don't work, at least the debugger output
is comprehensible).  So the problem apparently is related
to slime.

I would be happy to take any actions to try  to figure out
what the problem is, but I have to admit that I wouldn't
know where to start without some suggestions.
Permalink | Reply | 0 comments |

Gmane