- cells-devel - mailman3.common-lisp.net

[cells-devel] initial stab at using Cells
by Jack Unrue 10 Apr '06

10 Apr '06

Well, I've gotten through my first round of experimenting with Cells, and I have to say I like it a lot. If you want to see some code, take a look at: http://common-lisp.net/websvn/filedetails.php?repname=graphic-forms&path=%2… which is the data model for a simple game where you click on blocks to score points (and where the remaining blocks fall down to fill in the hole(s)). The game gets progressively harder as you pass from one level to the next. The UI displays the current level, current score, and points needed for the next level. Screenshot here: http://sourceforge.net/project/screenshots.php?group_id=163034&ssid=31548 If you look at the model, there is a shape-data slot that really ought to be ephemeral based on how the score slot's formula depends on it (the 'shape' score should only be added to the overall score tally once, no matter how many times the UI asks for the score). But I was seeing some intermittent behavior where it seemed like the shape-data value had already been propagated (and thus cleared) by the time that my UI code asked for the score. So I took the :ephemeral attribute off until I get some more time to figure that out. I'm sure this is a bug in my understanding, not in Cells. It would be good to document some typical strategies for dealing with self-referential slots (circularity) but in thinking about how I would describe what I did, I can see how it would be hard to explain to someone else :-) -- Jack Unrue

2 3

[cells-devel] tiny patch for running Cells 3 on CLISP
by Jack Unrue 09 Apr '06

09 Apr '06

Hello Kenny, As of CLISP 2.38 (if not earlier, I haven't checked), the #+clisp defun for slot-definition-name is no longer necessary. With that commented out, motor-control.lisp runs fine. Just thought you'd like to know :-) I'm starting to look at Cells 3 with my first experiment being to write a simple game as a demo for my Graphic-Forms project, using Cells as the data model of course. I'll let you know how it goes. -- Jack Unrue

2 3

[cells-devel] Celtk on ACL80 running on an MacBook Pro ...
by Frank Goenninger 07 Apr '06

07 Apr '06

Kenny, fellow Cells Users, Just to show that in you can do it principle here's a screenshot showing the result of running Celtk / Cells kicked off by (ctk::tk- test) in AllegroCL 8.0 on my brand new MacBook Pro ... Of course, there's other apps available also ;-) Have fun - and yes, ACL is running sloooooowww ... (Rosetta obviously keeping both cores of the Intel CPU busy at 80% load level ...) Frank

2 1

[cells-devel] Re: Celtk/Cells3 declarative, dataflow paradigm contrasted with Ltk/imperative paradigm
by Ken Tilton 05 Apr '06

05 Apr '06

John, Sorry, it was late, I reacted too hastily. That error message looks wild, not sure what to make of it. Unfortunately, I have no Linux system. But the original developers of LTk use Linux and are looking at Celtk these days. I have not heard from them, but a query on the cells-devel mailing list might turn up someone using both Linux and Celtk. Otherwise, it might helpt to track down tk-format-now and tweak it so that the branch to display messages sent to Tk always fires, then see (a) if those messages look okay (post them to cells-devel) and (b) how far you get before Celtk dies. kt On 4/5/06, John Landahl <john(a)landahl.org> wrote: > > > Right at the top of ltktest-cells-inside.lisp you will find: > > > > #+test-ltktest > > (progn > > (cells-reset 'tk-user-queue-handler) > > ; ...huge comment elided.... > > (tk-test-class 'ltktest-cells-inside)) > > Yep, saw that, and I suppose I expected that loading celtk would set > test-ltktest. Removing the #+ and re-evaluating does startup the > test. > > However, the result is a bunch of errors and a big traceback. Here's > what looks like the most useful chunk: > > 0> numparse | :=> 42read-data:UNKNOWNcolor name "systembuttonface" > > read-data:INVALIDcommand name ".cv-scroller.test-canvas" > > > INVALIDCOMMANDNAME.cv-scroller.test-canvasUNKNOWNCOLORNAMEsystembuttonfaceBADWINDOWPATHNAME.cv-scroller.test-canvasINVALIDCOMMANDNAME.cv-scroller.test-canvas.bkg-popINVALIDCOMMANDNAME.cv-scroller.test-canvas.bkg-popINVALIDCOMMANDNAME.cv-scroller.test-canvas.bkg-popINVALIDCOMMANDNAME.cv-scroller.test-canvas.bkg-popCANREAD.cv-scroller.test-canvas.bkg-pop.bkg > An error of type READER-ERROR has occured: READER-ERROR on > #<TWO-WAY-STREAM > :INPUT-STREAM > #<SB-SYS:FD-STREAM for "descriptor 14" {A755EC9}> > > :OUTPUT-STREAM #<SB-SYS:FD-STREAM for "descriptor 13" {A755541}>>: > illegal terminating character after a colon: #\ > 0: (SB-DEBUG:BACKTRACE > 536870911 > #<SB-SYS:FD-STREAM for "a constant string" {AFCDB01}>) > 1: (LTK::TRIVIAL-DEBUGGER #<READER-ERROR {AA34AF1}> #<unavailable > argument>) > 2: (INVOKE-DEBUGGER #<READER-ERROR {AA34AF1}>) > 3: (ERROR READER-ERROR) > 4: (SB-IMPL::%READER-ERROR > #<TWO-WAY-STREAM > :INPUT-STREAM #<SB-SYS:FD-STREAM for "descriptor 14" {A755EC9}> > :OUTPUT-STREAM #<SB-SYS:FD-STREAM for "descriptor 13" {A755541}>> > "illegal terminating character after a colon: ~S") > 5: (SB-IMPL::READ-TOKEN > #<TWO-WAY-STREAM > :INPUT-STREAM #<SB-SYS:FD-STREAM for "descriptor 14" {A755EC9}> > :OUTPUT-STREAM #<SB-SYS:FD-STREAM for "descriptor 13" {A755541}>> > #\:) > 6: (READ-PRESERVING-WHITESPACE > #<TWO-WAY-STREAM > :INPUT-STREAM #<SB-SYS:FD-STREAM for "descriptor 14" {A755EC9}> > :OUTPUT-STREAM #<SB-SYS:FD-STREAM for "descriptor 13" {A755541}>> > NIL > NIL > T) > 7: (READ-PRESERVING-WHITESPACE > #<TWO-WAY-STREAM > :INPUT-STREAM #<SB-SYS:FD-STREAM for "descriptor 14" {A755EC9}> > :OUTPUT-STREAM #<SB-SYS:FD-STREAM for "descriptor 13" {A755541}>> > NIL > NIL > NIL) > 8: (LTK::MAIN-ITERATION :BLOCKING T) > 9: ((LAMBDA ())) > 10: ((LABELS LTK::USE-DEBUGGER) > #<FUNCTION LTK::TRIVIAL-DEBUGGER> > #<CLOSURE (LAMBDA #) {AA3477D}>) > 11: (LTK:MAINLOOP :SERVE-EVENT NIL) > > I'm running SBCL 0.9.11 on Linux, if it matters. My Tk version is > 8.4. > > > No, but do compare it to 'ltktest-cells-inside to get a feel for how > > Cells transforms development. > > I definitely will. I'd very much like to do more declarative > programming where it makes sense, and (as you and others have pointed > out) UIs are a natural fit. So it will be good to compare your > approach to the more usual (less intelligent) approach. > -- > John Landahl <john(a)landahl.org> > http://landahl.org/john >

1 1

[cells-devel] Re: Cells: How to later set a formula for a cell ?
by Ken Tilton 03 Apr '06

03 Apr '06

Goenninger, Frank wrote: > Hi Kenny, > > currently heavily playing wit Celtk... > > I have up to now only set a cell's formula by using directly (c? ...) > with default-initargs. Now I am stopped cold by the fact that I have > to be able to set the formula by some setf form. How to do that in a > conforming way ? Download KR from the Garnet system, they have a million backdoors like that. :) What's the big picture on this? Maybe there is a way to do it without an alley-oop. If you sell me on the need, and you can at least start the slot as c-input, I imagine there is some way to do this. If you do not sell me on the need, I will still give you some pointers* on how to mess up the beauty of Cells yourself.:) But I think you will still need to start with a c-input, or you will lose the dependencies (Cells does not record dependencies on slots not mediated by Cells.) ken (going out for groceries soon) * Look at c-install, md-awaken, c-awaken-cell, and their neighbors.

2 2

Re: [cells-devel] Celtk: How to set title of Wish window and also name of app ?
by Ken Tilton 03 Apr '06

03 Apr '06

Frank Goenninger wrote: > Another question: > > Am 02.04.2006 um 06:12 schrieb Ken Tilton: > >> Goenninger, Frank wrote: >> >>> -----BEGIN PGP SIGNED MESSAGE----- >>> Hash: SHA512 >>> >>> Hi Kenny and all others: >>> >>> I do have the following mini app now: >>> >>> - -x-x-x-x-x- >>> >>> (defmodel psu-rc-app (window) >>> () >>> (:default-initargs >>> :id :psu-rc-app-w >>> :title$ "System Power Supply Remote Control (Version A.01.00)" >>> :kids (c? (the-kids >>> (app-menubar))))) >> > > Why does the title not get set via this default-initargs statement ? This is the same story: I just have not done anything yet to handle the window class itself. Background: If you expand any of the (deftk <widget>...) forms (actually, I strongly recommend you do that and stare at /all/ the varieties of code) you will see they expand into quite a bit of code, including a (defobserver <config parameter> ...) for every slot/tk config parameter. The way Celtk (and Cells-Gtk) work is that CLOS classes are defined for every TK widget type. A CLOS slot is defined for every tk config parameter. So with widgets you just set the corresponding CLOS instance slot and a defobserver sends that to Tk via: "<widget> configure -<config param name> <value from CLOS slot>". I just need to do for the window class what I did with all the widgets and canvas items (the other job deftk handles). What title$ needs is: (defobserver title$ ((self window)) (tk-format '(:configure "title") "wm title . ~s" (or new-value "Untitled"))) As I look at the other "wm..." commands I can see I first need to add some slots to the window class and then for each a defobserver to shoot the value across to Tk on any change. [Side note: if you track down the mechanism by which make-instance leads to a corresponding TK widget creation, you will find some code that sweeps throug all the slots populated at make-instance time and includes thos config options in the widget creation. Which is why the generated defobservers check for old-value-boundp; the first time the observer gets invoked (the only time old-value-boundp is nil) is during make-instance, and the observer need not do anything since make-instance (eventually) gets the configs across in the widget creation command. > > And another one: > > I see there is the title$ slot of a window. How do I setf this > correctly ? > The following does not work: > > (setf (title$ (fm^ :psi-rc-app-w)) "my new title") > > It (of course) emits an error that title$ is not a valid function - > yes, no, ahem, what? why? I am not sure I have the same code you have, since I have already changed that slot's initform thus: (title$ :initarg :title$ :accessor title$ :initform (c? (string-capitalize (class-name (class-of self))))) So your error could be for two reasons. Either I originally coded it thus (title$ :reader title$....) thinking the title$ would never change, which is nonsense because a doc window starts out "Untitled" and then that changes when they save. Anyway, if I do not tell CLOS to create a writer it will not do so. Simple. But I doubt I specified just "reader". More likely (because even i can see it on the current code base) title$ is not exported and you are playing around in celtk-user, as I did deliberately to find out which symbols need to be exported. So just add title$ to the symbols exported from the defpackage :celtk form. > > Thanks for some help here., The greater help I can offer is what I do when I cannot believe my eyes on an error: (apropos "title$"), though we both use AllegrroCL so we get to use the groovy Apropos dialog. That not only shows you if a symbol is exported, it will show you if there is a setf as well as a reader. In a case like yours where you have already compiled code and it has failed, you will see the symbol title$ in both the celtk and celtk-user packages. That is a tip-off that the celtk symbol was not exported, because had it been the compiler would not have interned a new symbol in celtk-user when it got to the reference to title$. kt

1 1

[Fwd: Re: [cells-devel] Celtk: How to set title of Wish window and also name of app ?]
by Ken Tilton 02 Apr '06

02 Apr '06

-------- Original Message -------- Subject: Re: [cells-devel] Celtk: How to set title of Wish window and also name of app ? Date: Sat, 01 Apr 2006 23:12:23 -0500 From: Ken Tilton <kentilton(a)gmail.com> Reply-To: kentilton(a)gmail.com To: Goenninger, Frank <fgoenninger(a)prion.de> References: <F43CECE8-7EFF-46B9-A4AA-106AA92E3265(a)prion.de> Goenninger, Frank wrote: >-----BEGIN PGP SIGNED MESSAGE----- >Hash: SHA512 > >Hi Kenny and all others: > >I do have the following mini app now: > >- -x-x-x-x-x- > >(defmodel psu-rc-app (window) > () > (:default-initargs > :id :psu-rc-app-w > :title$ "System Power Supply Remote Control (Version A.01.00)" > :kids (c? (the-kids > (app-menubar))))) > >;; THIS IS HOW I SET THE TITLE OF THE WISH SHELL WINDOW >;; Any better way? > >(defmethod initialize-instance :after ((self psu-rc-app) &key) > (tk-format '(:configure "title") "wm title . HI")) > > Yes and no. No in that, no, there is not a better way. :) Yes in that there should be. :) Since there is only one Tk class like this, a fancy macro might not have much value. What we need is a bunch of: (defobserver title$ ((self window)) (tk-format '(:configure "title") "wm title . ~a" (or new-value "Untitled"))) ...observers, one per "wm XXXX...." that might follow from a window attribute. Note that, unlike intervening during i-i, this works even on /changes/ to the title (which would be unusual, but for other window attributes perfectly normal. so let's hand;e the general case. Also... no, sorry, i forgot myself: the ii call to tk-format gets deferred until the tk-client-que-handler decides to dispatch configures, which is after things get made, so the window has been instantiated on the TK side before the tile change gets dispatched. Anyway, if someone wants to code up the observers I will happily add them to Celtk, if not I will certainly get to them eventually as my need arises. Right now I am going quietly mad trying to make sense of Tcl/Tk handling of fonts and text items in a canvas. :) As for the application name, er, can you copy wish.exe to frank.exe? Otherwise, if the doc does not help, I would try comp.lang.tcl. But if there is anyway to control that, I imagine we can get Celtk to handle it. ken

1 1

[cells-devel] Celtk: How to set title of Wish window and also name of app ?
by Goenninger, Frank 02 Apr '06

02 Apr '06

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA512 Hi Kenny and all others: I do have the following mini app now: - -x-x-x-x-x- (defmodel psu-rc-app (window) () (:default-initargs :id :psu-rc-app-w :title$ "System Power Supply Remote Control (Version A.01.00)" :kids (c? (the-kids (app-menubar))))) ;; THIS IS HOW I SET THE TITLE OF THE WISH SHELL WINDOW ;; Any better way? (defmethod initialize-instance :after ((self psu-rc-app) &key) (tk-format '(:configure "title") "wm title . HI")) (defun app-menubar () (mk-menubar :id :psu-rc-menu-bar :kids (c? (the-kids (mk-menu-entry-cascade-ex (:label "File") (mk-menu-entry-separator) (mk-menu-entry-command :label "Quit" :command "exit")))))) (defun run-psu-rc-app () (cells-reset 'tk-user-queue-handler) (tk-test-class 'psu-rc-app)) - -x-x-x-x-x- Hmm - now I wonder if this is really the right way to set the title of the window ... Next question: On the Mac there is the Application name displayed in the top left corner when an app gets activated. This always says "Wish" which is something I want to set to the my own application name .. Purely a Mac and Tcl/Tk problem but ... Thanks for any feedback. Cheers! Frank -----BEGIN PGP SIGNATURE----- Version: PGP Desktop 9.0.5 (Build 5050) iQIVAwUBRC8m8GAKVUddnkqnAQpQhA/9GkwObg42wTUXiS2faUVNEqP57/nS2OqF ++7HRu2s2NCGvanLRCoJZOMVXrsasHE0K7izLCuMJAqF67YDr6UsO8EOJVkPHWIg ogb6nc5FYO/pAR2MzbuH00JUURs4VSNTc61bC+JV47kzlkt5Uzx1BMNOWZoNWrUs lPWanXxDNcmlbPmSY6Jz252/kzR3jyGlbQViBqZtPNkkrCuip8Bs39VBqr4iaQRR LVMisvjas/is3KbBIVlAyLpX2gBvEivPa+oXfTmnG0rWUbfkYKms7EUT9uwWhxVG TOTP/8eZkFinPtObs/8hUcKNDSPs/4rAzPvH75ioOGZGkM2yAeNC6rTbOgyDvOvh bEBGhO20e064CmowKfbqZNFBHwmCtVvMopFhSMeaFEwBMPqjutAyWQ0F8/CXd121 BeA6ARRwi80q/hXPy38n80udY6XPkgNlNp7WAKIIoyiqZuaC9TMgik0t+W6WKdnq qrm1sZ4arT6tkzoBrobShZAk4pTl+k4ukB0wXIhwPPOUHR/mzPJzOKnkEFNtsAXi b1oF9A6deu8dB25HOv+aNRF6yUUAwf8eUvBhCpIGkfyjP6/hw2h56bgyPkmriH8Y 58BeT8tEoxNQhmoDryWbIRFgIVGjRw1d4D5yo8I6xeXmPXitGymaIDs9AVu9srhp shYQCEvr3NA= =+dKG -----END PGP SIGNATURE-----

1 0

[cells-devel] Evolving Ltktest-with-cells.lisp
by Ken Tilton 26 Mar '06

26 Mar '06

I have been refining both the doc and the functionality of ltktest-with-cells (in the Celtk module under Cells), which mirrors the LTk demo ltktest, for a few days now adding more and more connectivity between the widgets and the main display, as one would have in a more real-world application. At this point I think both doc and demo are adequate, and anyway I will not know how to improve them without feedback, whether it be suggestions or requests for clarification. btw, thx to Frank for his report that all runs well on Mac OS X. I myself saw recent versions working on Lispworks+win32, albeit with a problem in that the spinning lines spun at a glacial rate. I would be interested in hearing if Celtk runs on other platforms. Now here is a big heads up: every once in a while, except when I am trying to make it happen <g>, ACL8 on win32 hangs when I close ltktest-with-cells. Possibly something to do with my Timer class run amok? We will see. ken

1 0

[cells-devel] (+ macosx Celtk Cells3 ACL8.0): This combo is running ...
by Goenninger, Frank 26 Mar '06

26 Mar '06

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA512 Heya, just wanted to let you know that I got the above mentioned combination to work as expected by just downloading the stuff and then do an ASDF load. Well, not really - I had to adjust some hardcoded pathnames in demos.lisp to find appropriate external files (bitmap files, Kenny's eyes from year 1969...) Very nice! Will use that to make a GUI for an app that controls and monitors an HF amplifier for Ham Radio... Cheers Frank -----BEGIN PGP SIGNATURE----- Version: PGP Desktop 9.0.5 (Build 5050) iQIVAwUBRCXnvWAKVUddnkqnAQqnDQ/9Fg6honUHO2YCW14kFilFPZ57YRnDf5CF RIhPGiMOwEd4YCYYuldnQZHYJQJpKKz/+zZan5LL35u62w6rwtIm5eqz749nLF0I XDfEK5Rq3jwC64A20OEU8KIMCVp6hZ8Lr9t7cfBI5WElLEbyvCM+5TQOf5fPVmSf FEYpp8XUTXej4z0g846g2tnzrNy6LJVWG61Sn88ZjbrdsOlRHSElzKnU7rFnqIh7 xFQF/jz4lH6calkuXqafZWM9gQbQs/1ein5rdrb4KVR+it8W7MFpFR8pU/9mhZ99 HufJJZw4oKP9Cxevl2DVWCCGI8RT7UgUsYFofhwFOYr8j7GXzUgedNloIYuwEomt GpNvkpl0Qaz0ZtbY36Rb4Ae/NLd75jL85MQ4Xlk9R4CRBfErbM6dnfAxgCbL/D2D CFd4jq83b5i2rJXgakuV8Ow/HvhBYLjcYDf1a3Tb8J4a9tdSy5e60qnzLTEsKZAw Fqru+IG+cmTydTyn4yAnB8+3Q10uAwdUY293E3SWgLNOdgWgGIUibRCY+IoCcbCL IQS8CwcOxklkCdfkbmbvr4VIM7tWIXEO9IC6UrMHGcnlfLGsVHTeuG8rUwtAb2oR U+vLuOpPvLc6ccvQgJbJIX65a3MZmPUwpUgyNY4rK8RtDH2/0UDqxc0pTeLKb8HS WDVR42qnNwM= =a8QA -----END PGP SIGNATURE-----

1 0