Mac Carbon VM 3.8.15b5 pending

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
12 messages Options
Reply | Threaded
Open this post in threaded view
|

Mac Carbon VM 3.8.15b5 pending

johnmci
I've pushed out a 3.8.15b5 to address an error in the 3.8.15b4 VM  
with the new browser plugin interface and firefox.

http://www.smalltalkconsulting.com/squeak.html

I added three new features. One is SqueakDebug setting in the VM and  
browser plugin to enable you to print some
diagnostics messages for the browser interface and module loading,  
I'll add more message in the future.

The second is a SqueakQuitOnQuitAppleEvent flag that is used to quit  
the VM immediately if we get the quit event, which happens
when a user logs out. Likely this is the wrong choice, still for  
people doing vertical apps you need someone to prevent the VM from
cancelling the logout process.  MIT's Scratch will use this feature.

Finally I altered the logic when we look at the edit menu 'Quit do  
not save' If the application has used the VM's builtin feature
to modify the mac menus then they might have changed the menu item  
from 'Quit do not save' to perform Foobar.


3.8.15b5
                        squeakBrowser-info.plist add SqueakDebug parm (1) debug msgs, (0)  
no debug msgs

                        npsqueak.c streamline context usage a bit, make local var versus  
on structure
                                                        add logic to print msgs based on info.plist

                        sqMacMain.c Add SqueakDebug parm (1) debug msgs, (0) no debug msgs
                                                        debug messages for browser interface, and module loading
                                                        Add SqueakQuitOnQuitAppleEvent, true if we should quit  
immediately if
                                                        operating system sends us quit msg, as a result of logout.  
Normally false.

                        sqMacUIAppleEvents.c If you logout we get a quit apple event msg,  
normally we ignore,
                                                        unless SqueakQuitOnQuitAppleEvent is set.

                        sqMacUIEventsUniversal.c  If the user picks edit menu, item one,  
actually check to see if
                                                        the string is "Quit do not save", if so then quit, otherwise  
pass revised menu
                                                        choice to Smalltalk.

                        sqMacUnixExternalPrims.c Look to SqueakDebug for msg printing

               
                        sqMacWindowUniversal.c Look to SqueakDebug for msg printing, mind  
file does not print any msgs yet

--
========================================================================
===
John M. McIntosh <[hidden email]>
Corporate Smalltalk Consulting Ltd.  http://www.smalltalkconsulting.com
========================================================================
===



Reply | Threaded
Open this post in threaded view
|

Re: Mac Carbon VM 3.8.15b5 pending

Bruce O'Neel-5
Hi,

There seems to be a small problem with yellow on intel macs only, with
Squeak 3.8.15beta1U from the FTP site.  

If you download http://earth.pckswarms.ch/earthSwarms-1.0_Mac.dmg.gz and run it on a PPC system then the mountain peaks will be yellowish/brownish, but,
on an intel mac the mountain peaks will be purblish/bluish.  As you push the
button labeled topography until it says cloud fraction, it's different
yellows on a PPC but all white on an Intel mac.

The PPC behavior is the same as Linux and Windows.

Thanks very much for all your work on the Mac VM!

cheers

bruce
   
----- Message d'origine -----
De: John M McIntosh <[hidden email]>
Date: Sat, 10 Mar 2007 16:32:23 -0800
Sujet: Mac Carbon VM 3.8.15b5  pending
À: The general-purpose Squeak developers list <[hidden email]>

>I've pushed out a 3.8.15b5 to address an error in the 3.8.15b4 VM  
>with the new browser plugin interface and firefox.
>
>http://www.smalltalkconsulting.com/squeak.html
>
>I added three new features. One is SqueakDebug setting in the VM and  
>browser plugin to enable you to print some
>diagnostics messages for the browser interface and module loading,  
>I'll add more message in the future.
>
>The second is a SqueakQuitOnQuitAppleEvent flag that is used to quit  
>the VM immediately if we get the quit event, which happens
>when a user logs out. Likely this is the wrong choice, still for  
>people doing vertical apps you need someone to prevent the VM from
>cancelling the logout process.  MIT's Scratch will use this feature.
>
>Finally I altered the logic when we look at the edit menu 'Quit do  
>not save' If the application has used the VM's builtin feature
>to modify the mac menus then they might have changed the menu item  
>from 'Quit do not save' to perform Foobar.
>
>
>3.8.15b5
> squeakBrowser-info.plist add SqueakDebug parm (1) debug msgs, (0)  
>no debug msgs
>
> npsqueak.c streamline context usage a bit, make local var versus  
>on structure
> add logic to print msgs based on info.plist
>
> sqMacMain.c Add SqueakDebug parm (1) debug msgs, (0) no debug msgs
> debug messages for browser interface, and module loading
> Add SqueakQuitOnQuitAppleEvent, true if we should quit  
>immediately if
> operating system sends us quit msg, as a result of logout.  
>Normally false.
>
> sqMacUIAppleEvents.c If you logout we get a quit apple event msg,  
>normally we ignore,
> unless SqueakQuitOnQuitAppleEvent is set.
>
> sqMacUIEventsUniversal.c  If the user picks edit menu, item one,  
>actually check to see if
> the string is "Quit do not save", if so then quit, otherwise  
>pass revised menu
> choice to Smalltalk.
>
> sqMacUnixExternalPrims.c Look to SqueakDebug for msg printing
>
>
> sqMacWindowUniversal.c Look to SqueakDebug for msg printing, mind  
>file does not print any msgs yet
>
>--
>========================================================================
>===
>John M. McIntosh <[hidden email]>
>Corporate Smalltalk Consulting Ltd.  http://www.smalltalkconsulting.com
>========================================================================
>===
>
>
>
>

Reply | Threaded
Open this post in threaded view
|

Re: Mac Carbon VM 3.8.15b5 pending

johnmci
Actually if you click on appeaarance and set the window or desktop  
color you'll see the colors in the color picker *are* the same.
But what is going on here is that the application is calculating a  
Balloon 3D surface, however something is wrong somewhere so the
surface colors aren't the same between the two platforms.

Someone would have to dig deeper to understand why the colors are not  
correct.

On Mar 27, 2007, at 8:55 AM, Bruce O'Neel wrote:

> Hi,
>
> There seems to be a small problem with yellow on intel macs only, with
> Squeak 3.8.15beta1U from the FTP site.
>
> If you download http://earth.pckswarms.ch/ 
> earthSwarms-1.0_Mac.dmg.gz and run it on a PPC system then the  
> mountain peaks will be yellowish/brownish, but,
> on an intel mac the mountain peaks will be purblish/bluish.  As you  
> push the
> button labeled topography until it says cloud fraction, it's different
> yellows on a PPC but all white on an Intel mac.
>
> The PPC behavior is the same as Linux and Windows.
>
> Thanks very much for all your work on the Mac VM!
>
> cheers
>
> bruce
>
> ----- Message d'origine -----
> De: John M McIntosh <[hidden email]>
> Date: Sat, 10 Mar 2007 16:32:23 -0800
> Sujet: Mac Carbon VM 3.8.15b5  pending
> À: The general-purpose Squeak developers list <squeak-
> [hidden email]>
>
>> I've pushed out a 3.8.15b5 to address an error in the 3.8.15b4 VM
>> with the new browser plugin interface and firefox.
>>
>> http://www.smalltalkconsulting.com/squeak.html
>>
>> I added three new features. One is SqueakDebug setting in the VM and
>> browser plugin to enable you to print some
>> diagnostics messages for the browser interface and module loading,
>> I'll add more message in the future.
>>
>> The second is a SqueakQuitOnQuitAppleEvent flag that is used to quit
>> the VM immediately if we get the quit event, which happens
>> when a user logs out. Likely this is the wrong choice, still for
>> people doing vertical apps you need someone to prevent the VM from
>> cancelling the logout process.  MIT's Scratch will use this feature.
>>
>> Finally I altered the logic when we look at the edit menu 'Quit do
>> not save' If the application has used the VM's builtin feature
>> to modify the mac menus then they might have changed the menu item
>> from 'Quit do not save' to perform Foobar.
>>
>>
>> 3.8.15b5
>> squeakBrowser-info.plist add SqueakDebug parm (1) debug msgs, (0)
>> no debug msgs
>>
>> npsqueak.c streamline context usage a bit, make local var versus
>> on structure
>> add logic to print msgs based on info.plist
>>
>> sqMacMain.c Add SqueakDebug parm (1) debug msgs, (0) no debug  
>> msgs
>> debug messages for browser interface, and module loading
>> Add SqueakQuitOnQuitAppleEvent, true if we should quit
>> immediately if
>> operating system sends us quit msg, as a result of logout.
>> Normally false.
>>
>> sqMacUIAppleEvents.c If you logout we get a quit apple event msg,
>> normally we ignore,
>> unless SqueakQuitOnQuitAppleEvent is set.
>>
>> sqMacUIEventsUniversal.c  If the user picks edit menu, item one,
>> actually check to see if
>> the string is "Quit do not save", if so then quit, otherwise
>> pass revised menu
>> choice to Smalltalk.
>>
>> sqMacUnixExternalPrims.c Look to SqueakDebug for msg printing
>>
>>
>> sqMacWindowUniversal.c Look to SqueakDebug for msg printing, mind
>> file does not print any msgs yet
>>
>> --
>> =====================================================================
>> ===
>> ===
>> John M. McIntosh <[hidden email]>
>> Corporate Smalltalk Consulting Ltd.  http://
>> www.smalltalkconsulting.com
>> =====================================================================
>> ===
>> ===
>>
>>
>>
>>

--
========================================================================
===
John M. McIntosh <[hidden email]>
Corporate Smalltalk Consulting Ltd.  http://www.smalltalkconsulting.com
========================================================================
===



Reply | Threaded
Open this post in threaded view
|

Re: Re: Mac Carbon VM 3.8.15b5 pending

Bruce O'Neel-5
Hi,

Yea, sorry, that was a horrible description of the glitch.

Yes, what I see is that only on B3D surfaces does yellow not show up.
The colors normally are just perfect.

Thanks!

cheers

bruce

----- Message d'origine -----
De: John M McIntosh <[hidden email]>
Date: Tue, 27 Mar 2007 09:30:12 -0700
Sujet: Re: Mac Carbon VM 3.8.15b5  pending
À: "Bruce O'Neel" <[hidden email]>
Cc: The general-purpose Squeak developers list <[hidden email]>

>Actually if you click on appeaarance and set the window or desktop  
>color you'll see the colors in the color picker *are* the same.
>But what is going on here is that the application is calculating a  
>Balloon 3D surface, however something is wrong somewhere so the
>surface colors aren't the same between the two platforms.
>
>Someone would have to dig deeper to understand why the colors are not  
>correct.
>
>On Mar 27, 2007, at 8:55 AM, Bruce O'Neel wrote:
>
>> Hi,
>>
>> There seems to be a small problem with yellow on intel macs only, with
>> Squeak 3.8.15beta1U from the FTP site.
>>
>> If you download http://earth.pckswarms.ch/ 
>> earthSwarms-1.0_Mac.dmg.gz and run it on a PPC system then the  
>> mountain peaks will be yellowish/brownish, but,
>> on an intel mac the mountain peaks will be purblish/bluish.  As you  
>> push the
>> button labeled topography until it says cloud fraction, it's different
>> yellows on a PPC but all white on an Intel mac.
>>
>> The PPC behavior is the same as Linux and Windows.
>>
>> Thanks very much for all your work on the Mac VM!
>>
>> cheers
>>
>> bruce
>>
>> ----- Message d'origine -----
>> De: John M McIntosh <[hidden email]>
>> Date: Sat, 10 Mar 2007 16:32:23 -0800
>> Sujet: Mac Carbon VM 3.8.15b5  pending
>> À: The general-purpose Squeak developers list <squeak-
>> [hidden email]>
>>
>>> I've pushed out a 3.8.15b5 to address an error in the 3.8.15b4 VM
>>> with the new browser plugin interface and firefox.
>>>
>>> http://www.smalltalkconsulting.com/squeak.html
>>>
>>> I added three new features. One is SqueakDebug setting in the VM and
>>> browser plugin to enable you to print some
>>> diagnostics messages for the browser interface and module loading,
>>> I'll add more message in the future.
>>>
>>> The second is a SqueakQuitOnQuitAppleEvent flag that is used to quit
>>> the VM immediately if we get the quit event, which happens
>>> when a user logs out. Likely this is the wrong choice, still for
>>> people doing vertical apps you need someone to prevent the VM from
>>> cancelling the logout process.  MIT's Scratch will use this feature.
>>>
>>> Finally I altered the logic when we look at the edit menu 'Quit do
>>> not save' If the application has used the VM's builtin feature
>>> to modify the mac menus then they might have changed the menu item
>>> from 'Quit do not save' to perform Foobar.
>>>
>>>
>>> 3.8.15b5
>>> squeakBrowser-info.plist add SqueakDebug parm (1) debug msgs, (0)
>>> no debug msgs
>>>
>>> npsqueak.c streamline context usage a bit, make local var versus
>>> on structure
>>> add logic to print msgs based on info.plist
>>>
>>> sqMacMain.c Add SqueakDebug parm (1) debug msgs, (0) no debug  
>>> msgs
>>> debug messages for browser interface, and module loading
>>> Add SqueakQuitOnQuitAppleEvent, true if we should quit
>>> immediately if
>>> operating system sends us quit msg, as a result of logout.
>>> Normally false.
>>>
>>> sqMacUIAppleEvents.c If you logout we get a quit apple event msg,
>>> normally we ignore,
>>> unless SqueakQuitOnQuitAppleEvent is set.
>>>
>>> sqMacUIEventsUniversal.c  If the user picks edit menu, item one,
>>> actually check to see if
>>> the string is "Quit do not save", if so then quit, otherwise
>>> pass revised menu
>>> choice to Smalltalk.
>>>
>>> sqMacUnixExternalPrims.c Look to SqueakDebug for msg printing
>>>
>>>
>>> sqMacWindowUniversal.c Look to SqueakDebug for msg printing, mind
>>> file does not print any msgs yet
>>>
>>> --
>>> =====================================================================
>>> ===
>>> ===
>>> John M. McIntosh <[hidden email]>
>>> Corporate Smalltalk Consulting Ltd.  http://
>>> www.smalltalkconsulting.com
>>> =====================================================================
>>> ===
>>> ===
>>>
>>>
>>>
>>>
>
>--
>========================================================================
>===
>John M. McIntosh <[hidden email]>
>Corporate Smalltalk Consulting Ltd.  http://www.smalltalkconsulting.com
>========================================================================
>===
>
>
>
>

Reply | Threaded
Open this post in threaded view
|

Re: Mac Carbon VM 3.8.15b5 pending

Bert Freudenberg
That may well be not a VM problem.

- Bert -

On Mar 28, 2007, at 9:56 , Bruce O'Neel wrote:

> Hi,
>
> Yea, sorry, that was a horrible description of the glitch.
>
> Yes, what I see is that only on B3D surfaces does yellow not show up.
> The colors normally are just perfect.
>
> Thanks!
>
> cheers
>
> bruce
>
> ----- Message d'origine -----
> De: John M McIntosh <[hidden email]>
> Date: Tue, 27 Mar 2007 09:30:12 -0700
> Sujet: Re: Mac Carbon VM 3.8.15b5  pending
> À: "Bruce O'Neel" <[hidden email]>
> Cc: The general-purpose Squeak developers list <squeak-
> [hidden email]>
>
>> Actually if you click on appeaarance and set the window or desktop
>> color you'll see the colors in the color picker *are* the same.
>> But what is going on here is that the application is calculating a
>> Balloon 3D surface, however something is wrong somewhere so the
>> surface colors aren't the same between the two platforms.
>>
>> Someone would have to dig deeper to understand why the colors are not
>> correct.
>>
>> On Mar 27, 2007, at 8:55 AM, Bruce O'Neel wrote:
>>
>>> Hi,
>>>
>>> There seems to be a small problem with yellow on intel macs only,  
>>> with
>>> Squeak 3.8.15beta1U from the FTP site.
>>>
>>> If you download http://earth.pckswarms.ch/
>>> earthSwarms-1.0_Mac.dmg.gz and run it on a PPC system then the
>>> mountain peaks will be yellowish/brownish, but,
>>> on an intel mac the mountain peaks will be purblish/bluish.  As you
>>> push the
>>> button labeled topography until it says cloud fraction, it's  
>>> different
>>> yellows on a PPC but all white on an Intel mac.
>>>
>>> The PPC behavior is the same as Linux and Windows.
>>>
>>> Thanks very much for all your work on the Mac VM!
>>>
>>> cheers
>>>
>>> bruce
>>>
>>> ----- Message d'origine -----
>>> De: John M McIntosh <[hidden email]>
>>> Date: Sat, 10 Mar 2007 16:32:23 -0800
>>> Sujet: Mac Carbon VM 3.8.15b5  pending
>>> À: The general-purpose Squeak developers list <squeak-
>>> [hidden email]>
>>>
>>>> I've pushed out a 3.8.15b5 to address an error in the 3.8.15b4 VM
>>>> with the new browser plugin interface and firefox.
>>>>
>>>> http://www.smalltalkconsulting.com/squeak.html
>>>>
>>>> I added three new features. One is SqueakDebug setting in the VM  
>>>> and
>>>> browser plugin to enable you to print some
>>>> diagnostics messages for the browser interface and module loading,
>>>> I'll add more message in the future.
>>>>
>>>> The second is a SqueakQuitOnQuitAppleEvent flag that is used to  
>>>> quit
>>>> the VM immediately if we get the quit event, which happens
>>>> when a user logs out. Likely this is the wrong choice, still for
>>>> people doing vertical apps you need someone to prevent the VM from
>>>> cancelling the logout process.  MIT's Scratch will use this  
>>>> feature.
>>>>
>>>> Finally I altered the logic when we look at the edit menu 'Quit do
>>>> not save' If the application has used the VM's builtin feature
>>>> to modify the mac menus then they might have changed the menu item
>>>> from 'Quit do not save' to perform Foobar.
>>>>
>>>>
>>>> 3.8.15b5
>>>> squeakBrowser-info.plist add SqueakDebug parm (1) debug  
>>>> msgs, (0)
>>>> no debug msgs
>>>>
>>>> npsqueak.c streamline context usage a bit, make local var  
>>>> versus
>>>> on structure
>>>> add logic to print msgs based on info.plist
>>>>
>>>> sqMacMain.c Add SqueakDebug parm (1) debug msgs, (0) no debug
>>>> msgs
>>>> debug messages for browser interface, and module loading
>>>> Add SqueakQuitOnQuitAppleEvent, true if we should quit
>>>> immediately if
>>>> operating system sends us quit msg, as a result of logout.
>>>> Normally false.
>>>>
>>>> sqMacUIAppleEvents.c If you logout we get a quit apple event  
>>>> msg,
>>>> normally we ignore,
>>>> unless SqueakQuitOnQuitAppleEvent is set.
>>>>
>>>> sqMacUIEventsUniversal.c  If the user picks edit menu, item one,
>>>> actually check to see if
>>>> the string is "Quit do not save", if so then quit, otherwise
>>>> pass revised menu
>>>> choice to Smalltalk.
>>>>
>>>> sqMacUnixExternalPrims.c Look to SqueakDebug for msg printing
>>>>
>>>>
>>>> sqMacWindowUniversal.c Look to SqueakDebug for msg printing,  
>>>> mind
>>>> file does not print any msgs yet
>>>>
>>>> --
>>>> ===================================================================
>>>> ==
>>>> ===
>>>> ===
>>>> John M. McIntosh <[hidden email]>
>>>> Corporate Smalltalk Consulting Ltd.  http://
>>>> www.smalltalkconsulting.com
>>>> ===================================================================
>>>> ==
>>>> ===
>>>> ===
>>>>
>>>>
>>>>
>>>>
>>
>> --
>> =====================================================================
>> ===
>> ===
>> John M. McIntosh <[hidden email]>
>> Corporate Smalltalk Consulting Ltd.  http://
>> www.smalltalkconsulting.com
>> =====================================================================
>> ===
>> ===
>>
>>
>>
>>
>



Reply | Threaded
Open this post in threaded view
|

Re: Re: Mac Carbon VM 3.8.15b5 pending

Bruce O'Neel-5
Hi,

The only reason why I suspect a VM problem is that Windows, OpenBSD big endian, Linux little endian, Mac PPC and Mac Intel running the PPC VM all get the same display.  The only one which is different is the Mac Intel running the universal binary.

Is it possible that the problem is in  Squeak3D/b3dTypes.h?

/* An ugly hack but I can't find the global defs in CodeWarrior on the Mac */
#ifndef LSB_FIRST
        #define MSB_FIRST
#endif

#ifndef MSB_FIRST
        #define RED_INDEX 0
        #define GREEN_INDEX 1
        #define BLUE_INDEX 2
        #define ALPHA_INDEX 3
#else  
        #define ALPHA_INDEX 0
        #define BLUE_INDEX 1
        #define GREEN_INDEX 2
        #define RED_INDEX 3
#endif

Though, that said, I'm not sure.

I'll try to borrow the MacBook again in the next day or so to see if I can run enough experiements to find the problem.  I get the old PPC laptop, which, I prefer, but it's hard to test.

Thanks!

cheers

bruce



----- Message d'origine -----
De: Bert Freudenberg <[hidden email]>
Date: Wed, 28 Mar 2007 10:00:54 +0200
Sujet: Re: Mac Carbon VM 3.8.15b5  pending
À: The general-purpose Squeak developers list <[hidden email]>
Cc: [hidden email]

>That may well be not a VM problem.
>
>- Bert -
>
>On Mar 28, 2007, at 9:56 , Bruce O'Neel wrote:
>
>> Hi,
>>
>> Yea, sorry, that was a horrible description of the glitch.
>>
>> Yes, what I see is that only on B3D surfaces does yellow not show up.
>> The colors normally are just perfect.
>>
>> Thanks!
>>
>> cheers
>>
>> bruce
>>
>> ----- Message d'origine -----
>> De: John M McIntosh <[hidden email]>
>> Date: Tue, 27 Mar 2007 09:30:12 -0700
>> Sujet: Re: Mac Carbon VM 3.8.15b5  pending
>> À: "Bruce O'Neel" <[hidden email]>
>> Cc: The general-purpose Squeak developers list <squeak-
>> [hidden email]>
>>
>>> Actually if you click on appeaarance and set the window or desktop
>>> color you'll see the colors in the color picker *are* the same.
>>> But what is going on here is that the application is calculating a
>>> Balloon 3D surface, however something is wrong somewhere so the
>>> surface colors aren't the same between the two platforms.
>>>
>>> Someone would have to dig deeper to understand why the colors are not
>>> correct.
>>>
>>> On Mar 27, 2007, at 8:55 AM, Bruce O'Neel wrote:
>>>
>>>> Hi,
>>>>
>>>> There seems to be a small problem with yellow on intel macs only,  
>>>> with
>>>> Squeak 3.8.15beta1U from the FTP site.
>>>>
>>>> If you download http://earth.pckswarms.ch/
>>>> earthSwarms-1.0_Mac.dmg.gz and run it on a PPC system then the
>>>> mountain peaks will be yellowish/brownish, but,
>>>> on an intel mac the mountain peaks will be purblish/bluish.  As you
>>>> push the
>>>> button labeled topography until it says cloud fraction, it's  
>>>> different
>>>> yellows on a PPC but all white on an Intel mac.
>>>>
>>>> The PPC behavior is the same as Linux and Windows.
>>>>
>>>> Thanks very much for all your work on the Mac VM!
>>>>
>>>> cheers
>>>>
>>>> bruce
>>>>
>>>> ----- Message d'origine -----
>>>> De: John M McIntosh <[hidden email]>
>>>> Date: Sat, 10 Mar 2007 16:32:23 -0800
>>>> Sujet: Mac Carbon VM 3.8.15b5  pending
>>>> À: The general-purpose Squeak developers list <squeak-
>>>> [hidden email]>
>>>>
>>>>> I've pushed out a 3.8.15b5 to address an error in the 3.8.15b4 VM
>>>>> with the new browser plugin interface and firefox.
>>>>>
>>>>> http://www.smalltalkconsulting.com/squeak.html
>>>>>
>>>>> I added three new features. One is SqueakDebug setting in the VM  
>>>>> and
>>>>> browser plugin to enable you to print some
>>>>> diagnostics messages for the browser interface and module loading,
>>>>> I'll add more message in the future.
>>>>>
>>>>> The second is a SqueakQuitOnQuitAppleEvent flag that is used to  
>>>>> quit
>>>>> the VM immediately if we get the quit event, which happens
>>>>> when a user logs out. Likely this is the wrong choice, still for
>>>>> people doing vertical apps you need someone to prevent the VM from
>>>>> cancelling the logout process.  MIT's Scratch will use this  
>>>>> feature.
>>>>>
>>>>> Finally I altered the logic when we look at the edit menu 'Quit do
>>>>> not save' If the application has used the VM's builtin feature
>>>>> to modify the mac menus then they might have changed the menu item
>>>>> from 'Quit do not save' to perform Foobar.
>>>>>
>>>>>
>>>>> 3.8.15b5
>>>>> squeakBrowser-info.plist add SqueakDebug parm (1) debug  
>>>>> msgs, (0)
>>>>> no debug msgs
>>>>>
>>>>> npsqueak.c streamline context usage a bit, make local var  
>>>>> versus
>>>>> on structure
>>>>> add logic to print msgs based on info.plist
>>>>>
>>>>> sqMacMain.c Add SqueakDebug parm (1) debug msgs, (0) no debug
>>>>> msgs
>>>>> debug messages for browser interface, and module loading
>>>>> Add SqueakQuitOnQuitAppleEvent, true if we should quit
>>>>> immediately if
>>>>> operating system sends us quit msg, as a result of logout.
>>>>> Normally false.
>>>>>
>>>>> sqMacUIAppleEvents.c If you logout we get a quit apple event  
>>>>> msg,
>>>>> normally we ignore,
>>>>> unless SqueakQuitOnQuitAppleEvent is set.
>>>>>
>>>>> sqMacUIEventsUniversal.c  If the user picks edit menu, item one,
>>>>> actually check to see if
>>>>> the string is "Quit do not save", if so then quit, otherwise
>>>>> pass revised menu
>>>>> choice to Smalltalk.
>>>>>
>>>>> sqMacUnixExternalPrims.c Look to SqueakDebug for msg printing
>>>>>
>>>>>
>>>>> sqMacWindowUniversal.c Look to SqueakDebug for msg printing,  
>>>>> mind
>>>>> file does not print any msgs yet
>>>>>
>>>>> --
>>>>> ===================================================================
>>>>> ==
>>>>> ===
>>>>> ===
>>>>> John M. McIntosh <[hidden email]>
>>>>> Corporate Smalltalk Consulting Ltd.  http://
>>>>> www.smalltalkconsulting.com
>>>>> ===================================================================
>>>>> ==
>>>>> ===
>>>>> ===
>>>>>
>>>>>
>>>>>
>>>>>
>>>
>>> --
>>> =====================================================================
>>> ===
>>> ===
>>> John M. McIntosh <[hidden email]>
>>> Corporate Smalltalk Consulting Ltd.  http://
>>> www.smalltalkconsulting.com
>>> =====================================================================
>>> ===
>>> ===
>>>
>>>
>>>
>>>
>>
>
>
>
>
>

Reply | Threaded
Open this post in threaded view
|

Re: Mac Carbon VM 3.8.15b5 pending

timrowledge
>
> #ifndef MSB_FIRST
>         #define RED_INDEX 0
>         #define GREEN_INDEX 1
>         #define BLUE_INDEX 2
>         #define ALPHA_INDEX 3
> #else
>         #define ALPHA_INDEX 0
>         #define BLUE_INDEX 1
>         #define GREEN_INDEX 2
>         #define RED_INDEX 3
> #endif
That's not a particularly sensible bit of code. Not all systems mix  
the r,g,b,a in the same order. Some probably don't support 'a'.


tim
--
tim Rowledge; [hidden email]; http://www.rowledge.org/tim
Strange OpCodes: ESC: Emulate Small Child



Reply | Threaded
Open this post in threaded view
|

Re: Mac Carbon VM 3.8.15b5 pending

johnmci
In reply to this post by Bruce O'Neel-5

On Mar 28, 2007, at 2:13 AM, Bruce O'Neel wrote:

> Hi,
>
> The only reason why I suspect a VM problem is that Windows, OpenBSD  
> big endian, Linux little endian, Mac PPC and Mac Intel running the  
> PPC VM all get the same display.  The only one which is different  
> is the Mac Intel running the universal binary.
>
> Is it possible that the problem is in  Squeak3D/b3dTypes.h?
>
> /* An ugly hack but I can't find the global defs in CodeWarrior on  
> the Mac */
> #ifndef LSB_FIRST
>         #define MSB_FIRST
> #endif


Ah, well an interesting question then would be what is LSB_FIRST, who  
defines it, where does it come from, what value should it have?
Why does it work on other platform compiles, and yes CodeWarrior  
likely has been dead 10 years now...  (well maybe less)


--
========================================================================
===
John M. McIntosh <[hidden email]>
Corporate Smalltalk Consulting Ltd.  http://www.smalltalkconsulting.com
========================================================================
===



Reply | Threaded
Open this post in threaded view
|

Re: Re: Mac Carbon VM 3.8.15b5 pending

Bruce O'Neel-5
Hi,

On the unix/linux build it seems to be defined by configure.  From the Makefile you have a line:

 CFLAGS=          -g -O2 -fomit-frame-pointer -DLSB_FIRST=1

This on an x86 linux system.  On a Solaris SPARC system -DLSB_FIRST is not there.

cheers

bruce

----- Message d'origine -----
De: John M McIntosh <[hidden email]>
Date: Wed, 28 Mar 2007 12:02:08 -0700
Sujet: Re: Mac Carbon VM 3.8.15b5  pending
À: "Bruce O'Neel" <[hidden email]>
Cc: The general-purpose Squeak developers list <[hidden email]>

>
>On Mar 28, 2007, at 2:13 AM, Bruce O'Neel wrote:
>
>> Hi,
>>
>> The only reason why I suspect a VM problem is that Windows, OpenBSD  
>> big endian, Linux little endian, Mac PPC and Mac Intel running the  
>> PPC VM all get the same display.  The only one which is different  
>> is the Mac Intel running the universal binary.
>>
>> Is it possible that the problem is in  Squeak3D/b3dTypes.h?
>>
>> /* An ugly hack but I can't find the global defs in CodeWarrior on  
>> the Mac */
>> #ifndef LSB_FIRST
>>         #define MSB_FIRST
>> #endif
>
>
>Ah, well an interesting question then would be what is LSB_FIRST, who  
>defines it, where does it come from, what value should it have?
>Why does it work on other platform compiles, and yes CodeWarrior  
>likely has been dead 10 years now...  (well maybe less)
>
>
>--
>========================================================================
>===
>John M. McIntosh <[hidden email]>
>Corporate Smalltalk Consulting Ltd.  http://www.smalltalkconsulting.com
>========================================================================
>===
>
>
>

Reply | Threaded
Open this post in threaded view
|

Re: Mac Carbon VM 3.8.15b5 pending

johnmci
Ok, in my idisk experimental folder try

Squeak 3.8.16beta5U.app

That contains the recompiled macintel vm that should show colors  
properly for Squeak3D surfaces

On Mar 29, 2007, at 2:37 AM, Bruce O'Neel wrote:

> -DLSB_FIRST=1

--
========================================================================
===
John M. McIntosh <[hidden email]>
Corporate Smalltalk Consulting Ltd.  http://www.smalltalkconsulting.com
========================================================================
===



Reply | Threaded
Open this post in threaded view
|

Re: Re: Mac Carbon VM 3.8.15b5 pending

Bruce O'Neel-5
Thank you Thank you Thank you!

It works perfectly.

Just out of interest, was my guess correct or was it something completely unrelated?

Thanks!

cheers

bruce

----- Message d'origine -----
De: John M McIntosh <[hidden email]>
Date: Fri, 30 Mar 2007 11:59:48 -0700
Sujet: Re: Mac Carbon VM 3.8.15b5  pending
À: "Bruce O'Neel" <[hidden email]>
Cc: The general-purpose Squeak developers list <[hidden email]>

>Ok, in my idisk experimental folder try
>
>Squeak 3.8.16beta5U.app
>
>That contains the recompiled macintel vm that should show colors  
>properly for Squeak3D surfaces
>
>On Mar 29, 2007, at 2:37 AM, Bruce O'Neel wrote:
>
>> -DLSB_FIRST=1
>
>--
>========================================================================
>===
>John M. McIntosh <[hidden email]>
>Corporate Smalltalk Consulting Ltd.  http://www.smalltalkconsulting.com
>========================================================================
>===
>
>
>

Reply | Threaded
Open this post in threaded view
|

Re: Mac Carbon VM 3.8.15b5 pending

johnmci
Yes, needed the -DLSB_FIRST=1 for macintel compile so the byte  
ordering was correct.

On Apr 2, 2007, at 12:10 AM, Bruce O'Neel wrote:

> Thank you Thank you Thank you!
>
> It works perfectly.
>
> Just out of interest, was my guess correct or was it something  
> completely unrelated?
>
> Thanks!
>
> cheers
>
> bruce
>
> ----- Message d'origine -----
> De: John M McIntosh <[hidden email]>
> Date: Fri, 30 Mar 2007 11:59:48 -0700
> Sujet: Re: Mac Carbon VM 3.8.15b5  pending
> À: "Bruce O'Neel" <[hidden email]>
> Cc: The general-purpose Squeak developers list <squeak-
> [hidden email]>
>
>> Ok, in my idisk experimental folder try
>>
>> Squeak 3.8.16beta5U.app
>>
>> That contains the recompiled macintel vm that should show colors
>> properly for Squeak3D surfaces
>>
>> On Mar 29, 2007, at 2:37 AM, Bruce O'Neel wrote:
>>
>>> -DLSB_FIRST=1
>>
>> --
>> =====================================================================
>> ===
>> ===
>> John M. McIntosh <[hidden email]>
>> Corporate Smalltalk Consulting Ltd.  http://
>> www.smalltalkconsulting.com
>> =====================================================================
>> ===
>> ===
>>
>>
>>

--
========================================================================
===
John M. McIntosh <[hidden email]>
Corporate Smalltalk Consulting Ltd.  http://www.smalltalkconsulting.com
========================================================================
===