gmane.lisp.cells.devel

Re: How to get started?

Ken Tilton — 2012-04-18T18:28:36

Hey, thx everyone for pitching in. I spend most of my time as ktilton
< at > mcna dot net (and the google "easy account switching" kinda did not
work so well when I tried it, tho it seems they are working on it).

I agree with all said and would add only that the regression test
suite has good examples.

-kt

On Tue, Apr 17, 2012 at 6:42 PM, Peter Hildebrandt
<peter.hildebrandt< at >gmail.com> wrote:

Re: can one make dependencies between models?

Frank Goenninger — 2012-04-18T08:39:35

Am 18.04.2012 um 05:29 schrieb Mirko Vukovic:


How exactly did you do the setf ?


What does the inspection of your model and the cells slot reveal ?


Recalculation is simply forced by "reading" / accessing a cells slot... As soon as you say (c? (some-slot model)) or even just do a (format ..."~s" (some-slot model)) this triggers a recompute cycle in cells - the slot needs to be current to be correct so Cells triggers all dependencies...


Frank

Re: can one make dependencies between models?

Mirko Vukovic — 2012-04-18T03:32:02

Thanks Peter.

That activity is still in the future.  But from my own historical
perspective, the very reason I switched to lisp several years ago was the
desire for such self-GUI-fying models.  I had a blast lisping, but never
got to this original goal :-)

On Tue, Apr 17, 2012 at 6:57 PM, Peter Hildebrandt <
peter.hildebrandt< at >gmail.com> wrote:

Re: can one make dependencies between models?

Mirko Vukovic — 2012-04-18T03:29:58

Thank you very much Frank, I think I got it.

A follow-up question.  With this technique, I have several models, and the
data is flowing from one model to the next.  Nice.

But is there a suggested procedure to temporarily isolate the model, so
that it can be tested by itself?  In other words, can one explicitly set
values to cells whose initform defines them in terms of other model's slots.

I tried setf on a cell that was linked to another model.  It seems to work.

But, is its initform (c? (let ((m (fm^ :other-model))) (blah (slot m)
...))) still active?
How can I test that?  By forcing that model to recompute itself.  How do I
do that?  I did not find an obvious candidate in the exported symbols.

Thanks again,

Mirko

On Tue, Apr 17, 2012 at 5:13 PM, Frank Goenninger <frgo< at >me.com> wrote:

Re: can one make dependencies between models?

Peter Hildebrandt — 2012-04-17T22:57:55

I used cells-gtk3 for exactly this purpose, i.e. linking the
position/text/color of graphic elements to the state of cells models.

Depending on your objectives, some of the more complex widgets might
be very helpful:
- a treeview that reflects a hierarchical object structure (supporting
drag and drop if I recall correctly)
- a listview (grid) that shows slots of models in a list (can even be editable)
- a canvas (based on cairo) that can be populated by visual primitives
(boxes, circles, lines, text fields) that mirror cells models (e.g.
start point of a line can be linked to the position of a box, and if
the user drags/drops the box, the line follows like a connector)

Let me know if you have further questions.  I wrote a fairly complex
application using cells/cells-gtk (including a simple physics
simulator using cells-ode and opengl), and I'm happy to dig up sample
code if needed.

Peter



_______________________________________________
cells-devel site list
cells-devel< at >common-lisp.net
http://common-lis

Re: How to get started?

Peter Hildebrandt — 2012-04-17T22:42:03

Hi Jim,

cells-gtk and celtk are probably the best choices for desktop GUI
toolkits.  You probably want to look at both to decide which you
prefer.  (full disclaimer, I did a bit of work on cells-gtk).
Cells-gtk has the advantage of looking native on Ubuntu/Gnome (and is
somewhat portable to windows).  Celtk is more cross-platform and has a
lower threshold if you are familiar with Tk.

you can download the latest fork of cells-gtk3 here:
https://github.com/Ramarren/cells-gtk3

Let me know if you have questions.
Peter


On Tue, Apr 17, 2012 at 11:12 AM, Frank Goenninger <frgo< at >me.com> wrote:

_______________________________________________
cells-devel site list
cells-devel< at >common-lisp.net
http://common-lisp.net/mailman/listinfo/cells-devel

Re: can one make dependencies between models?

Frank Goenninger — 2012-04-17T21:13:51

... Output of the short demo below:

(m-test)

M1: New value for slot a => 1.
M1: New value for slot b => 1.
M2: New value for slot c => 1.

STEP 1

M1: New value for slot a => 2.
M2: New value for slot c => 2.

STEP 2 

M1: New value for slot b => 3.
M2: New value for slot c => 6.

... cute, ha ?

;-)))

Frank

Am 17.04.2012 um 23:08 schrieb Frank Goenninger:

Re: can one make dependencies between models?

Frank Goenninger — 2012-04-17T21:08:14

Hi Mirko,

Am 17.04.2012 um 20:17 schrieb Mirko Vukovic:


Sure:

(in-package #:cells)

;;; ------------------
;;; ***  Model M1  ***
;;; ------------------
  
(defmd m1 ()
  a
  b
  :a (c-in 1)
  :b (c-in 1))

(defobserver a ((self m1))
  (when new-value
    (format *debug-io* "~%~S: New value for slot a => ~S."
      self (a self))))

(defobserver b ((self m1))
  (when new-value
    (format *debug-io* "~%~S: New value for slot b => ~S."
      self (b self))))

(defmacro mk-m1 (id)
  `(make-instance 'm1
     :fm-parent *parent*
     :md-name ,id))

;;; ------------------
;;; ***  Model M2  ***
;;; ------------------

(defmd m2 ()
  (c (c? (let ((m1 (fm^ :m1)))   ;; -> fm^ searches for :m1 in the current family
           (* (a m1) (b m1))))))

(defmacro mk-m2 (id)
  `(make-instance 'm2
     :fm-parent *parent*
     :md-name ,id))

(defobserver c ((self m2))
  (when new-value
    (format *debug-io* "~%~S: New value for slot c => ~S."
      self (c self))))

;;; ------------------
;;; ***  Family M  ***
;;; -

can one make dependencies between models?

Mirko Vukovic — 2012-04-17T18:17:22

Hello,

I am doing very basic cell-stuff, much like the ones in the doc folder:
Not liking excel and its cousins, I am implementing spread-sheet like
calculations in cells.

My question:

1) Can I build two models (model1, model2) and specify that a slot in
model2 depends on changes in some other slot in model1?
2) Related: can I change a slot specification in a model.  For example from
`c-in' to `c?'.  I assume that I can, but I would have to re-initialize the
model somehow.  Correct?

I am also very interested in the question posted just a few minutes ago.  I
would like to build an automated way of generating a GUI front end my cell
models.

Thanks,

Mirko

Re: How to get started?

Frank Goenninger — 2012-04-17T18:12:56

Hi Jim,

Am 17.04.2012 um 19:18 schrieb Jim Prewett:


Celtk is running fine here, also Cells-Gtk is still in use AFAIK.


https://github.com/kennytilton/cells/wiki


defmodel is like defclass - with just a few more slot options to declare a cell.
defmd is a convenience macro that helps in writing less boilerplate code:

With defmodel a cellsified class might look like this:

(DEFMODEL CAT (MODEL)
                 ((NAME :INITFORM NIL :INITARG :NAME)
                  (FUR-COLOR :INITFORM NIL :INITARG :FUR-COLOR)
                  (TITLE :INITARG :TITLE :INITFORM
                         (C? (UTILS-KT::CONC$ (^NAME) " with " (^FUR-COLOR) " fur."))))
                 (:DEFAULT-INITARGS 
                       :NAME (C-IN "") 
                       :FUR-COLOR (C-IN "black")))

the same could be written as:

(defmd cat ()
              name
              fur-color
              (title (c? (utils-kt::conc$ (^name) " with " (^fur-color) " fur.")))
              :name (c-in "")
              :fur-color (c-in "bla

How to get started?

Jim Prewett — 2012-04-17T17:18:07

Hello,

I've recently stumbled across Cells and it sounds pretty interesting.  I'm 
wondering how to get started learning about how to use Cells.  In general, 
I've got several Lisp projects that could use a GUI and the various 
Cells-based GUI projects look interesting to me.  I think I should 
probably learn how to use Cells (at some basic level) before trying to 
dive into one of the GUI packages.  Is there a Cells-based GUI package 
that is currently preferred?

In general, I learn best by first following a simple example, then 
modifying it to do my bidding.  Can someone show me how to get started?

I'm also wondering if someone can explain the difference between DEFMODEL 
and DEFMD to me.

Thanks so much!,
Jim

James E. Prewett                    Jim< at >Prewett.org download< at >hpc.unm.edu 
Systems Team Leader           LoGS: http://www.hpc.unm.edu/~download/LoGS/ 
Designated Security Officer         OpenPGP key: pub 1024D/31816D93    
HPC Systems Engineer III   UNM HPC  505.277.8210

Re: TripleCells: A new twist to the unbound variable CACHE challenge ...

Kenneth Tilton — 2012-03-22T17:20:50

cache needs the explicit package as well:

(cond ((string-equal h \"arrive\") \"home\")
             ((string-equal h \"leave\") \"away\")
             (triple-cells::cache? triple-cells::cache) ;;<-- need package on both
             (t \"away\")))))"


HTH, ken

Re: Triple Cells: AllegroGraph 4.5 "porting"

Frank Goenninger — 2012-03-20T17:50:48

Am 20.03.2012 um 18:27 schrieb Kenneth Tilton:


No need to! Just DL VMware Fusion and run a pre-packaged, complete VM from Franz with AG4.5 server! Really easy ...


No.


ah ... Beginning to understand ...


See above, The VMware Fusion of VMware Player is available for Windows, too!


oh yeah ...


On the test build already. 

;; Frank

Re: Triple Cells: AllegroGraph 4.5 "porting"

Kenneth Tilton — 2012-03-20T17:27:20

Ugh, and find a Linux machine. I forgot about all that.

Is AG3 still available? You could use it to figure out what should be 
happening, and confirm my guess that you can just take 3c-cell-pulse and 
make it look like 3c-pulse: test that the triple is found before 
returning the object. You may then blow up on testing upi= with nil, but 
just do my suggested upi-equal.

I took some time to look at the code and I do not see that tcells get 
their !pulse set when created, nor do I see ensure-current being called 
necessarily up front, so it might just work if you make this one change.

The danger is that some incompatibility somewhere else is changing the 
code path so that initialization or ensure-current is different, in 
which case the above fix would just mask the real problem -- but then it 
should break again soon enough.

If AG3 can be had for Windows I could give it a try, if not it might be 
a while before I can get to a Linux set-up.

Falback is a painful email debugging, with more diagnostics.

b

Re: Triple Cells: AllegroGraph 4.5 "porting"

Frank Goenninger — 2012-03-20T16:29:47

Am 20.03.2012 um 15:10 schrieb Kenneth Tilton:


Yep, true.


hehe, hope that whatever is keeping you busy is kind of cool and worth the sweat ;-)


Would be good ... ;-)


I only do AG4.5. Earlier versions don't work anymore - many API changes. You see a (require :agraph4) instead of just :agraph in my Triple-Cells variant (file defpackage.lisp). AG4.x is the currently distributed stuff, so no reason to do earlier versions.


I ran into an error due to sp being nil: (object sp) barfs then, of course.


Ah - yes, that sounds like something worth exploring! 


Yep, I see - only thing is to also download AG4.5 server and also AG4.5 clients from Franz and run against that AG4.5 server.


Thanks much! 

Cheers
   Frank

Re: Triple Cells: AllegroGraph 4.5 "porting"

Kenneth Tilton — 2012-03-20T14:10:15

Hey, good idea. Thanks.


Yeah, blank nodes are hard to look at.

In the above get-sp is supposed to return a triple, the object of which 
is the datapulse. So the question is why get-sp returns nil?

You are right, btw, I am so busy I forgot to look at the code when I got 
home. I just spent ten minutes just looking at the code and, wow, this 
will take some refreshing and I'll have to run it and trace a bit to see 
what is up. I'll see if I can get to it tonight.

Have you ever run it on any AG4.x? I do know there have been enough 
changes that I am not surprised we have to get the hood up.

One thought: did you add the bit that complains about not getting a 
pulse? Since this is early in the life of the cell I can imagine that 
the intent was to treat a missing !pulse as "uninitialized". In which 
case I think the error you got was in the caller where upi= tries to 
compare UPIs and it gets upset over being passed nil. In which case it 
would just mean they tightened up something and we need to do our ow

Re: Triple Cells: AllegroGraph 4.5 "porting"

Frank Goenninger — 2012-03-19T22:04:56

Am 19.03.2012 um 15:47 schrieb Kenneth Tilton:


Wow - I thought BLOBs would be the way to go. For PostgreSQL, it's this:
http://www.postgresql.org/docs/current/static/datatype-binary.html

[snip]


It is! I just instrumented it a bit with lots of calls to #'trc.


Yes, as in:

(defun 3c-cell-pulse (c)
  (trc "3c-cell-pulse: c = " c)
  (let ((sp (get-sp c !ccc:pulse)))
    (trc "3c-cell-pulse: sp = " sp)
    (if sp
      (progn
        (let ((object (object sp)))
          (trc "3c-cell-pulse object" object)
          object)) ;; <<<<----- This is the datapulse, I assume ...
      (progn
        (trc "3c-cell-pulse: sp is NIL !")
        (break)))))

It is right at the first time #'3c-cell-pulse is called when it fails... I am still trying to wrap my head around your use of subject, predicate, object in 3c ;-) It get's pretty "anonymous" with all those "anon" blank nodes ...


I /think/ it's one pulse (there's only one call to #'3c-pulse-advance)...


Sure - see attachment. I suggest you unzip it somewhere a

Re: Triple Cells: AllegroGraph 4.5 "porting"

Kenneth Tilton — 2012-03-19T14:47:26

Great. I love the data model (and you know I love Cells). On my project 
we have a PostGres table called "triple". Only problem is the object 
column because SQL does not like untyped columns (so I create object 
columns for any type I need to store <g>).


Not sure I follow the output, but that could be because I have not 
looked at triplecells in ages. I do not even have that source on this 
laptop!

Big Q: the above looks like it is from the Hello World example/test. 
Yes? Is sp the current datapulse? I notice your log does not include 
that, and I would think that would be interesting (unless of course as 
it seems the whole history is one pulse <g>)

I can try looking at this tonight. Are you free to send me your triple 
cells project so I have your diagnostics (and just to be sure we are 
debugging the same thing)?

Good to hear from you!

ken

Additional info AllegroGraph 4.5 porting: Screendumpof RDF DB

Frank Goenninger — 2012-03-18T22:04:09

.... see attached screendump.

TIA again.

Cheers
    Frank

Re: functional programming

Kenneth Tilton — 2012-03-19T14:24:21

I think you have it about right: Cell rules are nicely declarative and 
generally functional, but in the end Cells is all about managing a 
complex pile of application state representing <whatever real-world 
system we are modelling>. So, for example, the RoboCells implementation 
kept its own internal model of the game, with players and the ball 
modelled as CLOS instances. This model got updated as new sensory inputs 
were received. Sounds like "state" to me. :) That said, I imagine even 
functional programmers have to deal with long-lived models so maybe 
....a monad? Sorry, I wish I knew more about the functional game.

One thought that crosses my mind is Clojure with its functional state 
(if I have that right) and I know some folks over there were playing 
with implementing something Cells-ish in Clojure. Maybe chack in with 
those folks for thoughts from folks more functionally inclined than myself.


Haha, I always love when during refactoring I come to a comment that is 
no longer needed because th

Triple Cells: AllegroGraph 4.5 "porting"

Frank Goenninger — 2012-03-18T21:53:13

Hi Kenny, 

I just began using Triple Cells. We have planned a rather large AllegroGraph-based implementation and as a fan of Cells I am porting Triple Cells to AllegroGraph 4.5.

I am running in a situation where I get a problem with 3c-pulse-advance:
(I have instrumented 3c a bit ...;-)

0> PULSE> ---- advancing: (:change)
0> 3c-make storing type !hw:computer future-part
0> 3c-in nil t equal !hw:echo-happen {_:anon4}
0> stmt-new called with {_:anon3} !hw:happen {_:anon4} {_:anon4}
0> stmt-new (3c-cell? o) <_:anon4 type input>
0> 3c-install-cell called with s p o {_:anon3} !hw:happen {_:anon4}
0> subject-cells-node called with s =  {_:anon3}
0> subject-cells-node tr nil
0> subject-cells-node new-blank-node {_:anon5}
0> subject-cells-node (add-triple s !ccc:cells n) 13
0> 3c-install-cell subject-cells-node s = {_:anon5}
0> subject-cells-node called with s =  {_:anon3}
0> subject-cells-node tr <_:anon3 cells _:anon5>
0> subject-cells-node (object tr) =  {_:anon5}
0> 3c-install-cell (add-triple (subject-cells-