f-script-talk Mailing List for F-Script (Page 14)

SourceForge Headquarters 225 Broadway Suite 1600 San Diego, CA 92101 +1 (858) 422-6466

Ken Ferry has released a SIMBL-based fix for running F-Script Anywhere 
on Mac OS X 10.3. You can get it fom Ken's software page at 
http://homepage.mac.com/kenferry/software.html

Best,

Phil

And this would also open up the possibility for lots of fun
with f-script:  <http://www.fscript.org/>

Since f-script lets you interactively play with Cocoa objects,
you could do all sorts of rapid-prototyping or just plain
playing around.

hmmm....  this may even be possible now with the plugin architecture,
i.e. to make an f-script plugin for Xgrid

-Nathan

On Jan 22, 2004, at 10:41 AM, Drew McCormack wrote:

> I don't know if this is in the works, but I would love to see Apple
> supply a Cocoa API for the xgrid agent. That is, a class that can be
> called from inside Cocoa code, and acts as an agent, communicating with
> the Controller. This would presumably open up untold uses. Imagine
> being able to just schedule some work to run on the local xgrid from
> inside a cocoa app. Need to process 100 images? Send it to the xgrid
> via the NSGridAgent class. Cool!
>
> Drew McCormack
>
> Trade Strategist
> Stock market strategy design software
>
> http://www.trade-strategist.com
> _______________________________________________
> xgrid-users mailing list | xgr...@li...
> Help/Unsubscribe/Archives: 
> http://www.lists.apple.com/mailman/listinfo/xgrid-users
> Do not post admin requests to the list. They will be ignored.
>
>

Le 16 janv. 04, =E0 03:34, Frank Hunter a =E9crit :
> Could anyone please tell me what program, and where obtained, to use=20=

> to open download >fscript_esug2002.ppt<
>

This file is in Microsoft PowerPoint format. To open it you'll need=20
PowerPoint or another program able to read PowerPoint files.

See:
http://www.microsoft.com/mac/products/powerpointx/powerpointx.aspx

Best,

Phil

Could anyone please tell me what program, and where obtained, to use to 
open download >fscript_esug2002.ppt<

Thank you

Frank Hunter
Robson Valley, BC
Canada

IMPORTANT BUG ALERT
-----------------------------------

A bug that can lead to incorrect results has been found in an old 
version of F-Script (v 1.2.3) and some older versions. The bug manifest 
itself when using a message pattern expression to compare two arrays of 
numbers, in certain cases, with the <= method (i.e., A <= B, where A 
and B are arrays of numbers).
You are advised to immediately update to newer versions (to download 
the latest version, go to 
http://www.fscript.org/download/download.htm).

-- Philippe Mougin

Le Mercredi, 22 octo 2003, =E0 21:07 America/Montreal, Philippe Mougin a=20=

=E9crit :

> Hi,
>
> The release candidate announced here a few days ago has been released
> as "F-Script 1.2.5".
>

Bravo Phil! Continue cet excellent travail!

A+

----------------------
Pierre

Hi,

The release candidate announced here a few days ago has been released
as "F-Script 1.2.5".

Release notes are available on-line at 
http://www.fscript.org/releaseNotes/
You can download F-Script 1.2.5 from 
http://www.fscript.org/download/download.htm

In other news, F-Script will be presented during the OOPSLA 2003 
conference and at the October BANG event (hosted by the O'Reilly Mac OS 
X Conference). See the "news page" at http://www.fscript.org/news.htm 
for more informations.

Also, note that ESUG, the European Smalltalk User Group, has released a 
new version of its CD-ROM. The CD-ROM contains various Smalltalk 
Systems as Test/Full/Non Commercial Version  (VisualWorks 7.1 NC, IBM 
VisualAge Smalltalk, Squeak 3.5, Smalltalk/X etc.). Also included: free 
out of print books, tutorials, lectures, columns, articles about 
smalltalk, ST-ANSI Standard and more. You can download the contents for 
free at http://www.ira.uka.de/~marcus/EsugcD.html

Best,

Phil 

Hi,

F-Script 1.2.5 candidate release is now available for download and test.

Binaries + documentation ->
http://www.fscript.org/download/FScriptBin-20031020.tar.gz

Sources ->
http://www.fscript.org/download/FScriptSources-20031020.tar.gz

The main new feature is support for Mac OS X 10.3.

I plan to publicly announce the release in a few days if no major 
problems are found in this version.

Best,

Phil

Le mardi, 16 sep 2003, =E0 10:56 Europe/Paris, Knud M=F6ller a =E9crit :

>> Finally, I should add that you can directly save and load blocks=20
>> to/from files. See section 9 in the manual.
>
> Yes, I had seen that. but [Block save/load] will serialize/deserialize=20=

> the Block-object, and not save a text representation, correct?

Yes you're right.

I hope your game will do well!

Best,

Phil=

Hi,

thanks for the quick and detailed answer. It really helped me to =20
understand what is going on behind the scenes a little bit better and =20=

solves my problem.
Am Montag, 15.09.03, um 23:20 Uhr (Europe/Berlin) schrieb Philippe =20
Mougin:

> Finally, I should add that you can directly save and load blocks =20
> to/from files. See section 9 in the manual.

Yes, I had seen that. but [Block save/load] will serialize/deserialize =20=

the Block-object, and not save a text representation, correct? And =20
since I want to be able to generate and manipulate my scripts in a text =20=

editor, I think a cannot use this method.

> BTW, feel free to let us know about your game and how you use F-Script =
=20
> in this project.

Well, it's nothing professional, but a project I want to submit to a =20
Macintosh game programming contest called uDevGames (see =20
http://www.idevgames.com/content/contest.php?id=3D10). The game ("Retro =20=

Maze") is an oldfashioned role playing game with a very simple, tile =20
based 3d-engine. The game's basic framework is done i Cocoa/ObjC. Now I =20=

plan to use fscript to glue everything together. More important, I want =20=

to attach scripts to various game objects (tiles, monster, walls, ...). =20=

These scripts will be executed whenever some condition is met (e.g. a =20=

tile's script is executed when the player walks over it) and provide =20
the game with the actual action.
I guess I could have used just about any scripting language to do that. =20=

However, fscript caught my eye through the articles on the =20
O'Reilly-network. It looked different, but very interesting and elegant =20=

- I was really fascinated by the fact that control structures are =20
implemented using blocks and messages! Since portability to other =20
platforms is not a concern for me, it doesn't matter that fscript is =20
only available on OSX at the moment.

Merci bien,

Knud
------------------------------------------------------------------------=20=

----------------
Knud M=F6ller -  Institut f=FCr Sprachliche Informationsverarbeitung,
Universit=E4t zu K=F6ln
(Department of Linguistic Data Processing,
University of Cologne, Germany)
http://www.spinfo.uni-koeln.de/~kmoeller

------------------------------------------------------------------------=20=

----------------

Dear Knud,

You get the expected behavior.

However, note that, actually, it is not the fact that you load the =20
source code of the block from a file that gets you this behavior but =20
the use of the "asBlock" method on a string. As stated in the manual: =20=

"... each block created by this method is in charge of its own =20
top-level workspace". In other words, the block created by this method =20=

has no knowledge of your F-Script session's workspace. Instead, it runs =20=

in its own workspace.

This may seems surprising at first. To better understand the rational =20=

of this behavior, ask yourself: how could the NSString asked to create =20=

a block (with the 'asBlock' method) know about your F-Script session? =20=

The answer is: it can't. Your NSString is an Objective-C object, like =20=

every object manipulated with F-Script. When it receives the 'asBlock' =20=

message, it doesn't know that this message comes from an F-Script =20
interpreter. Indeed, you can also directly send this message from an =20
Objective-C program as well.

What you need in order to achieve what you want is a way to ask the =20
current F-Script interpreter to create a block configured to use the =20
current workspace. By chance, the pre-defined identifier 'sys' refers =20=

to an object (of class System) which represent the current F-Script =20
interpreter, and provides the method you want (i.e., blockFromString:). =20=

Thus, if you write "sys blockFromString:codeFromFile", a new block will =20=

be created which will be configured to use the workspace of the =20
interpreter represented by the receiver (i.e., the current F-Script =20
interpreter).

Finally, I should add that you can directly save and load blocks =20
to/from files. See section 9 in the manual.

Hope this clarify things.

BTW, feel free to let us know about your game and how you use F-Script =20=

in this project.

Best,

Phil

Le lundi, 15 sep 2003, =E0 19:37 Europe/Paris, Knud M=F6ller a =E9crit :

> Hi,
>
> I'm new to fscript, and stumbled upon something that really puzzles =20=

> me. When I define a block directly from the prompt within the =20
> fscript-application, and define a variable within that block, the =20
> variable will be visible outside the block, too (after the block has =20=

> been sent 'value'). This is as I expected it. See this session for =20
> example:
>
> ---------------------------------------------------------------------
> > blockFromPrompt :=3D [ otherString :=3D 'this works!'. ]
>
> > blockFromPrompt value
>
> > otherString
> 'this works!'
> ---------------------------------------------------------------------
>
> However, when I load a block from a file, any variable defined in this =
=20
> block will NOT be visible outside. Compare this:
>
> ---------------------------------------------------------------------
> > codeFromFile :=3D NSString alloc initWithContentsOfFile: =20
> '/Users/dunk/Desktop/testScript.fscript'
>
> > blockFromFile :=3D codeFromFile asBlock
>
> > blockFromFile
> [
>     someString :=3D 'test me!'.
> ]
>
> > blockFromFile value
>
> > someString
>
> error: undefined identifier "someString"
> ---------------------------------------------------------------------
>
> Why is the 'someString' variable not visible, just as the =20
> 'otherString' variable? Is this a bug or a feature? If it is a =20
> feature, I don't really get it. Someone please enlighten me! =3D8-)
>
> Apart from that, I think that fscript is very cool! I plan to use it =20=

> in a game project of mine.
>
> Cheers,
> Knud
>
> PS: I use version 1.2.4
>
> =
-----------------------------------------------------------------------=20=

> --------
> Knud M=F6ller - Institut f=FCr Sprachliche Informationsverarbeitung,
> Universit=E4t zu K=F6ln
> (Department of Linguistic Data Processing, University of Cologne, =20
> Germany)
> http://www.spinfo.uni-koeln.de/~kmoeller

Hi,

I'm new to fscript, and stumbled upon something that really puzzles me. =20=

When I define a block directly from the prompt within the =20
fscript-application, and define a variable within that block, the =20
variable will be visible outside the block, too (after the block has =20
been sent 'value'). This is as I expected it. See this session for =20
example:

---------------------------------------------------------------------
 > blockFromPrompt :=3D [ otherString :=3D 'this works!'. ]

 > blockFromPrompt value

 > otherString
'this works!'
---------------------------------------------------------------------

However, when I load a block from a file, any variable defined in this =20=

block will NOT be visible outside. Compare this:

---------------------------------------------------------------------
 > codeFromFile :=3D NSString alloc initWithContentsOfFile: =20
'/Users/dunk/Desktop/testScript.fscript'

 > blockFromFile :=3D codeFromFile asBlock

 > blockFromFile
[
     someString :=3D 'test me!'.
]

 > blockFromFile value

 > someString

error: undefined identifier "someString"
---------------------------------------------------------------------

Why is the 'someString' variable not visible, just as the 'otherString' =20=

variable? Is this a bug or a feature? If it is a feature, I don't =20
really get it. Someone please enlighten me! =3D8-)

Apart from that, I think that fscript is very cool! I plan to use it in =20=

a game project of mine.

Cheers,
Knud

PS: I use version 1.2.4

------------------------------------------------------------------------=20=

-------
Knud M=F6ller - Institut f=FCr Sprachliche Informationsverarbeitung,
Universit=E4t zu K=F6ln
(Department of Linguistic Data Processing, University of Cologne, =20
Germany)
http://www.spinfo.uni-koeln.de/~kmoeller

Hi Joerg,

As you suggested, it would probably requires providing a specific=20
Objective-C method for block evaluation managing the creation and=20
destruction of the an autorelease pool (it would be easy too add it=20
yourself, for instance in a category). However, you should completely=20
ignore the problem for now, because of a much bigger issue: in the=20
current implementation, the F-Script run-time is not thread safe. That=20=

is, two threads in the same task should not access the F-Script=20
run-time concurrently; something which will happen in your case since=20
you're detaching a thread executing an F-Script block from an F-Script=20=

program.

Can you, by chance, re-formulate your application with a single active=20=

F-Script thread ?

Best,

Phil

Le jeudi, 7 ao=FB 2003, =E0 17:02 Europe/Paris, Joerg Garbers a =E9crit =
:

> Hi,
> what is the right way to make a new thread with an FScript block as=20
> target?
>
> The problem with
> NSThread detachNewThreadSelector:#value: toTarget:myBlock=20
> withObject:someObject.
> is that there is no NSAutoreleasePool in Place for the execution of=20
> myBlock.
>
> So myBlock must be shielded by some initialization:
> NSThread detachNewThreadSelector:#value: toTarget:(makeThreadableBlock=20=

> value:myBlock) withObject:someObject.
>
> Is there a really good implementation in FScript for this?
>
> makeThreadableBlock:=3D[:innerBlock |
>   [:arg ||p|
>     p:=3DNSAutoreleasePool alloc init.
>     innerBlock value:arg.
>     p release.
>     ]].
>
> yields an error, because NSAutoreleasePools may not be retained (by=20
> assignment to p). So no assignment? But how to empty the pool then?
>
> makeThreadableBlock:=3D[:innerBlock |
>   [:arg ||p|
>     NSAutoreleasePool alloc init.
>     innerBlock value:arg.
>     ]].
>
> but even the following does not catch all of the autoreleases (see=20
> console warnings)
>
> makeThreadableBlock:=3D[:innerBlock |
>   [:arg ||p|
>     p:=3DNSValue valueWithNonretainedObject:(NSAutoreleasePool alloc=20=

> init).
>     innerBlock value:arg.
>     p nonretainedObjectValue release.
>     ]].
>
> Any other solutions? Or should the Block method
> -(id) valueArgs:(id*)args count:(unsigned)count
> contain NSAutoreleasePool initialization and termination?
>
> Joerg
>

Hi,
what is the right way to make a new thread with an FScript block as 
target?

The problem with
NSThread detachNewThreadSelector:#value: toTarget:myBlock 
withObject:someObject.
is that there is no NSAutoreleasePool in Place for the execution of 
myBlock.

So myBlock must be shielded by some initialization:
NSThread detachNewThreadSelector:#value: toTarget:(makeThreadableBlock 
value:myBlock) withObject:someObject.

Is there a really good implementation in FScript for this?

makeThreadableBlock:=[:innerBlock |
   [:arg ||p|
     p:=NSAutoreleasePool alloc init.
     innerBlock value:arg.
     p release.
     ]].

yields an error, because NSAutoreleasePools may not be retained (by 
assignment to p). So no assignment? But how to empty the pool then?

makeThreadableBlock:=[:innerBlock |
   [:arg ||p|
     NSAutoreleasePool alloc init.
     innerBlock value:arg.
     ]].

but even the following does not catch all of the autoreleases (see 
console warnings)

makeThreadableBlock:=[:innerBlock |
   [:arg ||p|
     p:=NSValue valueWithNonretainedObject:(NSAutoreleasePool alloc 
init).
     innerBlock value:arg.
     p nonretainedObjectValue release.
     ]].

Any other solutions? Or should the Block method
-(id) valueArgs:(id*)args count:(unsigned)count
contain NSAutoreleasePool initialization and termination?

Joerg

An F-Script syntax coloring module for the Hydra text editor has been 
implemented by Andy Stark and is publicly available at 
http://hydra.globalse.org/developers.html

Best,

Phil

The current F-Script code already try to handle this situation=20
gracefully (although, as you point out, this does not work!). In the=20
current code, we use Apple's NSExceptionHandler class to set an=20
exceptionHandling mask to 63, which, according to Cocoa's doc should=20
raise an exception when a error like bad memory access happen. But=20
instead, F-Script hangs (it seems to be stuck in the=20
NSExceptionHandlerUncaughtSignalHandler() system function, which seems=20=

to be called by a signal handler).

Best,

Phil

Le mercredi, 11 jun 2003, =E0 08:34 Europe/Paris, Nicholas Seckar a =
=E9crit=20
:

> 	I've noticed that F-Script crashes when it tries to access a=20
> prematurely deallocated object. The crash is not a 'clean' failure, as=20=

> F-Script simply hangs until I kill it.
> 	While this isn't overly important, it would still be nicer to =
see=20
> this fixed. As usual, GDB logs EXC_BAD_ACCESS when this occurs. Anyone=20=

> want to write a signal handler?
>

	I've noticed that F-Script crashes when it tries to access a 
prematurely deallocated object. The crash is not a 'clean' failure, as 
F-Script simply hangs until I kill it.
	While this isn't overly important, it would still be nicer to see this 
fixed. As usual, GDB logs EXC_BAD_ACCESS when this occurs. Anyone want 
to write a signal handler?

Hi all,

Andy Stark has produced a bookmarked PDF version of the F-Script guide 
which is much easier to use than the version included in the current 
distribution. You can download it at 
http://www.fscript.org/download/FScriptGuide.pdf

Best,

Phil

Hi,

There is a quite well informed discussion about F-Script going on at 
the MacSlash web site.
Included are comparisons between F-Script, PyObjC, RubyCocoa and 
AppleScript/AppleScript Studio.

link: http://www.macslash.org/articles/03/06/05/1446248.shtml

Best,

Phil

Dear all,

The release candidate announced here a few days ago has been released 
as "F-Script 1.2.4".

Release notes available on-line at http://www.fscript.org/releaseNotes/
You can download from http://www.fscript.org/download/download.htm

Spread the word!

Best,

Phil 

Hi,

A problem in the first candidate release has been found and fixed. So 
we now have a new candidate release available for download and test.

Binaries + documentation ->
http://www.fscript.org/download/FScriptBin-20030529.tar.gz

Sources ->
http://www.fscript.org/download/FScriptSources-20030529.tar.gz

I plan to publicly announce the release in a few days if no major 
problems are found in this version.

  Best,

  Phil

Sorry. Your message could not be delivered to:

Andy Stark (Mailbox or Conference is full.)

Sorry. Your message could not be delivered to:

Andy Stark (Mailbox or Conference is full.)

Sorry. Your message could not be delivered to:

Andy Stark (Mailbox or Conference is full.)

Sorry. Your message could not be delivered to:

Andy Stark (Mailbox or Conference is full.)

2001	Jan	Feb	Mar (4)	Apr (1)	May (1)	Jun (1)	Jul	Aug (1)	Sep	Oct	Nov (1)	Dec (2)
2002	Jan (3)	Feb (1)	Mar (2)	Apr (8)	May (12)	Jun (3)	Jul (5)	Aug (4)	Sep (1)	Oct (1)	Nov (2)	Dec (4)
2003	Jan (1)	Feb	Mar (2)	Apr	May (14)	Jun (5)	Jul (1)	Aug (2)	Sep (4)	Oct (2)	Nov (2)	Dec
2004	Jan (3)	Feb (9)	Mar (1)	Apr (6)	May	Jun	Jul (3)	Aug (3)	Sep (11)	Oct (29)	Nov	Dec (3)
2005	Jan (8)	Feb (12)	Mar	Apr (1)	May (31)	Jun	Jul (7)	Aug (1)	Sep	Oct (2)	Nov (1)	Dec (2)
2006	Jan (13)	Feb (1)	Mar (5)	Apr (3)	May (1)	Jun (13)	Jul (2)	Aug (22)	Sep (15)	Oct (1)	Nov (1)	Dec
2007	Jan (7)	Feb (3)	Mar (2)	Apr (5)	May	Jun (6)	Jul (2)	Aug (1)	Sep (2)	Oct (3)	Nov (12)	Dec (5)
2008	Jan (2)	Feb	Mar (2)	Apr (2)	May (4)	Jun (31)	Jul (9)	Aug (10)	Sep (3)	Oct (4)	Nov	Dec
2009	Jan (3)	Feb	Mar	Apr	May	Jun	Jul	Aug (1)	Sep	Oct	Nov	Dec

f-script-talk Mailing List for F-Script (Page 14)

f-script-talk — Discussion about F-Script and related subjects