From stefan-ml at snobis.de  Thu Nov  3 21:33:05 2005
From: stefan-ml at snobis.de (Stefan Nobis)
Date: Thu, 03 Nov 2005 22:33:05 +0100
Subject: [cells-gtk-devel] Error compiling CVS version of cells-gtk
Message-ID: <87u0et8lim.fsf@snobis.de>

Hi.

I just tried to compile the CVS version of cells-gtk (and CVS
versions of cells and hello-c) with CMUCL (19c-pre1-20051019-1)
and got this error:

,----
| ; loading system definition from CELLS-GTK:CGTK;GTK-FFI;GTK-FFI.ASD.NEWEST into
| ; #<The ASDF2037 package>
| ;; Loading #P"/home/stefan/src/extern/cells-gtk/cgtk/gtk-ffi/gtk-ffi.asd".
| ; registering #<SYSTEM :GTK-FFI {5815739D}> as GTK-FFI
| ;; Loading #P"/home/stefan/src/extern/cells-gtk/cgtk/gtk-ffi/gtk-ffi.x86f".
| ;; Loading #P"/home/stefan/src/extern/cells-gtk/cgtk/gtk-ffi/gtk-core.x86f".
| 
| Undefined foreign symbol: "g_free"
|    [Condition of type KERNEL:SIMPLE-PROGRAM-ERROR]
| 
| Restarts:
|   0: [CONTINUE] Return NIL from load of #P"/home/stefan/src/extern/cells-gtk/cgtk/gtk-ffi/gtk-core.x86f".
|   1: [RETRY   ] Retry performing #<ASDF:LOAD-OP NIL {581CAE1D}> on
|                 #<ASDF:CL-SOURCE-FILE "gtk-core" {58163FED}>.
|   2: [ACCEPT  ] Continue, treating #<ASDF:LOAD-OP NIL {581CAE1D}> on
|                 #<ASDF:CL-SOURCE-FILE "gtk-core" {58163FED}> as
|                 having been successful.
|   3:            Return NIL from load of "load".
|   4: [ABORT   ] Return to Top-Level.
| 
| Debug  (type H for help)
| 
| (KERNEL:UNDEFINED-FOREIGN-SYMBOL-ERROR-HANDLER "<error finding name>"
|                                                #.(SYSTEM:INT-SAP #x3FFFC34C)
|                                                #<Alien (* #) at #x3FFFBFE8>
|                                                (14))
| Source: Error finding source:
| Error in function DEBUG::GET-FILE-TOP-LEVEL-FORM:  Source file no longer exists:
|   target:code/interr.lisp.
`----

If i try the cells-gtk-2005-05-29 tarball, everythings works
fine. Any hints what's going wrong?

-- 
Until the next mail...,
Stefan.


From asimon at math.bme.hu  Thu Nov  3 21:45:55 2005
From: asimon at math.bme.hu (Andras Simon)
Date: Thu, 3 Nov 2005 22:45:55 +0100 (CET)
Subject: [cells-gtk-devel] Error compiling CVS version of cells-gtk
In-Reply-To: <87u0et8lim.fsf@snobis.de>
References: <87u0et8lim.fsf@snobis.de>
Message-ID: <Pine.LNX.4.60.0511032237510.16048@csusza.math.bme.hu>



On Thu, 3 Nov 2005, Stefan Nobis wrote:

> Hi.
>
> I just tried to compile the CVS version of cells-gtk (and CVS
> versions of cells and hello-c) with CMUCL (19c-pre1-20051019-1)
> and got this error:
>
> ,----
> | ; loading system definition from CELLS-GTK:CGTK;GTK-FFI;GTK-FFI.ASD.NEWEST into
> | ; #<The ASDF2037 package>
> | ;; Loading #P"/home/stefan/src/extern/cells-gtk/cgtk/gtk-ffi/gtk-ffi.asd".
> | ; registering #<SYSTEM :GTK-FFI {5815739D}> as GTK-FFI
> | ;; Loading #P"/home/stefan/src/extern/cells-gtk/cgtk/gtk-ffi/gtk-ffi.x86f".
> | ;; Loading #P"/home/stefan/src/extern/cells-gtk/cgtk/gtk-ffi/gtk-core.x86f".
> |
> | Undefined foreign symbol: "g_free"
> |    [Condition of type KERNEL:SIMPLE-PROGRAM-ERROR]
> |
> | Restarts:
> |   0: [CONTINUE] Return NIL from load of #P"/home/stefan/src/extern/cells-gtk/cgtk/gtk-ffi/gtk-core.x86f".
> |   1: [RETRY   ] Retry performing #<ASDF:LOAD-OP NIL {581CAE1D}> on
> |                 #<ASDF:CL-SOURCE-FILE "gtk-core" {58163FED}>.
> |   2: [ACCEPT  ] Continue, treating #<ASDF:LOAD-OP NIL {581CAE1D}> on
> |                 #<ASDF:CL-SOURCE-FILE "gtk-core" {58163FED}> as
> |                 having been successful.
> |   3:            Return NIL from load of "load".
> |   4: [ABORT   ] Return to Top-Level.
> |
> | Debug  (type H for help)
> |
> | (KERNEL:UNDEFINED-FOREIGN-SYMBOL-ERROR-HANDLER "<error finding name>"
> |                                                #.(SYSTEM:INT-SAP #x3FFFC34C)
> |                                                #<Alien (* #) at #x3FFFBFE8>
> |                                                (14))
> | Source: Error finding source:
> | Error in function DEBUG::GET-FILE-TOP-LEVEL-FORM:  Source file no longer exists:
> |   target:code/interr.lisp.
> `----
>
> If i try the cells-gtk-2005-05-29 tarball, everythings works
> fine. Any hints what's going wrong?

Looks like the shared libraries (libgthread libglib libgobject libgdk
libgtk) didn't get loaded in time. Try to load them by hand before you
start asdf:oos and see what happens.

Andras


From peter.denno at nist.gov  Thu Nov  3 22:23:55 2005
From: peter.denno at nist.gov (Peter Denno)
Date: Thu, 3 Nov 2005 17:23:55 -0500
Subject: [cells-gtk-devel] Error compiling CVS version of cells-gtk
In-Reply-To: <Pine.LNX.4.60.0511032237510.16048@csusza.math.bme.hu>
References: <87u0et8lim.fsf@snobis.de>
	<Pine.LNX.4.60.0511032237510.16048@csusza.math.bme.hu>
Message-ID: <200511031723.56207.peter.denno@nist.gov>

Hi,

I think Andras is right. In fact, in gtk-ffi.asd I see:

;#+cmu(setf *gtk-lib-path* "/usr/lib/")
#+(OR cmu sbcl)(setf *gtk-lib-path* "/opt/gnome/lib/")  ; For my Suse machine

Which might be a problem if you are not on my Suse machine ;^)

The comment before this code says that defining the path to the library should 
not be necessary (for CMUCL). At least in principle that is true. But I 
recall having problems when I didn't specify it.


I tried the HEAD with CMUCL and it ran OK for me, but there was an error in 
compilation of drawing.lisp. I continued past it (:c 0). drawing.lisp is a 
work in progress -- a few weeks maybe.



On Thursday 03 November 2005 16:45, Andras Simon wrote:
> On Thu, 3 Nov 2005, Stefan Nobis wrote:
> > Hi.
> >
> > I just tried to compile the CVS version of cells-gtk (and CVS
> > versions of cells and hello-c) with CMUCL (19c-pre1-20051019-1)
> > and got this error:
> >
> > ,----
> >
> > | ; loading system definition from
> > | CELLS-GTK:CGTK;GTK-FFI;GTK-FFI.ASD.NEWEST into ; #<The ASDF2037
> > | package>
> > | ;; Loading
> > | #P"/home/stefan/src/extern/cells-gtk/cgtk/gtk-ffi/gtk-ffi.asd". ;
> > | registering #<SYSTEM :GTK-FFI {5815739D}> as GTK-FFI
> > | ;; Loading
> > | #P"/home/stefan/src/extern/cells-gtk/cgtk/gtk-ffi/gtk-ffi.x86f". ;;
> > | Loading
> > | #P"/home/stefan/src/extern/cells-gtk/cgtk/gtk-ffi/gtk-core.x86f".
> > |
> > | Undefined foreign symbol: "g_free"
> > |    [Condition of type KERNEL:SIMPLE-PROGRAM-ERROR]
> > |
> > | Restarts:
> > |   0: [CONTINUE] Return NIL from load of
> > | #P"/home/stefan/src/extern/cells-gtk/cgtk/gtk-ffi/gtk-core.x86f". 1:
> > | [RETRY   ] Retry performing #<ASDF:LOAD-OP NIL {581CAE1D}> on
> > | #<ASDF:CL-SOURCE-FILE "gtk-core" {58163FED}>.
> > |   2: [ACCEPT  ] Continue, treating #<ASDF:LOAD-OP NIL {581CAE1D}> on
> > |                 #<ASDF:CL-SOURCE-FILE "gtk-core" {58163FED}> as
> > |                 having been successful.
> > |   3:            Return NIL from load of "load".
> > |   4: [ABORT   ] Return to Top-Level.
> > |
> > | Debug  (type H for help)
> > |
> > | (KERNEL:UNDEFINED-FOREIGN-SYMBOL-ERROR-HANDLER "<error finding name>"
> > |                                                #.(SYSTEM:INT-SAP
> > | #x3FFFC34C) #<Alien (* #) at #x3FFFBFE8> (14))
> > | Source: Error finding source:
> > | Error in function DEBUG::GET-FILE-TOP-LEVEL-FORM:  Source file no
> > | longer exists: target:code/interr.lisp.
> >
> > `----
> >
> > If i try the cells-gtk-2005-05-29 tarball, everythings works
> > fine. Any hints what's going wrong?
>
> Looks like the shared libraries (libgthread libglib libgobject libgdk
> libgtk) didn't get loaded in time. Try to load them by hand before you
> start asdf:oos and see what happens.
>
> Andras
> _______________________________________________
> cells-gtk-devel site list
> cells-gtk-devel at common-lisp.net
> http://common-lisp.net/mailman/listinfo/cells-gtk-devel

-- 
- Best regards,
   Peter


From peter.denno at nist.gov  Thu Nov  3 22:35:21 2005
From: peter.denno at nist.gov (Peter Denno)
Date: Thu, 3 Nov 2005 17:35:21 -0500
Subject: [cells-gtk-devel] Error compiling CVS version of cells-gtk
In-Reply-To: <200511031723.56207.peter.denno@nist.gov>
References: <87u0et8lim.fsf@snobis.de>
	<Pine.LNX.4.60.0511032237510.16048@csusza.math.bme.hu>
	<200511031723.56207.peter.denno@nist.gov>
Message-ID: <200511031735.22627.peter.denno@nist.gov>

Hi,

One other thing I forgot to mention: I had problems running cells-gtk under 
slime with cmucl. When I tried to run the demo, it just sits there. I 
couldn't even break out of it properly. Is anyone else having that problem?

This is a new problem (with slime?) it worked not too long ago.



On Thursday 03 November 2005 17:23, Peter Denno wrote:
> Hi,
>
> I think Andras is right. In fact, in gtk-ffi.asd I see:
>
> ;#+cmu(setf *gtk-lib-path* "/usr/lib/")
> #+(OR cmu sbcl)(setf *gtk-lib-path* "/opt/gnome/lib/")  ; For my Suse
> machine
>
> Which might be a problem if you are not on my Suse machine ;^)
>
> The comment before this code says that defining the path to the library
> should not be necessary (for CMUCL). At least in principle that is true.
> But I recall having problems when I didn't specify it.
>
>
> I tried the HEAD with CMUCL and it ran OK for me, but there was an error in
> compilation of drawing.lisp. I continued past it (:c 0). drawing.lisp is a
> work in progress -- a few weeks maybe.
>
> On Thursday 03 November 2005 16:45, Andras Simon wrote:
> > On Thu, 3 Nov 2005, Stefan Nobis wrote:
> > > Hi.
> > >
> > > I just tried to compile the CVS version of cells-gtk (and CVS
> > > versions of cells and hello-c) with CMUCL (19c-pre1-20051019-1)
> > > and got this error:
> > >
> > > ,----
> > >
> > > | ; loading system definition from
> > > | CELLS-GTK:CGTK;GTK-FFI;GTK-FFI.ASD.NEWEST into ; #<The ASDF2037
> > > | package>
> > > | ;; Loading
> > > | #P"/home/stefan/src/extern/cells-gtk/cgtk/gtk-ffi/gtk-ffi.asd". ;
> > > | registering #<SYSTEM :GTK-FFI {5815739D}> as GTK-FFI
> > > | ;; Loading
> > > | #P"/home/stefan/src/extern/cells-gtk/cgtk/gtk-ffi/gtk-ffi.x86f". ;;
> > > | Loading
> > > | #P"/home/stefan/src/extern/cells-gtk/cgtk/gtk-ffi/gtk-core.x86f".
> > > |
> > > | Undefined foreign symbol: "g_free"
> > > |    [Condition of type KERNEL:SIMPLE-PROGRAM-ERROR]
> > > |
> > > | Restarts:
> > > |   0: [CONTINUE] Return NIL from load of
> > > | #P"/home/stefan/src/extern/cells-gtk/cgtk/gtk-ffi/gtk-core.x86f". 1:
> > > | [RETRY   ] Retry performing #<ASDF:LOAD-OP NIL {581CAE1D}> on
> > > | #<ASDF:CL-SOURCE-FILE "gtk-core" {58163FED}>.
> > > |   2: [ACCEPT  ] Continue, treating #<ASDF:LOAD-OP NIL {581CAE1D}> on
> > > |                 #<ASDF:CL-SOURCE-FILE "gtk-core" {58163FED}> as
> > > |                 having been successful.
> > > |   3:            Return NIL from load of "load".
> > > |   4: [ABORT   ] Return to Top-Level.
> > > |
> > > | Debug  (type H for help)
> > > |
> > > | (KERNEL:UNDEFINED-FOREIGN-SYMBOL-ERROR-HANDLER "<error finding name>"
> > > |                                                #.(SYSTEM:INT-SAP
> > > | #x3FFFC34C) #<Alien (* #) at #x3FFFBFE8> (14))
> > > | Source: Error finding source:
> > > | Error in function DEBUG::GET-FILE-TOP-LEVEL-FORM:  Source file no
> > > | longer exists: target:code/interr.lisp.
> > >
> > > `----
> > >
> > > If i try the cells-gtk-2005-05-29 tarball, everythings works
> > > fine. Any hints what's going wrong?
> >
> > Looks like the shared libraries (libgthread libglib libgobject libgdk
> > libgtk) didn't get loaded in time. Try to load them by hand before you
> > start asdf:oos and see what happens.
> >
> > Andras
> > _______________________________________________
> > cells-gtk-devel site list
> > cells-gtk-devel at common-lisp.net
> > http://common-lisp.net/mailman/listinfo/cells-gtk-devel

-- 
- Best regards,
   Peter


From asimon at math.bme.hu  Thu Nov  3 23:40:27 2005
From: asimon at math.bme.hu (Andras Simon)
Date: Fri, 4 Nov 2005 00:40:27 +0100 (CET)
Subject: [cells-gtk-devel] Error compiling CVS version of cells-gtk
In-Reply-To: <200511031735.22627.peter.denno@nist.gov>
References: <87u0et8lim.fsf@snobis.de>
	<Pine.LNX.4.60.0511032237510.16048@csusza.math.bme.hu>
	<200511031723.56207.peter.denno@nist.gov>
	<200511031735.22627.peter.denno@nist.gov>
Message-ID: <Pine.LNX.4.60.0511040019380.16048@csusza.math.bme.hu>



On Thu, 3 Nov 2005, Peter Denno wrote:

> Hi,
>
> One other thing I forgot to mention: I had problems running cells-gtk under
> slime with cmucl. When I tried to run the demo, it just sits there. I
> couldn't even break out of it properly. Is anyone else having that problem?

I'll try this with a fresh cells-gtk and slime, but it has to wait a
few days, because my hands are full right now... Sorry! In the
meantime, asking about this on the slime list may turn up a good idea
or two. (And this is probably unavoidable whether cells-gtk works with
slime for me or not.)

Just a shot in the dark: does changing swank's preferred-communication-style
make any difference?

Andras


From stefan-ml at snobis.de  Fri Nov  4 08:05:39 2005
From: stefan-ml at snobis.de (Stefan Nobis)
Date: Fri, 04 Nov 2005 09:05:39 +0100
Subject: [cells-gtk-devel] Error compiling CVS version of cells-gtk
In-Reply-To: <200511031723.56207.peter.denno@nist.gov> (Peter Denno's message
	of "Thu, 3 Nov 2005 17:23:55 -0500")
References: <87u0et8lim.fsf@snobis.de>
	<Pine.LNX.4.60.0511032237510.16048@csusza.math.bme.hu>
	<200511031723.56207.peter.denno@nist.gov>
Message-ID: <87fyqckfcc.fsf@snobis.de>

Peter Denno <peter.denno at nist.gov> writes:

> I think Andras is right. In fact, in gtk-ffi.asd I see:

> ;#+cmu(setf *gtk-lib-path* "/usr/lib/")
> #+(OR cmu sbcl)(setf *gtk-lib-path* "/opt/gnome/lib/")  ; For my Suse machine

I tried to comment out both lines and to follow the comment (leave
path empty), but that doesn't work. I had to comment out the
second line and uncomment the first one (my GTK libs are in
/usr/lib -- Debian).

After i ignored the error with drawing.lisp, the rest compiles
fine. But the demo raises an error (while showing the
splash-screen):

,----
| * (test-gtk::gtk-demo)
| 
| Error in function LISP::ASSERT-ERROR:
|    Callback ACTIVATE not defined in *widget-callbacks*
|    [Condition of type SIMPLE-ERROR]
| 
[...]
| 
| (LISP::ASSERT-ERROR CELLS-GTK::CB
|                     NIL
|                     "Callback ~a not defined in *widget-callbacks*"
|                     CELLS-GTK::ACTIVATE)
| Source: Error finding source:
| Error in function DEBUG::GET-FILE-TOP-LEVEL-FORM:  Source file no longer exists:
|   target:code/macros.lisp.
| 0] cells-gtk::*widget-callbacks*
| 
| ((CELLS-GTK::CLICKED) (CELLS-GTK::CHANGED) (CELLS-GTK::ACTIVATE)
|  (CELLS-GTK::VALUE-CHANGED) (CELLS-GTK::DAY-SELECTED)
|  (CELLS-GTK::SELECTION-CHANGED) (CELLS-GTK::TOGGLED) (CELLS-GTK::DELETE-EVENT)
|  (CELLS-GTK::MODIFIED-CHANGED))
`----

That assertion seems to be part of the macro DEF-GTK. But
according to the source, ACTIVATE is indeed defined in
*widget-callbacks* and the REPL says the same. So I'm a bit
confused and i think my Lisp knowledge is not good enough to fully
debug this (any hints and tips are welcome).

BTW: What do those "Source: Error finding source" errors mean?

-- 
Until the next mail...,
Stefan.


From stefan-ml at snobis.de  Fri Nov  4 08:26:28 2005
From: stefan-ml at snobis.de (Stefan Nobis)
Date: Fri, 04 Nov 2005 09:26:28 +0100
Subject: [cells-gtk-devel] Error compiling CVS version of cells-gtk
In-Reply-To: <87fyqckfcc.fsf@snobis.de> (Stefan Nobis's message of "Fri, 04
	Nov 2005 09:05:39 +0100")
References: <87u0et8lim.fsf@snobis.de>
	<Pine.LNX.4.60.0511032237510.16048@csusza.math.bme.hu>
	<200511031723.56207.peter.denno@nist.gov> <87fyqckfcc.fsf@snobis.de>
Message-ID: <87br10kedn.fsf@snobis.de>

Stefan Nobis <stefan-ml at snobis.de> writes:

> ,----
> | * (test-gtk::gtk-demo)
> | 
> | Error in function LISP::ASSERT-ERROR:
> |    Callback ACTIVATE not defined in *widget-callbacks*
> |    [Condition of type SIMPLE-ERROR]
> | 
> [...]
> | 
> | (LISP::ASSERT-ERROR CELLS-GTK::CB
> |                     NIL
> |                     "Callback ~a not defined in *widget-callbacks*"
> |                     CELLS-GTK::ACTIVATE)
> | Source: Error finding source:
> | Error in function DEBUG::GET-FILE-TOP-LEVEL-FORM:  Source file no longer exists:
> |   target:code/macros.lisp.
> | 0] cells-gtk::*widget-callbacks*
> | 
> | ((CELLS-GTK::CLICKED) (CELLS-GTK::CHANGED) (CELLS-GTK::ACTIVATE)
> |  (CELLS-GTK::VALUE-CHANGED) (CELLS-GTK::DAY-SELECTED)
> |  (CELLS-GTK::SELECTION-CHANGED) (CELLS-GTK::TOGGLED) (CELLS-GTK::DELETE-EVENT)
> |  (CELLS-GTK::MODIFIED-CHANGED))
> `----

I forgot the backtrace, hope it helps:

,----
| 0] backtrace
| 
| 0: (LISP::ASSERT-ERROR CELLS-GTK::CB
|                        NIL
|                        "Callback ~a not defined in *widget-callbacks*"
|                        CELLS-GTK::ACTIVATE)
| 1: ((METHOD CELLS::C-OUTPUT-SLOT-NAME (PROGN) (# CELLS-GTK:ENTRY T T T))
|     (#() . #(#)) #<unused-arg> #<unused-arg> ENTRY ...)
| 2: ("LAMBDA (G2134 G2135)" #<unused-arg> #<unused-arg> CELLS-GTK:ON-ACTIVATE
|     ENTRY ...)
| 3: ("DEFUN C-OUTPUT-SLOT")
| 4: (CELLS::FINISH-BUSINESS)
| 5: (CELLS::CALL-WITH-INTEGRITY #<unused-arg>
|                                (:OUTPUT NIL)
|                                #<Closure Over Function "DEFUN C-OUTPUT-SLOT"
|                                  {58FA2291}>)
| 6: ((METHOD CELLS:MD-AWAKEN NIL (CELLS::MODEL-OBJECT)) (#(0) . #())
|     #<unused-arg> TEST-GTK)
| 7: (CELLS:TO-BE TEST-GTK)
| 8: (CELLS-GTK:START-APP TEST-GTK::TEST-GTK :DEBUG NIL)
| 9: (INTERACTIVE-EVAL (TEST-GTK::GTK-DEMO))
| 10: (LISP::%TOP-LEVEL)
| 11: ((LABELS LISP::RESTART-LISP
|        SAVE-LISP))
`----

-- 
Until the next mail...,
Stefan.


From peter.denno at nist.gov  Fri Nov  4 15:28:15 2005
From: peter.denno at nist.gov (Peter Denno)
Date: Fri, 4 Nov 2005 10:28:15 -0500
Subject: [cells-gtk-devel] Error compiling CVS version of cells-gtk
In-Reply-To: <87br10kedn.fsf@snobis.de>
References: <87u0et8lim.fsf@snobis.de> <87fyqckfcc.fsf@snobis.de>
	<87br10kedn.fsf@snobis.de>
Message-ID: <200511041028.15756.peter.denno@nist.gov>

Hi Stefan,

I'm not sure I tried the same CMUCL as you, but I tried this one:
ftp://ftp.common-lisp.net/pub/project/cmucl/release/19c/pre-release/ built on 
the same day as what you referenced, 19c-pre1-20051019-1 and did not have any 
problems. (x86, right?) I even tried it with libcellsgtk.so not built. Still 
no problem. 

Where did you get cells and hello-c? Try taking the cells and hello-c code 
from the  2005-05-29 tarball, if that is not what you are using. 

In other news, updating my slime/cmucl got around the problem I mentioned 
earlier. I also cleaned up the compile problems with CVS HEAD. 


On Friday 04 November 2005 03:26, Stefan Nobis wrote:
> Stefan Nobis <stefan-ml at snobis.de> writes:
> > ,----
> >
> > | * (test-gtk::gtk-demo)
> > |
> > | Error in function LISP::ASSERT-ERROR:
> > |    Callback ACTIVATE not defined in *widget-callbacks*
> > |    [Condition of type SIMPLE-ERROR]
> >
> > [...]
> >
> > | (LISP::ASSERT-ERROR CELLS-GTK::CB
> > |                     NIL
> > |                     "Callback ~a not defined in *widget-callbacks*"
> > |                     CELLS-GTK::ACTIVATE)
> > | Source: Error finding source:
> > | Error in function DEBUG::GET-FILE-TOP-LEVEL-FORM:  Source file no
> > | longer exists: target:code/macros.lisp.
> > | 0] cells-gtk::*widget-callbacks*
> > |
> > | ((CELLS-GTK::CLICKED) (CELLS-GTK::CHANGED) (CELLS-GTK::ACTIVATE)
> > |  (CELLS-GTK::VALUE-CHANGED) (CELLS-GTK::DAY-SELECTED)
> > |  (CELLS-GTK::SELECTION-CHANGED) (CELLS-GTK::TOGGLED)
> > | (CELLS-GTK::DELETE-EVENT) (CELLS-GTK::MODIFIED-CHANGED))
> >
> > `----
>
> I forgot the backtrace, hope it helps:
>
> ,----
>
> | 0] backtrace
> |
> | 0: (LISP::ASSERT-ERROR CELLS-GTK::CB
> |                        NIL
> |                        "Callback ~a not defined in *widget-callbacks*"
> |                        CELLS-GTK::ACTIVATE)
> | 1: ((METHOD CELLS::C-OUTPUT-SLOT-NAME (PROGN) (# CELLS-GTK:ENTRY T T T))
> |     (#() . #(#)) #<unused-arg> #<unused-arg> ENTRY ...)
> | 2: ("LAMBDA (G2134 G2135)" #<unused-arg> #<unused-arg>
> | CELLS-GTK:ON-ACTIVATE ENTRY ...)
> | 3: ("DEFUN C-OUTPUT-SLOT")
> | 4: (CELLS::FINISH-BUSINESS)
> | 5: (CELLS::CALL-WITH-INTEGRITY #<unused-arg>
> |                                (:OUTPUT NIL)
> |                                #<Closure Over Function "DEFUN
> | C-OUTPUT-SLOT" {58FA2291}>)
> | 6: ((METHOD CELLS:MD-AWAKEN NIL (CELLS::MODEL-OBJECT)) (#(0) . #())
> |     #<unused-arg> TEST-GTK)
> | 7: (CELLS:TO-BE TEST-GTK)
> | 8: (CELLS-GTK:START-APP TEST-GTK::TEST-GTK :DEBUG NIL)
> | 9: (INTERACTIVE-EVAL (TEST-GTK::GTK-DEMO))
> | 10: (LISP::%TOP-LEVEL)
> | 11: ((LABELS LISP::RESTART-LISP
> |        SAVE-LISP))
>
> `----

-- 
- Best regards,
   Peter


From peter.denno at nist.gov  Mon Nov  7 21:51:13 2005
From: peter.denno at nist.gov (Peter Denno)
Date: Mon, 7 Nov 2005 16:51:13 -0500
Subject: [cells-gtk-devel] Error compiling CVS version of cells-gtk
In-Reply-To: <87pspcgnne.fsf@snobis.de>
References: <87u0et8lim.fsf@snobis.de>
	<200511041028.15756.peter.denno@nist.gov>
	<87pspcgnne.fsf@snobis.de>
Message-ID: <200511071651.13913.peter.denno@nist.gov>

On Monday 07 November 2005 16:21, Stefan Nobis wrote:
> Peter Denno <peter.denno at nist.gov> writes:
> > Where did you get cells and hello-c? Try taking the cells and
> > hello-c code from the 2005-05-29 tarball, if that is not what
> > you are using.
>
> That's the problem. I used the versions from the above tarball and
> it worked. With a bit searching and diffing i found the problem in
> hello-c from CVS:

In the past I have argued against the idea that we could simply asdf-install 
the individual packages cells-gtk, hello-c, and cells -- and expect things to 
work. This illustrates the problem. Each evolves independently. asdf-install 
doesn't solve the integration problem between them. That's one reason I put 
them together, test them and make a tarball.

Thanks for the info.

>
> ,----
>
> | diff -u -r1.2 callbacks.lisp
> | --- callbacks.lisp      6 Apr 2005 18:06:22 -0000       1.2
> | +++ callbacks.lisp      7 Nov 2005 21:14:48 -0000
> | @@ -31,7 +31,9 @@
> |                (fli:make-pointer :symbol-name (symbol-name callback-name)
> | ;; leak?
> |
> |                  :functionp t))))
> |
> |      (print (list :ff-register-callable-returns cb))
> | -    cb))
> | +    cb)
> | +  #+cmu
> | +  (alien::symbol-trampoline callback-name))
> |
> |  (defmacro ff-defun-callable (call-convention result-type name args &body
> | body) (declare (ignorable result-type))
> | @@ -47,8 +49,10 @@
> |         (declare (:convention ,(ecase call-convention
> |                                  (:cdecl :c)
> |                                  (:stdcall :stdcall))))
> | +       , at body)
> | +    #+cmu
> | +    `(alien:def-callback ,name (,(convert-from-uffi-type result-type
> | :return) , at native-args) , at body)))
> | -
> |
> |  #+test
> |  (ff-defun-callable :cdecl :int square ((arg-1 :int)(data (* :void)))
>
> `----
>
> When i add this patch (to the CVS version) everything works with
> current CVS versions of Cells, Hello-C and Cells-GTK.
>
> In both cases (Cells/Hello-C from CVS or from 2005-05-29 tarball)
> i get these warnings:
>
> ,----
>
> | (<unknown>:7651): Gtk-CRITICAL **: gtk_box_pack_start: assertion
> | `child->parent == NULL' failed
> |
> | (<unknown>:7651): Gtk-CRITICAL **: gtk_box_pack_start: assertion
> | `child->parent == NULL' failed
> |
> | (<unknown>:7651): Gtk-WARNING **: Attempting to add a widget with type
> | GtkAccelLabel to a GtkImageMenuItem, but as a GtkBin subclass a
> | GtkImageMenuItem can only contain one widget at a time; it already
> | contains a widget of type GtkAccelLabel
>
> `----

-- 
- Best regards,
   Peter