[Pharo-project] [updates] 10150-10152
10150 ------ -> Issue 333: A fix for animated GIF harvested from Squeak -> Mantis 0006343: Process does not need to be a special object Change proposed by martin v. Loewis: When changing Process, I was warned that it is a special object and therefore it is dangerous to change its class definition. Process currently lives at: 28 of Smalltalk specialObjectsArray. In reviewing all references to the special objects array, I could not find any reference to that slot, neither in the Squeak 3.9 image nor in the VM. 10151 ------ -> Issue 146: ClassDescription>>copyUnobstrusevely ClassDescription>>copyUnobstrusevely and friends are only invoked by kedama which should be removed -> Issue 128: M5711: TextFieldMorph changes text position when emptied -> now really do Postscript: Smalltalk recreateSpecialObjectsArray. 10152 ----- -> Issue 276: candidate for core: Process-specific variables Martin von Loewis did a nice implementation of thread-specific variables: This project provides process-specific variables (called TLS - thread- local storage) in other environments. Two kinds of variables are provided: ProcessLocalVariable (similar to TLS), and DynamicVariables (akin LISP's dynamic scope) -- Marcus Denker -- denker@iam.unibe.ch http://www.iam.unibe.ch/~denker
2008/11/16 Marcus Denker <denker@iam.unibe.ch>:
10150 ------ -> Issue 333: A fix for animated GIF harvested from Squeak -> Mantis 0006343: Process does not need to be a special object Change proposed by martin v. Loewis: When changing Process, I was warned that it is a special object and therefore it is dangerous to change its class definition. Process currently lives at: 28 of Smalltalk specialObjectsArray. In reviewing all references to the special objects array, I could not find any reference to that slot, neither in the Squeak 3.9 image nor in the VM.
Its not referencing the process class directly, however it assuming that Process having following slots: "Class Process" SuspendedContextIndex := 1. PriorityIndex := 2. MyListIndex := 3 So, changing the Process class definition may lead to unpredictable behavior in VM.
10151 ------ -> Issue 146: ClassDescription>>copyUnobstrusevely ClassDescription>>copyUnobstrusevely and friends are only invoked by kedama which should be removed -> Issue 128: M5711: TextFieldMorph changes text position when emptied -> now really do Postscript: Smalltalk recreateSpecialObjectsArray.
10152 ----- -> Issue 276: candidate for core: Process-specific variables Martin von Loewis did a nice implementation of thread-specific variables: This project provides process-specific variables (called TLS - thread-local storage) in other environments. Two kinds of variables are provided: ProcessLocalVariable (similar to TLS), and DynamicVariables (akin LISP's dynamic scope)
-- Marcus Denker -- denker@iam.unibe.ch http://www.iam.unibe.ch/~denker
_______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
-- Best regards, Igor Stasenko AKA sig.
marcus could you contact martin to ask him to sign the agreement? Or do we consider that he published that on mantis and that as such it is MIT which is true. Stef On Nov 16, 2008, at 6:05 PM, Marcus Denker wrote:
10150 ------ -> Issue 333: A fix for animated GIF harvested from Squeak -> Mantis 0006343: Process does not need to be a special object Change proposed by martin v. Loewis: When changing Process, I was warned that it is a special object and therefore it is dangerous to change its class definition. Process currently lives at: 28 of Smalltalk specialObjectsArray. In reviewing all references to the special objects array, I could not find any reference to that slot, neither in the Squeak 3.9 image nor in the VM.
10151 ------ -> Issue 146: ClassDescription>>copyUnobstrusevely ClassDescription>>copyUnobstrusevely and friends are only invoked by kedama which should be removed -> Issue 128: M5711: TextFieldMorph changes text position when emptied -> now really do Postscript: Smalltalk recreateSpecialObjectsArray.
10152 ----- -> Issue 276: candidate for core: Process-specific variables Martin von Loewis did a nice implementation of thread-specific variables: This project provides process-specific variables (called TLS - thread-local storage) in other environments. Two kinds of variables are provided: ProcessLocalVariable (similar to TLS), and DynamicVariables (akin LISP's dynamic scope)
-- Marcus Denker -- denker@iam.unibe.ch http://www.iam.unibe.ch/~denker
_______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
On 16.11.2008, at 19:43, Igor Stasenko wrote:
Its not referencing the process class directly, however it assuming that Process having following slots: "Class Process" SuspendedContextIndex := 1. PriorityIndex := 2. MyListIndex := 3
So, changing the Process class definition may lead to unpredictable behavior in VM.
Hmmm... ok, yes. So maybe we can leave it in the specialObjectarray, but put a comment that exactly says this... Ok, will do that. Marcus -- Marcus Denker -- denker@iam.unibe.ch http://www.iam.unibe.ch/~denker
On 16.11.2008, at 19:44, Stéphane Ducasse wrote:
marcus
could you contact martin to ask him to sign the agreement?
I will ask him. Marcus
Or do we consider that he published that on mantis and that as such it is MIT which is true.
Stef
On Nov 16, 2008, at 6:05 PM, Marcus Denker wrote:
10150 ------ -> Issue 333: A fix for animated GIF harvested from Squeak -> Mantis 0006343: Process does not need to be a special object Change proposed by martin v. Loewis: When changing Process, I was warned that it is a special object and therefore it is dangerous to change its class definition. Process currently lives at: 28 of Smalltalk specialObjectsArray. In reviewing all references to the special objects array, I could not find any reference to that slot, neither in the Squeak 3.9 image nor in the VM.
10151 ------ -> Issue 146: ClassDescription>>copyUnobstrusevely ClassDescription>>copyUnobstrusevely and friends are only invoked by kedama which should be removed -> Issue 128: M5711: TextFieldMorph changes text position when emptied -> now really do Postscript: Smalltalk recreateSpecialObjectsArray.
10152 ----- -> Issue 276: candidate for core: Process-specific variables Martin von Loewis did a nice implementation of thread-specific variables: This project provides process-specific variables (called TLS - thread-local storage) in other environments. Two kinds of variables are provided: ProcessLocalVariable (similar to TLS), and DynamicVariables (akin LISP's dynamic scope)
-- Marcus Denker -- denker@iam.unibe.ch http://www.iam.unibe.ch/~denker
_______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
_______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
-- Marcus Denker -- denker@iam.unibe.ch http://www.iam.unibe.ch/~denker
This is really showing that a cool classBuilder would really encode such constraint. Stef On Nov 16, 2008, at 7:57 PM, Marcus Denker wrote:
On 16.11.2008, at 19:43, Igor Stasenko wrote:
Its not referencing the process class directly, however it assuming that Process having following slots: "Class Process" SuspendedContextIndex := 1. PriorityIndex := 2. MyListIndex := 3
So, changing the Process class definition may lead to unpredictable behavior in VM.
Hmmm... ok, yes. So maybe we can leave it in the specialObjectarray, but put a comment that exactly says this...
Ok, will do that.
Marcus
-- Marcus Denker -- denker@iam.unibe.ch http://www.iam.unibe.ch/~denker
_______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
participants (3)
-
Igor Stasenko -
Marcus Denker -
Stéphane Ducasse