From peter.denno at nist.gov Thu Jun 1 14:46:08 2006 From: peter.denno at nist.gov (Peter Denno) Date: Thu, 1 Jun 2006 10:46:08 -0400 Subject: [cells-gtk-devel] SBCL patches In-Reply-To: References: <85sln43k0u.fsf@tristan.br-automation.com> Message-ID: <200606011046.08716.peter.denno@nist.gov> Hi all, OK. The patches that Marco Gidde provided are now in CVS, and the "News" section of the web page has been updated to mention this. Thanks to Marco for the patches and Andras for letting me know that I missed Marco's email! While I'm here, I'm thought I'd ask what people think is missing in cells-gtk. For example, a while back I started some stuff to do drawing, and I intend to someday pick that up again. I would use it to provide a graphing capability (at least draw and manipulate trees). But I'm not in great need of the capability now, and things just seem to work (I have a 3 cells-gtk apps now). So does anyone here have thoughts on how we can make things better? On Tuesday 30 May 2006 09:27, Andras Simon wrote: > Hi Peter, > > Are you considering applying Marco's patches > (http://common-lisp.net/pipermail/cells-gtk-devel/2006-May/000171.html)? > I tested them (except for the one in gtk-ffi.lisp which wraps > defpackage in an eval-when, because I think that the defpackage should > go instead in a separate file), and they not only make SBCL, but also > ACL 6.2 happy; and I see no regression in LWL either. (The context > menu in the Menu tab still doesn't work in LWL, but it didn't before > either.) > > Also, something like > > --- gtk-ffi/gtk-ffi.lisp 19 Feb 2006 20:17:41 -0000 1.19 > +++ gtk-ffi/gtk-ffi.lisp 30 May 2006 13:13:43 -0000 > @@ -133,12 +133,13 @@ > #+win32 "libgtk-win32-2.0-0.dll" > #+macosx "libgtk-win32-2.0-0.dylib") > #+libcellsgtk > - (cffi:load-foreign-library #+cffi-features:unix "libcellsgtk.so" > + (cffi:load-foreign-library #+cffi-features:unix > + (merge-pathnames "libcellsgtk.so" > #.*compile-file-pathname*) #+win32 "libcellsgtk.dll" > #+macosx "libcellsgtk.dylib"))) > ) ; eval > > might be a good idea. Or is there a better way to ensuring that > libcellsgtk.so is found? > > 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 Jun 1 19:30:30 2006 From: asimon at math.bme.hu (Andras Simon) Date: Thu, 1 Jun 2006 21:30:30 +0200 (CEST) Subject: [cells-gtk-devel] SBCL patches In-Reply-To: <200606011046.08716.peter.denno@nist.gov> References: <85sln43k0u.fsf@tristan.br-automation.com> <200606011046.08716.peter.denno@nist.gov> Message-ID: On Thu, 1 Jun 2006, Peter Denno wrote: > Hi all, > > OK. The patches that Marco Gidde provided are now in CVS, and the "News" > section of the web page has been updated to mention this. Great! Thanks! [...] > So does anyone here have thoughts on how we can make things better? I'd like to clean up the code a little bit to reduce the number of compiler notes of the different lisps. The attached patch and a new file (gtk-ffi/package.lisp) is a first step in this direction. The patch is pretty uncontroversial (I think!) except for the - (cffi:load-foreign-library #+cffi-features:unix "libcellsgtk.so" + (cffi:load-foreign-library #+cffi-features:unix + (merge-pathnames "libcellsgtk.so" #.*compile-file-pathname*) part. It works for me in all lisps I tried, but still... Or are users supposed to install libcellsgtk.so in /usr/lib/ or some such? One more thing: since Cells has become Cells2, I think we should have a version of Cells in the repository at least until (if ever) we convert cells-gtk to Cells2. What do you think? Andras -------------- next part -------------- Index: cells-gtk/conditions.lisp =================================================================== RCS file: /project/cells-gtk/cvsroot/root/cells-gtk/conditions.lisp,v retrieving revision 1.3 diff -u -r1.3 conditions.lisp --- cells-gtk/conditions.lisp 16 Feb 2006 18:12:15 -0000 1.3 +++ cells-gtk/conditions.lisp 1 Jun 2006 18:33:29 -0000 @@ -34,5 +34,4 @@ (define-condition gtk-user-signals-quit (condition) ()) -(export '(gtk-user-signals-quit gtk-continuable-error gtk-report-error)) Index: cells-gtk/menus.lisp =================================================================== RCS file: /project/cells-gtk/cvsroot/root/cells-gtk/menus.lisp,v retrieving revision 1.13 diff -u -r1.13 menus.lisp --- cells-gtk/menus.lisp 19 Feb 2006 20:12:16 -0000 1.13 +++ cells-gtk/menus.lisp 1 Jun 2006 18:33:29 -0000 @@ -18,8 +18,6 @@ (in-package :cgtk) -(export '(cgtk-set-active-item-by-path)) - (defmacro with-tree-iters (vars &body body) `(let (,@(loop for var in vars collect `(,var (gtk-adds-tree-iter-new)))) (unwind-protect @@ -305,4 +303,3 @@ (def-widget tearoff-menu-item (menu-item) () () ()) -(export '(gtk-combo-box-set-active)) \ No newline at end of file Index: cells-gtk/packages.lisp =================================================================== RCS file: /project/cells-gtk/cvsroot/root/cells-gtk/packages.lisp,v retrieving revision 1.5 diff -u -r1.5 packages.lisp --- cells-gtk/packages.lisp 19 Feb 2006 20:12:58 -0000 1.5 +++ cells-gtk/packages.lisp 1 Jun 2006 18:33:29 -0000 @@ -16,9 +16,16 @@ |# +(in-package :cl-user) + (defpackage :cells-gtk (:nicknames :cgtk) - (:use :common-lisp :pod :cells :gtk-ffi)) + (:use :common-lisp :pod :cells :gtk-ffi) + (:export #:gtk-user-signals-quit + #:gtk-continuable-error + #:gtk-report-error + #:cgtk-set-active-item-by-path + #:gtk-combo-box-set-active)) Index: gtk-ffi/gtk-ffi.asd =================================================================== RCS file: /project/cells-gtk/cvsroot/root/gtk-ffi/gtk-ffi.asd,v retrieving revision 1.13 diff -u -r1.13 gtk-ffi.asd --- gtk-ffi/gtk-ffi.asd 19 Feb 2006 20:16:57 -0000 1.13 +++ gtk-ffi/gtk-ffi.asd 1 Jun 2006 18:33:31 -0000 @@ -13,7 +13,8 @@ :name "gtk-ffi" :depends-on (:cells :pod-utils :cffi :cffi-uffi-compat) :components - ((:file "gtk-ffi") + ((:file "package") + (:file "gtk-ffi" :depends-on ("package")) (:file "gtk-core" :depends-on ("gtk-ffi")) (:file "gtk-other" :depends-on ("gtk-ffi")) (:file "gtk-button" :depends-on ("gtk-ffi")) Index: gtk-ffi/gtk-ffi.lisp =================================================================== RCS file: /project/cells-gtk/cvsroot/root/gtk-ffi/gtk-ffi.lisp,v retrieving revision 1.20 diff -u -r1.20 gtk-ffi.lisp --- gtk-ffi/gtk-ffi.lisp 1 Jun 2006 14:16:28 -0000 1.20 +++ gtk-ffi/gtk-ffi.lisp 1 Jun 2006 18:33:31 -0000 @@ -16,11 +16,6 @@ |# -(eval-when (:compile-toplevel :load-toplevel :execute) - (unless (find-package :gtk-ffi) - (defpackage :gtk-ffi - (:use :common-lisp :pod)))) - (in-package :gtk-ffi) ;;; POD throw-away utility to convert hello-c/uffi to cffi @@ -136,13 +131,13 @@ #+win32 "libgtk-win32-2.0-0.dll" #+macosx "libgtk-win32-2.0-0.dylib") #+libcellsgtk - (cffi:load-foreign-library #+cffi-features:unix "libcellsgtk.so" + (cffi:load-foreign-library #+cffi-features:unix + (merge-pathnames "libcellsgtk.so" #.*compile-file-pathname*) #+win32 "libcellsgtk.dll" #+macosx "libcellsgtk.dylib"))) ) ; eval (eval-when (:compile-toplevel :load-toplevel :execute) - (export '(+c-null+ int-slot-indexed load-gtk-libs)) (defun gtk-function-name (lisp-name) (substitute #\_ #\- lisp-name)) @@ -425,9 +420,6 @@ (defun cast (ptr type) (deref-pointer-runtime-typed ptr (ffi-to-uffi-type type))) -(eval-when (compile load eval) - (export '(uint c-pointer c-ptr-null c-array-ptr c-ptr c-string sint32 uint32 uint8 boolean - ulong int long single-float double-float otherwise *gtk-debug* load-gtk-libs - col-type-to-ffi-type deref-pointer-runtime-typed gtk-tree-iter +c-null+))) + -------------- next part -------------- #| Gtk ffi Copyright (c) 2004 by Vasilis Margioulas You have the right to distribute and use this software as governed by the terms of the Lisp Lesser GNU Public License (LLGPL): (http://opensource.franz.com/preamble.html) This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the Lisp Lesser GNU Public License for more details. |# (in-package :cl-user) (defpackage :gtk-ffi (:use :common-lisp :pod) (:export #:+c-null+ #:int-slot-indexed #:load-gtk-libs #:uint #:c-pointer #:c-ptr-null #:c-array-ptr #:c-ptr #:c-string #:sint32 #:uint32 #:uint8 #:boolean #:ulong #:int #:long #:single-float #:double-float #:otherwise #:*gtk-debug* #:load-gtk-libs #:col-type-to-ffi-type #:deref-pointer-runtime-typed #:gtk-tree-iter)) From peter.denno at nist.gov Thu Jun 1 20:42:39 2006 From: peter.denno at nist.gov (Peter Denno) Date: Thu, 1 Jun 2006 16:42:39 -0400 Subject: [cells-gtk-devel] SBCL patches In-Reply-To: References: <85sln43k0u.fsf@tristan.br-automation.com> <200606011046.08716.peter.denno@nist.gov> Message-ID: <200606011642.39878.peter.denno@nist.gov> On Thursday 01 June 2006 15:30, you wrote: > On Thu, 1 Jun 2006, Peter Denno wrote: > > Hi all, > > > > OK. The patches that Marco Gidde provided are now in CVS, and the "News" > > section of the web page has been updated to mention this. > > Great! Thanks! > > [...] > > > So does anyone here have thoughts on how we can make things better? > > I'd like to clean up the code a little bit to reduce the number of > compiler notes of the different lisps. The attached patch and a new > file (gtk-ffi/package.lisp) is a first step in this direction. > > The patch is pretty uncontroversial (I think!) except for the > > - (cffi:load-foreign-library #+cffi-features:unix "libcellsgtk.so" > + (cffi:load-foreign-library #+cffi-features:unix > + (merge-pathnames "libcellsgtk.so" #.*compile-file-pathname*) > > part. It works for me in all lisps I tried, but still... Or are users > supposed to install libcellsgtk.so in /usr/lib/ or some such? > > One more thing: since Cells has become Cells2, I think we should have > a version of Cells in the repository at least until (if ever) we > convert cells-gtk to Cells2. Good point. The tarball uses what Kenny Tilton was calling "cells 2," though it might have changed since I last checked. I'll at least check in what I've got. > > What do you think? These look fine. I'll check the merge-pathnames thing for libcellsgtk on lispworks. I think the use of /usr/lib or whatever is leftover from the days when we had to specify (or thought we had to specify) the library path. This is better if it works. I'll check lispworks, at least. It might be a few days before I get this in CVS. But I'll make a note so I don't forget ;^) Thanks! > Andras -- - Best regards, Peter From asimon at math.bme.hu Fri Jun 2 09:46:10 2006 From: asimon at math.bme.hu (Andras Simon) Date: Fri, 2 Jun 2006 11:46:10 +0200 (CEST) Subject: [cells-gtk-devel] SBCL patches In-Reply-To: <200606011642.39878.peter.denno@nist.gov> References: <85sln43k0u.fsf@tristan.br-automation.com> <200606011046.08716.peter.denno@nist.gov> <200606011642.39878.peter.denno@nist.gov> Message-ID: On Thu, 1 Jun 2006, Peter Denno wrote: > On Thursday 01 June 2006 15:30, you wrote: >> On Thu, 1 Jun 2006, Peter Denno wrote: >>> Hi all, >>> >>> OK. The patches that Marco Gidde provided are now in CVS, and the "News" >>> section of the web page has been updated to mention this. >> >> Great! Thanks! >> >> [...] >> >>> So does anyone here have thoughts on how we can make things better? >> >> I'd like to clean up the code a little bit to reduce the number of >> compiler notes of the different lisps. The attached patch and a new >> file (gtk-ffi/package.lisp) is a first step in this direction. >> >> The patch is pretty uncontroversial (I think!) except for the >> >> - (cffi:load-foreign-library #+cffi-features:unix "libcellsgtk.so" >> + (cffi:load-foreign-library #+cffi-features:unix >> + (merge-pathnames "libcellsgtk.so" #.*compile-file-pathname*) >> >> part. It works for me in all lisps I tried, but still... Or are users >> supposed to install libcellsgtk.so in /usr/lib/ or some such? > > >> >> One more thing: since Cells has become Cells2, I think we should have >> a version of Cells in the repository at least until (if ever) we >> convert cells-gtk to Cells2. > > Good point. The tarball uses what Kenny Tilton was calling "cells 2," though > it might have changed since I last checked. I'll at least check in what I've > got. I'm rather confused about Cells versions. All I know for sure is that cells-gtk is incompatible with the latest CVS Cells (Cells 3?), so it'd be nice to provide a cells-gtk-friendly version of Cells, either in a separate tarball, or (better) in CVS. In the long run, it'd probably make sense to switch to Cells 3. > >> >> What do you think? > > These look fine. I'll check the merge-pathnames thing for libcellsgtk on > lispworks. I think the use of /usr/lib or whatever is leftover from the days > when we had to specify (or thought we had to specify) the library path. This > is better if it works. I'll check lispworks, at least. I tested it with LWL/SBCL/ACL, but more testing never hurts... > > It might be a few days before I get this in CVS. But I'll make a note so I > don't forget ;^) That's fine with me. In the meantime, should I be making new patches against the patched version, or the CVS version? Andras From marco.gidde at tiscali.de Sat Jun 3 12:21:00 2006 From: marco.gidde at tiscali.de (Marco Gidde) Date: Sat, 03 Jun 2006 14:21:00 +0200 Subject: [cells-gtk-devel] SBCL patches In-Reply-To: <200606011046.08716.peter.denno@nist.gov> (Peter Denno's message of "Thu, 1 Jun 2006 10:46:08 -0400") References: <85sln43k0u.fsf@tristan.br-automation.com> <200606011046.08716.peter.denno@nist.gov> Message-ID: <85wtbyxwer.fsf@tristan.br-automation.com> Hi Peter, Peter Denno writes: > OK. The patches that Marco Gidde provided are now in CVS, and the "News" > section of the web page has been updated to mention this. > > Thanks to Marco for the patches and Andras for letting me know that I missed > Marco's email! thanks for applying the patches. After a week of silence I was almost offended :-) > While I'm here, I'm thought I'd ask what people think is missing in cells-gtk. > For example, a while back I started some stuff to do drawing, and I intend to > someday pick that up again. I would use it to provide a graphing capability > (at least draw and manipulate trees). But I'm not in great need of the > capability now, and things just seem to work (I have a 3 cells-gtk apps now). > So does anyone here have thoughts on how we can make things better? What I miss most are much better drawing capabilities. Though I managed to draw a few lines on a widget using the mentioned functions I really did not get the ideas behind the API and switched to clg which already has cairo bindings. Of course clg has some other problems, e.g. no cells inside :-) and it is not using cffi. I think joining together clg and cells-gtk would create a killer GUI library for common lisp! Just my 0.02? Regards, Marco From peter.denno at nist.gov Tue Jun 6 15:07:22 2006 From: peter.denno at nist.gov (Peter Denno) Date: Tue, 6 Jun 2006 11:07:22 -0400 Subject: [cells-gtk-devel] SBCL patches In-Reply-To: <85wtbyxwer.fsf@tristan.br-automation.com> References: <85sln43k0u.fsf@tristan.br-automation.com> <200606011046.08716.peter.denno@nist.gov> <85wtbyxwer.fsf@tristan.br-automation.com> Message-ID: <200606061107.23046.peter.denno@nist.gov> On Saturday 03 June 2006 08:21, Marco Gidde wrote: Hi Marco, > > What I miss most are much better drawing capabilities. Though I > managed to draw a few lines on a widget using the mentioned functions > I really did not get the ideas behind the API and switched to clg > which already has cairo bindings. Of course clg has some other > problems, e.g. no cells inside :-) and it is not using cffi. Do you mean you "did not get the ideas behind the API" of the drawing functions, or cells-gtk in general? I think the biggest problem with cell-gtk is the lack of documentation of Cells. And for me, some of the design choices of Cells are not intuitive, making documenting it difficult. The drawing functions are a rough work-in-progress. The idea behind them hasn't really been expressed yet (I had in mind something with closures, but it needs lots of work). > > I think joining together clg and cells-gtk would create a killer GUI > library for common lisp! Just my 0.02? I took a quick look at clg and Cairo. Is what you like most about clg the Cairo port? clg looks ok. Cairo looks like the interesting part of it. I am not a "graphics person" but just someone who needed a GUI for lisp and stumbled into cells-gtk. It is good that I saw Cairo now, before I got too deep into drawing. I'll play around with it next time I get a chance to work on Cells-gtk. Thanks! > > > Regards, > > Marco -- - Best regards, Peter From marco.gidde at tiscali.de Wed Jun 7 13:39:45 2006 From: marco.gidde at tiscali.de (Marco Gidde) Date: Wed, 07 Jun 2006 15:39:45 +0200 Subject: [cells-gtk-devel] SBCL patches In-Reply-To: <200606061107.23046.peter.denno@nist.gov> (Peter Denno's message of "Tue, 6 Jun 2006 11:07:22 -0400") References: <85sln43k0u.fsf@tristan.br-automation.com> <200606011046.08716.peter.denno@nist.gov> <85wtbyxwer.fsf@tristan.br-automation.com> <200606061107.23046.peter.denno@nist.gov> Message-ID: <85u06xhyou.fsf@tristan.br-automation.com> Hi Peter, Peter Denno writes: > On Saturday 03 June 2006 08:21, Marco Gidde wrote: > >> What I miss most are much better drawing capabilities. Though I >> managed to draw a few lines on a widget using the mentioned functions >> I really did not get the ideas behind the API and switched to clg >> which already has cairo bindings. Of course clg has some other >> problems, e.g. no cells inside :-) and it is not using cffi. > > Do you mean you "did not get the ideas behind the API" of the drawing > functions, or cells-gtk in general? I meant the drawing functions. For instance it wasn't clear to me whether it was possible to draw directly on the widget or whether I had to create a pixmap which then should be copied to the widget. Also the meaning of the special variables (e.g. *window* and friends) and when (and why) they were set to some value wasn't clear. But as I wrote: I succeeded in drawing a line:-) But this is certainly not a serious critique since it is obvious that all the drawing stuff in cells-gtk is very raw and work in progress. > I think the biggest problem with cell-gtk is the lack of documentation of > Cells. And for me, some of the design choices of Cells are not intuitive, > making documenting it difficult. Not sure about the cells docu: though I only read the docu and most of the cells related discussions on c.l.l. but never used cells seriously before, it wasn't too hard to write a small test application, always keeping one eye on test-gtk. But the devil is in the details: e.g. when to use fm^ and when fm-other is still a miracle to me. > The drawing functions are a rough work-in-progress. The idea behind them > hasn't really been expressed yet (I had in mind something with closures, but > it needs lots of work). It was especially the amount of work that made me switch to clg. For the basic GUI stuff (eg. buttons, menus, list-views etc.) cells-gtk seems to be appropriate, maybe even better than cgl because of cells. > I took a quick look at clg and Cairo. Is what you like most about clg the > Cairo port? clg looks ok. Cairo looks like the interesting part of it. I am > not a "graphics person" but just someone who needed a GUI for lisp and > stumbled into cells-gtk. It is good that I saw Cairo now, before I got too > deep into drawing. I'll play around with it next time I get a chance to work > on Cells-gtk. It is not Cairo that attracts me, it's more clg's completeness and its nice integration with CLOS (in contrast to e.g. lambda-gtk, which is also quite complete, but using it is like programming in C with lots of parenthesis). While I have no problem with extending existing packages or fixing bugs as they occur, I don't want to spend most of my time doing this, in this case (GUI) especially because I never used GTK in C/C++ and so my knowledge is quite limited. If I had a choice, I would use Qt/KDE bindings, but these are essentially not existing and creating them seems to be really hard because of C++ and the C++ extensions used by Qt. If you think of continuing your work on the drawing functions, you should definitely take a very close look at Cairo: Gtk+-2.8 uses it internally for drawing the widgets! Regards, Marco From vinay at cs.unc.edu Sat Jun 17 23:29:43 2006 From: vinay at cs.unc.edu (Vinay Reddy) Date: Sat, 17 Jun 2006 16:29:43 -0700 Subject: [cells-gtk-devel] cannot load "load.lisp" Message-ID: <537f59d10606171629x65382205o305f1c3f8059a3fd@mail.gmail.com> Hi, I set the correct path in load.lisp, but I get the following error when I type in (load "load.lisp") -- ;; Loading file /home/vbondhug/lisp/cells-gtk/load.lisp ... 0 errors, 0 warnings ;; Compiling file /home/vbondhug/lisp/cells-gtk/cffi/src/cffi-clisp.lisp ... *** - READ from #: there is no package with name "FFI" The following restarts are available: RETRY :R1 Retry performing # on #. ACCEPT :R2 Continue, treating # on # as having been successful. SKIP :R3 skip (LET # # ...) STOP :R4 stop loading file /home/vbondhug/lisp/cells-gtk/load.lisp ABORT :R5 ABORT ABORT :R6 ABORT ABORT :R7 ABORT I'm new to Lisp, so don't really know what's going on... I'd appreciate any help... Thanks, Vinay From peter.denno at nist.gov Sun Jun 18 02:21:36 2006 From: peter.denno at nist.gov (Peter Denno) Date: Sat, 17 Jun 2006 22:21:36 -0400 Subject: [cells-gtk-devel] cannot load "load.lisp" In-Reply-To: <537f59d10606171629x65382205o305f1c3f8059a3fd@mail.gmail.com> References: <537f59d10606171629x65382205o305f1c3f8059a3fd@mail.gmail.com> Message-ID: <200606172221.37304.peter.denno@nist.gov> Hi, I just tried clisp on a linux machine, and didn't have a problem but that doesn't prove too much. It says it can't find the package FFI. FFI is the name of a clisp package. It seems to be missing from your lisp image. What do you get when you evaluate (use-package "FFI") ? I suppose it tells you again that it can't find FFI. Also, how about (lisp-implementation-version) CL-USER> (lisp-implementation-version) "2.36 (2005-12-04) (built 3342781448) (memory 3345206865)" Since you are new to lisp and maybe not yet settled on clisp, you might try sbcl. It has threading that make using cells-gtk easier. Either that or solve the problem with clisp. On Saturday 17 June 2006 19:29, Vinay Reddy wrote: > Hi, > I set the correct path in load.lisp, but I get the following error > when I type in (load "load.lisp") > > -- ;; Loading file /home/vbondhug/lisp/cells-gtk/load.lisp ... > 0 errors, 0 warnings > ;; Compiling file /home/vbondhug/lisp/cells-gtk/cffi/src/cffi-clisp.lisp > ... *** - READ from # #P"/home/vbondhug/lisp/cells-gtk/cffi/src/cffi-clisp.lisp" @84>: there > is no package with name "FFI" > The following restarts are available: > RETRY :R1 Retry performing # #x000333DBD1F8> on #. > ACCEPT :R2 Continue, treating # #x000333DBD1F8> on # > as having been successful. > SKIP :R3 skip (LET # # ...) > STOP :R4 stop loading file > /home/vbondhug/lisp/cells-gtk/load.lisp > ABORT :R5 ABORT > ABORT :R6 ABORT > ABORT :R7 ABORT > > I'm new to Lisp, so don't really know what's going on... > > I'd appreciate any help... > > Thanks, > Vinay > _______________________________________________ > 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 Sun Jun 18 22:19:33 2006 From: peter.denno at nist.gov (Peter Denno) Date: Sun, 18 Jun 2006 18:19:33 -0400 Subject: [cells-gtk-devel] cannot load "load.lisp" In-Reply-To: <537f59d10606181203q5a4a8ca6j5b47e05ccc7913e6@mail.gmail.com> References: <537f59d10606171629x65382205o305f1c3f8059a3fd@mail.gmail.com> <200606172221.37304.peter.denno@nist.gov> <537f59d10606181203q5a4a8ca6j5b47e05ccc7913e6@mail.gmail.com> Message-ID: <200606181819.34243.peter.denno@nist.gov> Hi, I just tried compiling cells-gtk-2006-02-16 under sbcl-0.9.12-x86-64-linux-binary.tar.bz2 and it compiles fine for me. There is one place where you have to continue from a variable being redefined. We fixed this in the CVS, but for now, just using the tarball, you will get this: The constant +C-NULL+ is being redefined (from #.(SB-SYS:INT-SAP #X00000000) to #.(SB-SYS:INT-SAP #X00000000)) See also: The ANSI Standard, Macro DEFCONSTANT The SBCL Manual, Node "Idiosyncrasies" Type HELP for debugger help, or (SB-EXT:QUIT) to exit from SBCL. restarts (invokable by number or by possibly-abbreviated name): 0: [CONTINUE] Go ahead and change the value. ...At which point you type 0 and it continues (lots of text printing out, which is inevitable in sbcl) but after all that you should see: Done! Now try (test-gtk::gtk-demo) T * I can't try running cells-gtk on x64 because the server I tried this on does not run X11. But that isn't where you are getting stuck anyway. The error sbcl is showing you is the same sort of thing that you saw with clisp. A package is not found. Again I don't understand this. I will show you what I did to install sbcl and to compile cells-gtk: (0) mkdir ~/lisp (1) cd ~/lisp (2) tar xjf sbcl-0.9.12-x86-64-linux-binary.tar.bz2 (3) cd sbcl-0.9.12-x86-64-linux/ (4) INSTALL_ROOT=/home/pdenno sh install.sh (5) cd .. (6) tar zxf cells-gtk-2006-02-16.tgz (7) cd cells-gtk-2006-02-16/ (8) sbcl --core ~/lisp/sbcl-0.9.12-x86-64-linux/output/sbcl.core (89 (load "load") (10) enter 0 as described above. (11) (test-gtk::gtk-demo) On Sunday 18 June 2006 15:03, you wrote: > I tried sbcl, but that failed too... Here's the error when I do a > (load "load") with sbcl: > > ; compiling file > "/exit14/home/vbondhug/temp/cells-gtk/root/gtk-ffi/gtk-ffi.lisp" > (written 03 JAN 2006 12:02:10 PM): > ; compiling (DEFPACKAGE :GTK-FFI ...) > debugger invoked on a SB-KERNEL:SIMPLE-PACKAGE-ERROR in thread > #: > The name "LISP" does not designate any package. > > Type HELP for debugger help, or (SB-EXT:QUIT) to exit from SBCL. > > restarts (invokable by number or by possibly-abbreviated name): > 0: [RETRY ] Retry performing # on > #. > 1: [ACCEPT] Continue, treating # on > # as having been > successful. > 2: [ABORT ] Exit debugger, returning to top level. > > (SB-INT:%FIND-PACKAGE-OR-LOSE "LISP") > > I downloaded the latest sbcl binaries for amd 64 from sf.net. Also, I > went into the gtk-ffi directory and did a make, which completed > without errors... > > Vinay -- - Best regards, Peter From vinay at cs.unc.edu Sun Jun 18 18:22:11 2006 From: vinay at cs.unc.edu (Vinay Reddy) Date: Sun, 18 Jun 2006 11:22:11 -0700 Subject: [cells-gtk-devel] cannot load "load.lisp" In-Reply-To: <200606172221.37304.peter.denno@nist.gov> References: <537f59d10606171629x65382205o305f1c3f8059a3fd@mail.gmail.com> <200606172221.37304.peter.denno@nist.gov> Message-ID: <537f59d10606181122y66d14012v1f2bb14639c67db0@mail.gmail.com> > Hi, > > I just tried clisp on a linux machine, and didn't have a problem but that > doesn't prove too much. It says it can't find the package FFI. FFI is the > name of a clisp package. It seems to be missing from your lisp image. What > do you get when you evaluate (use-package "FFI") ? I suppose it tells you > again that it can't find FFI. Where do I get the FFI package? I googled but didn't find anything proper... I tried cells-gtk-2006-02-16.tgz and cells-gtk-2006-01-03.tgz, both of which give the FFI package error. Does FFI have to be added in separately? > Also, how about (lisp-implementation-version) > > CL-USER> (lisp-implementation-version) > "2.36 (2005-12-04) (built 3342781448) (memory 3345206865)" [1]> (lisp-implementation-version) "2.38 (2006-01-24) (built 3359570456) (memory 3359570552)" > Since you are new to lisp and maybe not yet settled on clisp, you might try > sbcl. It has threading that make using cells-gtk easier. Either that or solve > the problem with clisp. I'll try that, but I'd like to solve the problem with clisp. I won't be comfortable otherwise :) Cheers, Vinay PS: This seems to be a development mailing list, which I think is not the right place for questions about installation... Is there another mailing list appropriate for such issues? From asimon at math.bme.hu Mon Jun 19 01:50:02 2006 From: asimon at math.bme.hu (Andras Simon) Date: Mon, 19 Jun 2006 03:50:02 +0200 (CEST) Subject: [cells-gtk-devel] cannot load "load.lisp" In-Reply-To: <537f59d10606181122y66d14012v1f2bb14639c67db0@mail.gmail.com> References: <537f59d10606171629x65382205o305f1c3f8059a3fd@mail.gmail.com> <200606172221.37304.peter.denno@nist.gov> <537f59d10606181122y66d14012v1f2bb14639c67db0@mail.gmail.com> Message-ID: On Sun, 18 Jun 2006, Vinay Reddy wrote: >> Hi, >> >> I just tried clisp on a linux machine, and didn't have a problem but that >> doesn't prove too much. It says it can't find the package FFI. FFI is the >> name of a clisp package. It seems to be missing from your lisp image. >> What >> do you get when you evaluate (use-package "FFI") ? I suppose it tells you >> again that it can't find FFI. > > Where do I get the FFI package? I googled but didn't find anything proper... FFI is the Foreign Function Interface in CLISP. Depending on your platform, either you have it in your CLISP image, or you can't have it at all... See http://clisp.cons.org/impnotes.html#dffi But: you didn't answer Peter's question. So what does your CLISP say for (use-package "FFI") (or (find-package "FFI"))? > I tried cells-gtk-2006-02-16.tgz and cells-gtk-2006-01-03.tgz, both of > which give the FFI package error. Does FFI have to be added in > separately? No. Is the cffi package there in the tarball you downloaded? Can you try to run its test-suite? Andras From ray.subhasis at gmail.com Mon Jun 19 06:28:08 2006 From: ray.subhasis at gmail.com (Subhasis Ray) Date: Mon, 19 Jun 2006 11:58:08 +0530 Subject: [cells-gtk-devel] cannot load "load.lisp" In-Reply-To: References: <537f59d10606171629x65382205o305f1c3f8059a3fd@mail.gmail.com> <200606172221.37304.peter.denno@nist.gov> <537f59d10606181122y66d14012v1f2bb14639c67db0@mail.gmail.com> Message-ID: Hi, I had successfully compile and run Cells-GTK upto the verion of *2006-01-03 *on CLisp versions 2.36, 2.37 and 2.38 on both Windows XP and Fedora Core -1. I do not think it is a problem with cells-gtk, all the problems I had were due to mistakes in path settings. So my suggesstion is to double check that you do not miss out the paths anywhere. I shall shortly try out the *2006-02-16 version. Cheers. Su* -------------- next part -------------- An HTML attachment was scrubbed... URL: From ray.subhasis at gmail.com Mon Jun 19 09:36:09 2006 From: ray.subhasis at gmail.com (Subhasis Ray) Date: Mon, 19 Jun 2006 15:06:09 +0530 Subject: [cells-gtk-devel] cannot load "load.lisp" In-Reply-To: References: <537f59d10606171629x65382205o305f1c3f8059a3fd@mail.gmail.com> <200606172221.37304.peter.denno@nist.gov> <537f59d10606181122y66d14012v1f2bb14639c67db0@mail.gmail.com> Message-ID: Hi, Though it may seem redundant/ridiculous, but for new-comers very simple problems pose big hurdle to getting started (I am a newbie myself). Here are my steps for running cells gtk ,it runs fine when "load.lisp" is loaded from a clean cells-gtk. But as I tried to load it a second time I face the problem mentioned by Lelanthran: Download and install gtk - in windows it is installed in C:\Program Files\Common Files\GTK. Download ftp://common-lisp.net/pub/project/cells-gtk/cells-gtk-2006-02-16.tgz. Download CLisp 2.38 for win32 from http://prdownloads.sourceforge.net/clisp/clisp-2.38-win32-with-readline-and-gettext.zip?download Extract clisp-2.38-win32-with-readline-and-gettext.zip in "C:\", so the lisp installation directory becomes C:\clisp-2.38. Now run the install.batscript, enter 'y' to all questions. Extract cells-gtk-2006-02-16.tgz wherever you like. In my case it was "C:\subhasir\tmp". Copy root\gtk-ffi\libcellsgtk.dll from cells-gtk directory (i.e. C:\subhasir\tmp\cells-gtk-2006-02-16\root\gtk-ffi\libcellsgtk.dll) to the bin directory of your GTK installation (c:\Program Files\Common Files\GTK\2.0\bin for an installation with defaults). Edit the file load.lisp in the directory where cells-gtk was extracted (C:\subhasir\tmp\cells-gtk-2006-02-16\ in my case). Modify Line#12 in the file and place the path-components to the cells-gtk directory. I have had problem with drive-letter - if "C:" is added at the beginning clisp will complain about the path. This is what I put there: #+clisp (make-pathname :directory '(:absolute "subhasir" "tmp" "cells-gtk-2006-02-16")))) ; <=== CLISP users Now start clisp. Go to the cells-gtk directory by entering '(cd "C:/subhasir/tmp/cells-gtk-2006-02-16")' in clisp console - note the front-slash as path-separator. [ You don't have to cd to cells-gtk directory if you have asdf.lisp in your path ]. Enter '(load "load.lisp")' - and everything should compile fine. The last output should be: 0 errors, 0 warnings Done! Now try (test-gtk::gtk-demo) ;; Loaded file C:\subhasir\tmp\cells-gtk-2006-02-16\load.lisp T Now enter (test-gtk::gtk-demo) in clisp console and you should see the cells-gtk splash screen appear. Now if I restart clisp and come-back to reload the cells-gtk, this error comes: ;; Loading file C:\subhasir\tmp\cells-gtk-2006-02-16\root\gtk-ffi\gtk- ffi.fas . .. ;; Loaded file C:\subhasir\tmp\cells-gtk-2006-02-16\root\gtk-ffi\gtk- ffi.fas ;; Loading file C:\subhasir\tmp\cells-gtk-2006-02-16\root\gtk-ffi\gtk- core.fas ... ** - Continuable Error FFI::FOREIGN-LIBRARY-FUNCTION: no dynamic object named "g_free" in library :DEFAULT If you continue (by typing 'continue'): Skip foreign object creation The following restarts are also available: RETRY :R1 Retry performing # on #. ACCEPT :R2 Continue, treating # on # as having been successful. SKIP :R3 skip (LET # # ...) STOP :R4 stop loading file C:\subhasir\tmp\cells-gtk-2006-02-16\l oad.lisp ABORT :R5 ABORT Thanks and regards, Su -------------- next part -------------- An HTML attachment was scrubbed... URL: From vinay at cs.unc.edu Mon Jun 19 16:42:50 2006 From: vinay at cs.unc.edu (Vinay Reddy) Date: Mon, 19 Jun 2006 09:42:50 -0700 Subject: [cells-gtk-devel] cannot load "load.lisp" In-Reply-To: <200606181819.34243.peter.denno@nist.gov> References: <537f59d10606171629x65382205o305f1c3f8059a3fd@mail.gmail.com> <200606172221.37304.peter.denno@nist.gov> <537f59d10606181203q5a4a8ca6j5b47e05ccc7913e6@mail.gmail.com> <200606181819.34243.peter.denno@nist.gov> Message-ID: <537f59d10606190942r5752dd80i6fe4c0270197aaf3@mail.gmail.com> > I just tried compiling cells-gtk-2006-02-16 under > sbcl-0.9.12-x86-64-linux-binary.tar.bz2 and it compiles fine for me. There is > one place where you have to continue from a variable being redefined. We > fixed this in the CVS, but for now, just using the tarball, you will get > this: > > The constant +C-NULL+ is being redefined (from #.(SB-SYS:INT-SAP #X00000000) > to #.(SB-SYS:INT-SAP #X00000000)) > See also: > The ANSI Standard, Macro DEFCONSTANT > The SBCL Manual, Node "Idiosyncrasies" > > Type HELP for debugger help, or (SB-EXT:QUIT) to exit from SBCL. > > restarts (invokable by number or by possibly-abbreviated name): > 0: [CONTINUE] Go ahead and change the value. > > ...At which point you type 0 and it continues (lots of text printing out, > which is inevitable in sbcl) but after all that you should see: > > Done! Now try (test-gtk::gtk-demo) > T > * > I can't try running cells-gtk on x64 because the server I tried this on does > not run X11. But that isn't where you are getting stuck anyway. > > The error sbcl is showing you is the same sort of thing that you saw with > clisp. A package is not found. Again I don't understand this. I will show you > what I did to install sbcl and to compile cells-gtk: > > (0) mkdir ~/lisp > (1) cd ~/lisp > (2) tar xjf sbcl-0.9.12-x86-64-linux-binary.tar.bz2 > (3) cd sbcl-0.9.12-x86-64-linux/ > (4) INSTALL_ROOT=/home/pdenno sh install.sh > (5) cd .. > (6) tar zxf cells-gtk-2006-02-16.tgz > (7) cd cells-gtk-2006-02-16/ > (8) sbcl --core ~/lisp/sbcl-0.9.12-x86-64-linux/output/sbcl.core > (89 (load "load") > (10) enter 0 as described above. > (11) (test-gtk::gtk-demo) I tried the exact same steps. This time, the whole thing went through fine, but when I do (test-gtk::gtk-demo) it complains about libcellsgtk.so. (test-gtk::gtk-demo) debugger invoked on a CFFI:LOAD-FOREIGN-LIBRARY-ERROR in thread #: Unable to load foreign library: libcellsgtk.so Here's the error message: Type HELP for debugger help, or (SB-EXT:QUIT) to exit from SBCL. restarts (invokable by number or by possibly-abbreviated name): 0: [RETRY ] Try loading the foreign library again. 1: [USE-VALUE] Use another library instead. 2: [ABORT ] Exit debugger, returning to top level. (CFFI::HANDLE-LOAD-FOREIGN-LIBRARY-ERROR "libcellsgtk.so" "Unable to load foreign library: ~A" "libcellsgtk.so") I build libcellsgtk.so in root/gtk-ffi/ and placed it in /usr/lib, but sbcl couldn't find it. Even though cells-gtk loaded fine in sbcl, when I do a (use-package "FFI"), it still can't find FFI. (find-package "FFI") returns NIL. Thanks, Vinay PS: I'm using RHEL 4 on AMD 64. From asimon at math.bme.hu Mon Jun 19 17:56:09 2006 From: asimon at math.bme.hu (Andras Simon) Date: Mon, 19 Jun 2006 19:56:09 +0200 (CEST) Subject: [cells-gtk-devel] cannot load "load.lisp" In-Reply-To: <537f59d10606190951j2da2470fsb8824341ea0bd529@mail.gmail.com> References: <537f59d10606171629x65382205o305f1c3f8059a3fd@mail.gmail.com> <200606172221.37304.peter.denno@nist.gov> <537f59d10606181122y66d14012v1f2bb14639c67db0@mail.gmail.com> <537f59d10606190951j2da2470fsb8824341ea0bd529@mail.gmail.com> Message-ID: On Mon, 19 Jun 2006, Vinay Reddy wrote: >> Is the cffi package there in the tarball you downloaded? Can you try >> to run its test-suite? > > Yes, cffi is in the tarball I downloaded. Sadly, I'm unable to run the > test-suite. It couldn't find the package "ASDF". I'm sure it's a > problem with paths, but I really don't know how/where to fix it :( asdf.lisp must be in that tarball, too. Load that, and then load all the files with .asd extension in the cffi directory (this is not the usual procedure, but never mind), and say (asdf:oos 'asdf:load-op :cffi) If that works, try (asdf:oos 'asdf:load-op :cffi-tests) But wait! For this, you may need the rt (regression test) package, it's here: http://files.b9.com/rt/rt.tar.gz It's got its own .asd file, load that, too. > I seem to be the only one with the FFI problem... I compiled my clisp > from source (version 2.38). I even tried using --with-dynamic-ffi with > the configure script, but that didn't help. You still didn't tell us what CLISP says to (find-package "FFI"). In your other mail, you wrote that SBCL says NIL; but that wasn't the question, since it was CLISP, not SBCL that gave you `there is no package with name "FFI"'; SBCL's foreign function interface package is called SB-ALIEN. Andras From vinay at cs.unc.edu Mon Jun 19 18:03:19 2006 From: vinay at cs.unc.edu (Vinay Reddy) Date: Mon, 19 Jun 2006 11:03:19 -0700 Subject: [cells-gtk-devel] cannot load "load.lisp" In-Reply-To: References: <537f59d10606171629x65382205o305f1c3f8059a3fd@mail.gmail.com> <200606172221.37304.peter.denno@nist.gov> <537f59d10606181122y66d14012v1f2bb14639c67db0@mail.gmail.com> <537f59d10606190951j2da2470fsb8824341ea0bd529@mail.gmail.com> Message-ID: <537f59d10606191103t4093c40bod0746afa87cbaa7b@mail.gmail.com> > asdf.lisp must be in that tarball, too. Load that, and then load all > the files with .asd extension in the cffi directory (this is not the > usual procedure, but never mind), and say > > (asdf:oos 'asdf:load-op :cffi) > > If that works, try > > (asdf:oos 'asdf:load-op :cffi-tests) > > But wait! For this, you may need the rt (regression test) package, > it's here: http://files.b9.com/rt/rt.tar.gz It's got its own .asd > file, load that, too. > > > I seem to be the only one with the FFI problem... I compiled my clisp > > from source (version 2.38). I even tried using --with-dynamic-ffi with > > the configure script, but that didn't help. > > You still didn't tell us what CLISP says to (find-package "FFI"). Sorry about that. CLISP also returns NIL to that. With sbcl I was able to do a (load "load") successfully, but on running (test-gtk::gtk-demo), I get a memory fault error. debugger invoked on a SB-KERNEL::MEMORY-FAULT-ERROR in thread #: memory fault Type HELP for debugger help, or (SB-EXT:QUIT) to exit from SBCL. restarts (invokable by number or by possibly-abbreviated name): 0: [ABORT] Exit debugger, returning to top level. (SB-KERNEL::MEMORY-FAULT-ERROR) I'll try the clisp tests... Thanks, Vinay From asimon at math.bme.hu Mon Jun 19 19:19:27 2006 From: asimon at math.bme.hu (Andras Simon) Date: Mon, 19 Jun 2006 21:19:27 +0200 (CEST) Subject: [cells-gtk-devel] cannot load "load.lisp" In-Reply-To: <537f59d10606191103t4093c40bod0746afa87cbaa7b@mail.gmail.com> References: <537f59d10606171629x65382205o305f1c3f8059a3fd@mail.gmail.com> <200606172221.37304.peter.denno@nist.gov> <537f59d10606181122y66d14012v1f2bb14639c67db0@mail.gmail.com> <537f59d10606190951j2da2470fsb8824341ea0bd529@mail.gmail.com> <537f59d10606191103t4093c40bod0746afa87cbaa7b@mail.gmail.com> Message-ID: On Mon, 19 Jun 2006, Vinay Reddy wrote: >> asdf.lisp must be in that tarball, too. Load that, and then load all >> the files with .asd extension in the cffi directory (this is not the >> usual procedure, but never mind), and say >> >> (asdf:oos 'asdf:load-op :cffi) >> >> If that works, try >> >> (asdf:oos 'asdf:load-op :cffi-tests) >> >> But wait! For this, you may need the rt (regression test) package, >> it's here: http://files.b9.com/rt/rt.tar.gz It's got its own .asd >> file, load that, too. >> >> > I seem to be the only one with the FFI problem... I compiled my clisp >> > from source (version 2.38). I even tried using --with-dynamic-ffi with >> > the configure script, but that didn't help. >> >> You still didn't tell us what CLISP says to (find-package "FFI"). > > Sorry about that. CLISP also returns NIL to that. Hmmm, what's your OS? The CLISP doc says that FFI is "Platform Dependent: Many UNIX, Win32 platforms only." Anyway, if there's no FFI, don't bother with testing cffi; it won't work. > > With sbcl I was able to do a (load "load") successfully, but on So you got around the foreign library loading problem. Great! > running (test-gtk::gtk-demo), I get a memory fault error. > > debugger invoked on a SB-KERNEL::MEMORY-FAULT-ERROR in thread > #: > memory fault > > Type HELP for debugger help, or (SB-EXT:QUIT) to exit from SBCL. > > restarts (invokable by number or by possibly-abbreviated name): > 0: [ABORT] Exit debugger, returning to top level. > > (SB-KERNEL::MEMORY-FAULT-ERROR) I have no idea why this happens. Did you compile your SBCL? Can you run its self-tests? (You need the SBCL sources for this.) Andras > I'll try the clisp tests... > > Thanks, > Vinay > From vinay at cs.unc.edu Mon Jun 19 16:51:13 2006 From: vinay at cs.unc.edu (Vinay Reddy) Date: Mon, 19 Jun 2006 09:51:13 -0700 Subject: [cells-gtk-devel] cannot load "load.lisp" In-Reply-To: References: <537f59d10606171629x65382205o305f1c3f8059a3fd@mail.gmail.com> <200606172221.37304.peter.denno@nist.gov> <537f59d10606181122y66d14012v1f2bb14639c67db0@mail.gmail.com> Message-ID: <537f59d10606190951j2da2470fsb8824341ea0bd529@mail.gmail.com> > Is the cffi package there in the tarball you downloaded? Can you try > to run its test-suite? Yes, cffi is in the tarball I downloaded. Sadly, I'm unable to run the test-suite. It couldn't find the package "ASDF". I'm sure it's a problem with paths, but I really don't know how/where to fix it :( I seem to be the only one with the FFI problem... I compiled my clisp from source (version 2.38). I even tried using --with-dynamic-ffi with the configure script, but that didn't help. Thanks, Vinay From vinay at cs.unc.edu Mon Jun 19 19:55:45 2006 From: vinay at cs.unc.edu (Vinay Reddy) Date: Mon, 19 Jun 2006 12:55:45 -0700 Subject: [cells-gtk-devel] cannot load "load.lisp" In-Reply-To: References: <537f59d10606171629x65382205o305f1c3f8059a3fd@mail.gmail.com> <200606172221.37304.peter.denno@nist.gov> <537f59d10606181122y66d14012v1f2bb14639c67db0@mail.gmail.com> <537f59d10606190951j2da2470fsb8824341ea0bd529@mail.gmail.com> <537f59d10606191103t4093c40bod0746afa87cbaa7b@mail.gmail.com> Message-ID: <537f59d10606191255p5fa014a2h4d8e2059681fca8@mail.gmail.com> > Hmmm, what's your OS? The CLISP doc says that FFI is "Platform > Dependent: Many UNIX, Win32 platforms only." Anyway, if there's no > FFI, don't bother with testing cffi; it won't work. I'm using RHEL 4. > I have no idea why this happens. Did you compile your SBCL? Can you > run its self-tests? (You need the SBCL sources for this.) Nope, I downloaded sbcl binaries for AMD 64 from sf.net. okay, I'll try compiling sbcl from sources. Thanks, Vinay From asimon at math.bme.hu Mon Jun 19 21:45:08 2006 From: asimon at math.bme.hu (Andras Simon) Date: Mon, 19 Jun 2006 23:45:08 +0200 (CEST) Subject: [cells-gtk-devel] cannot load "load.lisp" In-Reply-To: <537f59d10606191255p5fa014a2h4d8e2059681fca8@mail.gmail.com> References: <537f59d10606171629x65382205o305f1c3f8059a3fd@mail.gmail.com> <200606172221.37304.peter.denno@nist.gov> <537f59d10606181122y66d14012v1f2bb14639c67db0@mail.gmail.com> <537f59d10606190951j2da2470fsb8824341ea0bd529@mail.gmail.com> <537f59d10606191103t4093c40bod0746afa87cbaa7b@mail.gmail.com> <537f59d10606191255p5fa014a2h4d8e2059681fca8@mail.gmail.com> Message-ID: On Mon, 19 Jun 2006, Vinay Reddy wrote: >> Hmmm, what's your OS? The CLISP doc says that FFI is "Platform >> Dependent: Many UNIX, Win32 platforms only." Anyway, if there's no >> FFI, don't bother with testing cffi; it won't work. > > I'm using RHEL 4. In that case I'm very much surprised that your CLISP comes without FFI. Perhaps you could ask about this on http://news.gmane.org/gmane.lisp.clisp.general > >> I have no idea why this happens. Did you compile your SBCL? Can you >> run its self-tests? (You need the SBCL sources for this.) > > Nope, I downloaded sbcl binaries for AMD 64 from sf.net. > > okay, I'll try compiling sbcl from sources. Again, another thing to try would be to ask on the sbcl-help list. I'm sorry I can't be of more help... Andras From ray.subhasis at gmail.com Tue Jun 20 05:22:08 2006 From: ray.subhasis at gmail.com (Subhasis Ray) Date: Tue, 20 Jun 2006 10:52:08 +0530 Subject: [cells-gtk-devel] cannot load "load.lisp" In-Reply-To: <537f59d10606190942r5752dd80i6fe4c0270197aaf3@mail.gmail.com> References: <537f59d10606171629x65382205o305f1c3f8059a3fd@mail.gmail.com> <200606172221.37304.peter.denno@nist.gov> <537f59d10606181203q5a4a8ca6j5b47e05ccc7913e6@mail.gmail.com> <200606181819.34243.peter.denno@nist.gov> <537f59d10606190942r5752dd80i6fe4c0270197aaf3@mail.gmail.com> Message-ID: > > > I build libcellsgtk.so in root/gtk-ffi/ and placed it in /usr/lib, but > sbcl couldn't find it. > I think gtk installs in /usr/lib/gtk .. though I don't have a RHEL installation right now - you can search for the other gtk-library files (like libatk-.so, libpango-.so , etc.) and place your libcellsgtk.so there. hth Su -------------- next part -------------- An HTML attachment was scrubbed... URL: From asimon at math.bme.hu Tue Jun 20 09:43:03 2006 From: asimon at math.bme.hu (Andras Simon) Date: Tue, 20 Jun 2006 11:43:03 +0200 (CEST) Subject: [cells-gtk-devel] trivial patches Message-ID: Hi Peter, Would you commit these? Thanks, Andras -------------- next part -------------- Index: cells-gtk/drawing.lisp =================================================================== RCS file: /project/cells-gtk/cvsroot/root/cells-gtk/drawing.lisp,v retrieving revision 1.7 diff -u -a -r1.7 drawing.lisp --- cells-gtk/drawing.lisp 7 Jun 2006 16:40:06 -0000 1.7 +++ cells-gtk/drawing.lisp 19 Jun 2006 23:15:41 -0000 @@ -69,7 +69,6 @@ (gtk-signal-connect-swap widget "expose-event" (cffi:get-callback 'drawing-expose-event-handler) :data data)) -(export '(gtk-drawing-set-handlers)) ;;;============================================================================ (in-package :cgtk) Index: cells-gtk/packages.lisp =================================================================== RCS file: /project/cells-gtk/cvsroot/root/cells-gtk/packages.lisp,v retrieving revision 1.6 diff -u -a -r1.6 packages.lisp --- cells-gtk/packages.lisp 7 Jun 2006 16:43:08 -0000 1.6 +++ cells-gtk/packages.lisp 19 Jun 2006 23:15:41 -0000 @@ -32,7 +32,6 @@ #:push-message #:pop-message #:pulse - #:gtk-drawing-set-handlers #:*gcontext* #:with-pixmap #:with-gc Index: gtk-ffi/package.lisp =================================================================== RCS file: /project/cells-gtk/cvsroot/root/gtk-ffi/package.lisp,v retrieving revision 1.1 diff -u -a -r1.1 package.lisp --- gtk-ffi/package.lisp 8 Jun 2006 14:58:24 -0000 1.1 +++ gtk-ffi/package.lisp 19 Jun 2006 23:15:42 -0000 @@ -64,5 +64,6 @@ #:gtk-tree-store-set-kids #:gtk-tree-model-get-cell #:gtk-tree-view-render-cell - #:gtk-file-chooser-get-filenames-strs)) + #:gtk-file-chooser-get-filenames-strs + #:gtk-drawing-set-handlers)) From peter.denno at nist.gov Tue Jun 20 12:34:59 2006 From: peter.denno at nist.gov (Peter Denno) Date: Tue, 20 Jun 2006 08:34:59 -0400 Subject: [cells-gtk-devel] trivial patches In-Reply-To: References: Message-ID: <200606200834.59797.peter.denno@nist.gov> On Tuesday 20 June 2006 05:43, Andras Simon wrote: > Hi Peter, > > Would you commit these? Thanks, Hi Andras, Sure, I will if you'd like. But didn't you get CVS access? > > Andras -- - Best regards, Peter From asimon at math.bme.hu Tue Jun 20 15:04:24 2006 From: asimon at math.bme.hu (Andras Simon) Date: Tue, 20 Jun 2006 17:04:24 +0200 (CEST) Subject: [cells-gtk-devel] trivial patches In-Reply-To: <200606200834.59797.peter.denno@nist.gov> References: <200606200834.59797.peter.denno@nist.gov> Message-ID: On Tue, 20 Jun 2006, Peter Denno wrote: > On Tuesday 20 June 2006 05:43, Andras Simon wrote: >> Hi Peter, >> >> Would you commit these? Thanks, > > Hi Andras, > > Sure, I will if you'd like. But didn't you get CVS access? Looks like I did, it's just that I wasn't aware of it! Thanks! Andras From peter.denno at nist.gov Tue Jun 20 18:11:10 2006 From: peter.denno at nist.gov (Peter Denno) Date: Tue, 20 Jun 2006 14:11:10 -0400 Subject: [cells-gtk-devel] cannot load "load.lisp" In-Reply-To: <537f59d10606191255p5fa014a2h4d8e2059681fca8@mail.gmail.com> References: <537f59d10606171629x65382205o305f1c3f8059a3fd@mail.gmail.com> <537f59d10606191255p5fa014a2h4d8e2059681fca8@mail.gmail.com> Message-ID: <200606201411.10773.peter.denno@nist.gov> On Monday 19 June 2006 15:55, Vinay Reddy wrote: > > Hmmm, what's your OS? The CLISP doc says that FFI is "Platform > > Dependent: Many UNIX, Win32 platforms only." Anyway, if there's no > > FFI, don't bother with testing cffi; it won't work. > > I'm using RHEL 4. > > > I have no idea why this happens. Did you compile your SBCL? Can you > > run its self-tests? (You need the SBCL sources for this.) > > Nope, I downloaded sbcl binaries for AMD 64 from sf.net. > > okay, I'll try compiling sbcl from sources. I would guess that the chances of success along this path, given that you are an inexperience user, would not be very good. -- - Best regards, Peter From asimon at math.bme.hu Tue Jun 20 18:36:15 2006 From: asimon at math.bme.hu (Andras Simon) Date: Tue, 20 Jun 2006 20:36:15 +0200 (CEST) Subject: [cells-gtk-devel] cannot load "load.lisp" In-Reply-To: <200606201411.10773.peter.denno@nist.gov> References: <537f59d10606171629x65382205o305f1c3f8059a3fd@mail.gmail.com> <537f59d10606191255p5fa014a2h4d8e2059681fca8@mail.gmail.com> <200606201411.10773.peter.denno@nist.gov> Message-ID: On Tue, 20 Jun 2006, Peter Denno wrote: > On Monday 19 June 2006 15:55, Vinay Reddy wrote: >>> Hmmm, what's your OS? The CLISP doc says that FFI is "Platform >>> Dependent: Many UNIX, Win32 platforms only." Anyway, if there's no >>> FFI, don't bother with testing cffi; it won't work. >> >> I'm using RHEL 4. >> >>> I have no idea why this happens. Did you compile your SBCL? Can you >>> run its self-tests? (You need the SBCL sources for this.) >> >> Nope, I downloaded sbcl binaries for AMD 64 from sf.net. >> >> okay, I'll try compiling sbcl from sources. > > I would guess that the chances of success along this path, given that you are > an inexperience user, would not be very good. Right. Perhaps starting with Allegro (www.franz.com) or LispWorks (www.lispworks.com) would be easier. I was worried that the heap size restriction in ACL 8 Express might be a problem, but fortunately no. Andras From shaun.kruger at gmail.com Wed Jun 21 23:19:35 2006 From: shaun.kruger at gmail.com (Shaun Kruger) Date: Wed, 21 Jun 2006 17:19:35 -0600 Subject: [cells-gtk-devel] CLisp win32 FFI related errors Message-ID: I've been doing more reading and found a manual that describes uffi:def-function syntax. In my google searches I have found that (uffi:def-function) sometimes has a :module "libsomething" value set. The error I get from starting after doing (saveinitmem) and opening lisp -M is this: [4]> (test-gtk::gtk-demo) ** - Continuable Error FFI::FOREIGN-CALL-OUT: no dynamic object named "g_signal_connect_data" in library :DEFAULT If you continue (by typing 'continue'): Skip foreign object creation The following restarts are also available: ABORT :R1 ABORT I found the definition of (uffi:def-function) for g_signal_connect_data in gtk-ffi\gtk-utilities.lisp and it is missing a :module value. Would adding a :module value be a reasonable fix? How does this fit in with the (cffi:define-foreign-library) declarations in cgk-fii\gtk-ffi.lisp? Is there any reason why the one for gobject is 'gobject instead of :gobject? Changing it didn't make any difference to my problem... If I make a call to (cells-gtk::load-gtk-libs) I get a # return, but then it fails with the following: [8]> (test-gtk::gtk-demo) *** - FFI::FOREIGN-CALL-OUT: # comes from a previous Lisp session and is invalid The following restarts are available: ABORT :R1 ABORT I'm trying everything I know how to figure out. What else should I be looking at? Once again, thoughts and suggestions would be a real help. Shaun Kruger -- Visit my blog at http://hackerlog.blogspot.com ===================================================== If more of us valued food and cheer and song above hoarded gold, it would be a merrier world. -- J.R.R. Tolkien From shaun.kruger at gmail.com Wed Jun 21 21:28:50 2006 From: shaun.kruger at gmail.com (Shaun Kruger) Date: Wed, 21 Jun 2006 15:28:50 -0600 Subject: [cells-gtk-devel] cannot load "load.lisp" In-Reply-To: References: <537f59d10606171629x65382205o305f1c3f8059a3fd@mail.gmail.com> <200606172221.37304.peter.denno@nist.gov> <537f59d10606181122y66d14012v1f2bb14639c67db0@mail.gmail.com> Message-ID: On 6/19/06, Subhasis Ray wrote: > Hi, > Though it may seem redundant/ridiculous, but for new-comers very simple > problems pose big hurdle to getting started (I am a newbie myself). Here are > my steps for running cells gtk ,it runs fine when "load.lisp " is loaded > from a clean cells-gtk. But as I tried to load it a second time I face the > problem mentioned by Lelanthran: > I am having this same problem with the same version levels and process described. I am wondering if it has something to do with it saying "in library :DEFAULT" instead of listing the GTK library by path. I really don't know enough about FFI to debug a problem like this. Would it help if the gtk libraries were in the same folder as clisp? That seems to take care of wierdness with the PostgreSQL FFI stuff. For the short term however, I have determined that deleting all the .fas files in cells-gtk\root\gtk-ffi and then running (load "load") will work. That doesn't do much for me as I want to generate a new lispinit.mem file and have cells-gtk available at startup. Suggestions on where to start debugging this issue myself would be very helpful. Where do I find the FFI documentation? What are some common warnings when working with FFI? Shaun Kruger > > Now if I restart clisp and come-back to reload the cells-gtk, this error > comes: > ;; Loading file > C:\subhasir\tmp\cells-gtk-2006-02-16\root\gtk-ffi\gtk-ffi.fas > . > .. > ;; Loaded file > C:\subhasir\tmp\cells-gtk-2006-02-16\root\gtk-ffi\gtk-ffi.fas > ;; Loading file > C:\subhasir\tmp\cells-gtk-2006-02-16\root\gtk-ffi\gtk- > core.fas > ... > ** - Continuable Error > FFI::FOREIGN-LIBRARY-FUNCTION: no dynamic object named "g_free" in library > :DEFAULT > If you continue (by typing 'continue'): Skip foreign object creation > The following restarts are also available: > RETRY :R1 > Retry performing # on # "gtk-core" #x19F461C1>. > ACCEPT :R2 > Continue, treating # on # "gtk-core" #x19F461C1> as having been successful. > SKIP :R3 skip (LET # # ...) > STOP :R4 stop loading file > C:\subhasir\tmp\cells-gtk-2006-02-16\l > oad.lisp > ABORT :R5 ABORT > > Thanks and regards, > Su > > > _______________________________________________ > cells-gtk-devel site list > cells-gtk-devel at common-lisp.net > http://common-lisp.net/mailman/listinfo/cells-gtk-devel > > -- Visit my blog at http://hackerlog.blogspot.com ===================================================== If more of us valued food and cheer and song above hoarded gold, it would be a merrier world. -- J.R.R. Tolkien From peter.denno at nist.gov Fri Jun 30 16:20:51 2006 From: peter.denno at nist.gov (Peter Denno) Date: Fri, 30 Jun 2006 12:20:51 -0400 Subject: [cells-gtk-devel] New tarballs Message-ID: <200606301220.52292.peter.denno@nist.gov> Hi, I put new tarballs/zip of cells-gtk on common-lisp.net: http://common-lisp.net/project/cells-gtk/ CVS is currently synchronized with these tarballs. The tarballs use the 2006-06-26 version of CFFI, and should load flawlessly in: Linux: Lispworks sbcl clisp Windows: clisp I haven't tested Allegro (I don't have it) and cmucl seems to have problems loading the libraries. I had a problem with LW/Win32, that was a bit strange, but another user seems to be OK in that regard. Drop the email group a note if you have any problems with these, or if LW/Win32 works, or if you discover what's up with cmu... etc. To see what's really new relative to CVS, I guess you'd have to look in the repository. Functionally this isn't much different than other recent releases, it just loads cleaner. See ya, -- - Best regards, Peter