Re: [croquet] Anyone want to Croquet from C5 in Japan?

Not a clue.

The other demos working tells us that openGL is working, and by  
connecting KAT locally you're taking the network out of it.

But errors are good! What do yours say?

-H

On Jan 26, 2007, at 9:05 PM, Bill Allen wrote:

Bill,
I had the the same errors, but only without OpenAl proper install. I think, you
need to check out for openal libraries in system folders.

Regards,
Nikolay

Howard Stearns wrote: The sequence that occurs is that I run Croquet.sh from the
CroquetCollaborative directory, click on the "Click Me" button, clear
the address, and accept. The screen tells me it's waiting for connection
forever until I click in the red part, or just mouse over the blank part
of the screen above the red part, then it freezes.

Pardon the screen shot, but when the screen freezes, I can grab the text.

Regards,
Bill

croquet collaborative error

Nikolay Suslov wrote:
> Bill,
> I had the the same errors, but only without OpenAl proper install. I think, you
> need to check out for openal libraries in system folders.
>
> Regards,
> Nikolay
>  
Hi Nikolay,

II have OpenAL installed. It's my understanding that the Croquet Demo
Master, et al, wouldn't run without it. I know on my AMD64 desktop, this
was a problem since ubuntu doesn't put the lib32 version in by default
and it's a manual exercise to get it.

 ls -l /usr/lib/libopenal*
lrwxrwxrwx 1 root root     18 2006-03-27 18:56 /usr/lib/libopenal.so.0
-> libopenal.so.0.0.8
-rw-r--r-- 1 root root 307808 2006-03-27 05:51 /usr/lib/libopenal.so.0.0.8

Regards,
Bill

Bill Allen wrote:

Bill Allen wrote: Croquet.sh is attempting to make a link from libopenal.so.1 to
libopenal.so in /usr/lib. Since I'm not running sudo or root, this was
failing silently. I made the link by hand. The result is the same, but
the error is different:

Ah, the picture you attached is showing the first few frames of the  
debugger, without details such as instance variables and arguments.    
But already we know that during the execution of a method called  
#mouseEnter:, something was nil - aka the "UndefinedObject."  
Something inside of #mouseEnter: was sending a message to an object  
that it expected to be non-nil, but which was nil. The  
UndefinedObject did not know how to respond to whatever the message  
was that was sent to it.

We can find out what message and object by looking at the  
#mouseEnter: frame in the debugger.
Just click on that line in the debugger.  If you're lucky, the  
debugger summary will get bigger to show you the source code for the  
#mouseEnter: method. The method that it's trying to send will be  
highlighted in green. The thing just to the left of that the thing  
that is nil and isn't supposed to be.

[If you're not lucky, nothing at all will happen when you click on  
the #mouseEnter: line in the debugger summary. If this is the case,  
try pressing cmd/alt-. (that's a period).]

Now, in this particular case, I happen to know that there's some  
subtlety that I don't quite understand between the Morphic  
initialization and Croquet, and that the KAT participant tickles this  
issue more than other demos. If you look in  
KCroquetParticipant>>mouseMove:, you'll see that there's a little guard:

mouseMove: evt
        "Sometimes happens before we get initialized!"
        harness ifNil: [^false].
        ^super mouseMove: evt.

We probably need the same sort of thing in #mouseEnter:. But you need  
to look at the frame as described above to see WHAT we have to guard  
for being nil.


On Jan 27, 2007, at 11:26 AM, Bill Allen wrote:

This is different. A failure within OpenAL>>isCurrent is almost  
certainly an openAL problem.

I've found it quite hard to debug openAL. A lot our usage assumes it  
is working and doesn't tell you nicely when it isn't.  If you don't  
get this OpenAL>>isCurrent problem on Windows, I would suggests  
getting things working on Windows first, and then going back to Linux  
to work on OpenAL without having worry about getting it complicated  
by other issues.

Now, I HAVE had such problems on Windows, too.  I had a machine that  
had a bunch of games installed on it by someone else, and these had  
completely screwed up the OpenAL libraries such that even  
reinstalling different forms of OpenAL didn't help.  I ended up  
uninstalling all the games and then reinstalling OpenAL, and that  
fixed things. (I didn't have time to test after uninstalling each  
game, so I can't tell you which one was being evil.)

On Jan 27, 2007, at 11:40 AM, Bill Allen wrote:

On Jan 27, 2007, at 18:40 , Bill Allen wrote:

> I noticed that Croquet.sh is attempting to make a link from  
> libopenal.so.1 to libopenal.so in /usr/lib.

Is it? That would be very odd. We used to have the script make a link  
to /usr/lib/libGL.so.1, but the link was not created in a system  
directory, but rather inside the Croquet directory structure. The  
same, of at all, should happen for libopenal. The better alternative  
would be to refer to "libopenal.so.1" from the Croquet code when  
running under Linux. We had that discussion before (wrt OpenGL), not  
sure what the current status is on that.

- Bert -

Bert Freudenberg wrote: You're right that the script in the sdk distribution doesn't have the
openal.so link, but the one with the CroquetCollaborative dist does have it.

Bill

Howard Stearns wrote:

You're right. I installed openal on windows and I can get on locally.
When I try to do it to www.croquetccollaborative.org however, the screen
just stays red forever.

Bill

On Jan 27, 2007, at 19:19 , Bill Allen wrote:

Well, then maybe you should notify the maintainers of that distribution.


- Bert -

Because it crashed at 3 am this morning. I've reset. Should be fine  
now. Thanks for the head's up.

It was running unattended for three days and 37 visits. Getting there...

We didn't  cache any new shapshots while it was up, so we've lost the  
little gifts that folks made. (I liked the balloon-head stick  
figure!)  I haven't reached any conclusions yet, but I'm thinking  
that we should try to make little stop-the-world garbage collector,  
and that every time the last person leaves, we should GC or cache a  
snapshot or both. And we've got to do some sort of watchdog...

On Jan 27, 2007, at 1:28 PM, Bill Allen wrote:
> ... When I try to do it to www.croquetccollaborative.org however,  
> the screen just stays red forever.
>

I believe I've been added to this listserve by mistake. I imagine there must be some jfoster out there who would understand all of this much better than I do.

Regards,
Jamie Foster

----- Original Message -----
From: Bill Allen <[hidden email]>
Date: Saturday, January 27, 2007 12:21 pm
Subject: Re: [croquet] Anyone want to Croquet from C5 in Japan?
To: [hidden email], Bert Freudenberg <[hidden email]>

It looks like the libGL stuff (SDK distro and CroquetCollaborative)  
still do things in the .sh rather than in Squeak, but it does try to  
make the link locally:

if [ ! -x /usr/lib/libGL.so -a -x /usr/lib/libGL.so.1 ] ; then
        if [ ! -x  "$EXE/libGL.so" ] ; then
                echo Creating libGL.so symlink in $EXE
                ln -sf /usr/lib/libGL.so.1 "$EXE/libGL.so"
        fi
fi

The old SDK distro doesn't say anything at all about openAL in  
Croquet.sh.
A Linux colleague pointed out to me that it should, so we added this  
to Croquet.sh in the CroquetCollaborative distro:

if [ ! -x /usr/local/lib/libopenal.so ] ; then
        ln -s /usr/lib/libopenal.so.0 /usr/local/lib/libopenal.so
fi

I think Bert's points are well taken -- that if it isn't done in  
Squeak, it should at least create the link in the executable  
directory.  I don't have a Linux box to test, so if someone wants to  
test and post corrected code, that would be great.

I don't see a bug ticket on this, so I just entered one at http://
bugs.impara.de/view.php?id=5849

On Jan 27, 2007, at 12:16 PM, Bert Freudenberg wrote:

Howard Stearns wrote:
I tried out the CroquetCollaborative this week and couldn't get it to
work --- it stuck on openal (debugger opened in squeak) when it tried to
start. I just didn't have the time to track down the problem, but I
assumed it was a link problem after I looked at the bash script.

I'll try it out, where's the script?

(deja vu... didn't we have this same conversation a while back.)



--
brad fuller
 http://www.Sonaural.com/
 +1 (408) 799-6124

On Jan 28, 2007, at 1:40 , Brad Fuller wrote:
> I'll try it out, where's the script?

Look for Croquet.sh

> (deja vu... didn't we have this same conversation a while back.)

Oh yes we did:

        https://lists.wisc.edu/read/messages?id=857047

- Bert -

I get a unresponsive image on my Mac with CroquetCollaborative 1.0.17.7

Here is the console.log and SqueakDebug.log

Mac OS X Version 10.3.9 (Build 7W98)
2007-01-28 10:32:06 +0100
2007-01-28 10:36:24.292 SystemUIServer[264] CPSLaunchAppWithData (/Applications/Roxio Toast 6 Titanium/Toast 6 Titanium.app, ..)
Warning: unrecognized command line flag -psn_0_2097153
/Applications/Firefox.app/Contents/MacOS/firefox-bin: can't map file: /Library/Internet Plug-Ins/Flash Player.plugin ((os/kern) invalid argument)
2007-01-28 15:47:18.174 Squeak VM Opt[380] *** _NSAutoreleaseNoPool(): Object 0x40b040 of class NSCFString autoreleased with no pool in place - just leaking
2007-01-28 15:47:18.175 Squeak VM Opt[380] *** _NSAutoreleaseNoPool(): Object 0x40b7e0 of class NSPathStore2 autoreleased with no pool in place - just leaking
2007-01-28 15:47:18.176 Squeak VM Opt[380] *** _NSAutoreleaseNoPool(): Object 0x40cf60 of class NSPathStore2 autoreleased with no pool in place - just leaking
2007-01-28 15:57:16.052 Squeak VM Opt[384] *** _NSAutoreleaseNoPool(): Object 0x40b040 of class NSCFString autoreleased with no pool in place - just leaking
2007-01-28 15:57:16.053 Squeak VM Opt[384] *** _NSAutoreleaseNoPool(): Object 0x40b7e0 of class NSPathStore2 autoreleased with no pool in place - just leaking
2007-01-28 15:57:16.053 Squeak VM Opt[384] *** _NSAutoreleaseNoPool(): Object 0x40cf60 of class NSPathStore2 autoreleased with no pool in place - just leaking
2007-01-28 15:59:50.030 Squeak VM Opt[388] *** _NSAutoreleaseNoPool(): Object 0x40b040 of class NSCFString autoreleased with no pool in place - just leaking
2007-01-28 15:59:50.031 Squeak VM Opt[388] *** _NSAutoreleaseNoPool(): Object 0x40b7e0 of class NSPathStore2 autoreleased with no pool in place - just leaking
2007-01-28 15:59:50.031 Squeak VM Opt[388] *** _NSAutoreleaseNoPool(): Object 0x40cf60 of class NSPathStore2 autoreleased with no pool in place - just leaking
2007-01-28 15:59:50.185 Squeak VM Opt[388] *** _NSAutoreleaseNoPool(): Object 0x414660 of class NSPathStore2 autoreleased with no pool in place - just leaking
2007-01-28 15:59:54.656 Squeak VM Opt[389] *** _NSAutoreleaseNoPool(): Object 0x447640 of class NSPathStore2 autoreleased with no pool in place - just leaking
2007-01-28 16:01:43.019 Squeak VM Opt[392] *** _NSAutoreleaseNoPool(): Object 0x40b040 of class NSCFString autoreleased with no pool in place - just leaking
2007-01-28 16:01:43.020 Squeak VM Opt[392] *** _NSAutoreleaseNoPool(): Object 0x40b7e0 of class NSPathStore2 autoreleased with no pool in place - just leaking
2007-01-28 16:01:43.020 Squeak VM Opt[392] *** _NSAutoreleaseNoPool(): Object 0x40cf60 of class NSPathStore2 autoreleased with no pool in place - just leaking

MessageNotUnderstood: UndefinedObject>>isCurrent
28 January 2007 4:02:04 pm

VM: Mac OS - a SmalltalkImage
Image: Croquet1.0beta [latest update: #0]

SecurityManager state:
Restricted: false
FileAccess: true
SocketAccess: true
Working Dir /Users/karl/Desktop/CroquetCollaborative-1.0
Trusted Dir /Users/karl/Desktop/CroquetCollaborative-1.0
Untrusted Dir /Users/karl/Library/Preferences/Squeak/Internet

UndefinedObject(Object)>>doesNotUnderstand:
        Receiver: nil
        Arguments and temporary variables:
                aMessage: isCurrent
        Receiver's instance variables:
nil

KCroquetParticipant(CroquetParticipantWithMenu)>>mouseEnter:
        Receiver: a KCroquetParticipant(1151)
        Arguments and temporary variables:
                evt: [171@100 mouseEnter 11827]
        Receiver's instance variables:
                bounds: 0@110 corner: 800@600
                owner: a PasteUpMorph<world>(4026) [world]
                submorphs: #(a KNavMorph(934) a KZoomNavigatorMorph(3908))
                fullBounds: 0@110 corner: 800@600
                color: Color red
                extension: a MorphExtension (599) [other:  (clipSubmorphs -> true) (tweakWorld ...etc...
                harness: a KStandardHarness
                lastBounds: nil
                dock: a KZoomNavigatorMorph(3908)
                windowData: nil
                entry: a TPostcardOID:[c2be09189969d56cc3073ec4d256cb6d]:'WisconsinWorld'-#main...etc...
                address: nil
                port: nil
                nav: a KNavMorph(934)

KCroquetParticipant(Morph)>>handleMouseEnter:
        Receiver: a KCroquetParticipant(1151)
        Arguments and temporary variables:
                anEvent: [171@100 mouseEnter 11827]
        Receiver's instance variables:
                bounds: 0@110 corner: 800@600
                owner: a PasteUpMorph<world>(4026) [world]
                submorphs: #(a KNavMorph(934) a KZoomNavigatorMorph(3908))
                fullBounds: 0@110 corner: 800@600
                color: Color red
                extension: a MorphExtension (599) [other:  (clipSubmorphs -> true) (tweakWorld ...etc...
                harness: a KStandardHarness
                lastBounds: nil
                dock: a KZoomNavigatorMorph(3908)
                windowData: nil
                entry: a TPostcardOID:[c2be09189969d56cc3073ec4d256cb6d]:'WisconsinWorld'-#main...etc...
                address: nil
                port: nil
                nav: a KNavMorph(934)

MouseButtonEvent(MouseEvent)>>sentTo:
        Receiver: [171@100 mouseEnter 11827]
        Arguments and temporary variables:
                anObject: a KCroquetParticipant(1151)
        Receiver's instance variables:
                timeStamp: 11827
                source: a HandMorph(2149)
                type: #mouseEnter
                buttons: 0
                position: 171@100
                handler: nil
                wasHandled: true
                whichButton: 4


--- The full stack ---
UndefinedObject(Object)>>doesNotUnderstand:
KCroquetParticipant(CroquetParticipantWithMenu)>>mouseEnter:
KCroquetParticipant(Morph)>>handleMouseEnter:
MouseButtonEvent(MouseEvent)>>sentTo:
 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
MouseButtonEvent>>sentTo:
KCroquetParticipant(Morph)>>handleEvent:
KCroquetParticipant(CroquetParticipantWithMenu)>>handleEvent:
[] in MouseOverHandler>>processMouseOver: {[:m | (m == focus    or: [m hasOwner: focus])   ifTrue: [localEvt := evt    ...]}
Array(SequenceableCollection)>>reverseDo:
MouseOverHandler>>processMouseOver:
HandMorph>>processEvents
[] in WorldState>>doOneCycleNowFor: {[:h |  ActiveHand := h.  h processEvents.  capturingGesture := capturingGest...]}
Array(SequenceableCollection)>>do:
WorldState>>handsDo:
WorldState>>doOneCycleNowFor:
WorldState>>doOneCycleFor:
PasteUpMorph>>doOneCycle
[] in Project class>>spawnNewProcess {[[World doOneCycle.  Processor yield.  false] whileFalse.  nil]}
[] in BlockContext>>newProcess {[self value.  Processor terminateActive]}

#isCurrent is certainly an OpenAL issue.

-H

On Jan 28, 2007, at 9:07 AM, Karl wrote: