Re: [Pharo-project] PseudoContext?
On Jan 29, 2012, at 6:02 PM, Stéphane Ducasse wrote:
Hi
I have the impression that PseudoContext is not used anymore. At least compact class array and special object array do not use it anymore. So what is the use for this class?
I think it was only used by Ian's original Jitter... which means 10 years unused. We should remove it. Marcus -- Marcus Denker -- http://marcusdenker.de
I don't see any PseudoContext in squeak trunk, you can reasonably not afford to be dirtier ;) 2012/1/29 Marcus Denker <marcus.denker@inria.fr>:
On Jan 29, 2012, at 6:02 PM, Stéphane Ducasse wrote:
Hi
I have the impression that PseudoContext is not used anymore. At least compact class array and special object array do not use it anymore. So what is the use for this class?
I think it was only used by Ian's original Jitter... which means 10 years unused. We should remove it.
    Marcus
-- Marcus Denker -- http://marcusdenker.de
On Sun, Jan 29, 2012 at 10:53 PM, Nicolas Cellier < nicolas.cellier.aka.nice@gmail.com> wrote:
I don't see any PseudoContext in squeak trunk, you can reasonably not afford to be dirtier ;)
YES!! And this class got me crazy once because it subclass from nil !! BTW, we should also fix the senders of #isPseudoContext if we remoev the class.
2012/1/29 Marcus Denker <marcus.denker@inria.fr>:
On Jan 29, 2012, at 6:02 PM, Stéphane Ducasse wrote:
Hi
I have the impression that PseudoContext is not used anymore. At least
compact class array and special object array
do not use it anymore. So what is the use for this class?
I think it was only used by Ian's original Jitter... which means 10 years unused. We should remove it.
Marcus
-- Marcus Denker -- http://marcusdenker.de
-- Mariano http://marianopeck.wordpress.com
I will with pleasureâ¦. Stef On Jan 30, 2012, at 9:50 AM, Mariano Martinez Peck wrote:
On Sun, Jan 29, 2012 at 10:53 PM, Nicolas Cellier <nicolas.cellier.aka.nice@gmail.com> wrote: I don't see any PseudoContext in squeak trunk, you can reasonably not afford to be dirtier ;)
YES!! And this class got me crazy once because it subclass from nil !! BTW, we should also fix the senders of #isPseudoContext if we remoev the class.
2012/1/29 Marcus Denker <marcus.denker@inria.fr>:
On Jan 29, 2012, at 6:02 PM, Stéphane Ducasse wrote:
Hi
I have the impression that PseudoContext is not used anymore. At least compact class array and special object array do not use it anymore. So what is the use for this class?
I think it was only used by Ian's original Jitter... which means 10 years unused. We should remove it.
Marcus
-- Marcus Denker -- http://marcusdenker.de
-- Mariano http://marianopeck.wordpress.com
On Mon, Jan 30, 2012 at 2:39 AM, Stéphane Ducasse <stephane.ducasse@inria.fr
wrote:
I will with pleasureâ¦.
Also think of getting rid of TranslatedMethod and BlockContext if you haven't already. Also, collapsing ContextPart and MethodContext onto just Context (or ActivationContext (?)) would be nice.
Stef
On Jan 30, 2012, at 9:50 AM, Mariano Martinez Peck wrote:
On Sun, Jan 29, 2012 at 10:53 PM, Nicolas Cellier <
nicolas.cellier.aka.nice@gmail.com> wrote:
I don't see any PseudoContext in squeak trunk, you can reasonably not afford to be dirtier ;)
YES!! And this class got me crazy once because it subclass from nil !! BTW, we should also fix the senders of #isPseudoContext if we remoev the class.
2012/1/29 Marcus Denker <marcus.denker@inria.fr>:
On Jan 29, 2012, at 6:02 PM, Stéphane Ducasse wrote:
Hi
I have the impression that PseudoContext is not used anymore. At
least compact class array and special object array
do not use it anymore. So what is the use for this class?
I think it was only used by Ian's original Jitter... which means 10 years unused. We should remove it.
Marcus
-- Marcus Denker -- http://marcusdenker.de
-- Mariano http://marianopeck.wordpress.com
-- best, Eliot
Also think of getting rid of TranslatedMethod
yes we already did that one
and BlockContext if you haven't already.
it is deprecated
Also, collapsing ContextPart and MethodContext onto just Context (or ActivationContext (?)) would be nice.
Thanks for the suggestion I will do it. What is the best choice to avoid incompatibility? Having MethodContext I guess. Stef
On Mon, Jan 30, 2012 at 11:47 AM, Stéphane Ducasse < stephane.ducasse@inria.fr> wrote:
Also think of getting rid of TranslatedMethod
yes we already did that one
and BlockContext if you haven't already.
it is deprecated
Also, collapsing ContextPart and MethodContext onto just Context (or ActivationContext (?)) would be nice.
Thanks for the suggestion I will do it. What is the best choice to avoid incompatibility? Having MethodContext I guess.
Yes. But it is perhaps worth making the change incompatible to say goodbye to BlockContext? One can always do Smalltalk at: #MethodContext put: ActivationContext to get older packages to load. I don't know what's best, but for once I like the clean-up :)
Stef
-- best, Eliot
Also, collapsing ContextPart and MethodContext onto just Context (or ActivationContext (?)) would be nice.
Thanks for the suggestion I will do it. What is the best choice to avoid incompatibility? Having MethodContext I guess.
Yes. But it is perhaps worth making the change incompatible to say goodbye to BlockContext? One can always do Smalltalk at: #MethodContext put: ActivationContext to get older packages to load. I don't know what's best, but for once I like the clean-up :)
I see :) I was more concerned about the fact that VM people may rely on MethodContext (name) while ContextPart as a superclass should not be used (or has a lower probably to be used). Stef
On Mon, Jan 30, 2012 at 1:21 PM, Stéphane Ducasse <stephane.ducasse@inria.fr
wrote:
Also, collapsing ContextPart and MethodContext onto just Context (or ActivationContext (?)) would be nice.
Thanks for the suggestion I will do it. What is the best choice to avoid incompatibility? Having MethodContext I guess.
Yes. But it is perhaps worth making the change incompatible to say goodbye to BlockContext? One can always do Smalltalk at: #MethodContext put: ActivationContext to get older packages to load. I don't know what's best, but for once I like the clean-up :)
I see :)
I was more concerned about the fact that VM people may rely on MethodContext (name) while ContextPart as a superclass should not be used (or has a lower probably to be used).
Right. My point is that MethodContext is now a bad name since MethodContext represents both method and block activations.
Stef
-- best, Eliot
Indeed ActivationContext is nicer. I will use that one :) Stef
On Mon, Jan 30, 2012 at 1:21 PM, Stéphane Ducasse <stephane.ducasse@inria.fr> wrote:
Also, collapsing ContextPart and MethodContext onto just Context (or ActivationContext (?)) would be nice.
Thanks for the suggestion I will do it. What is the best choice to avoid incompatibility? Having MethodContext I guess.
Yes. But it is perhaps worth making the change incompatible to say goodbye to BlockContext? One can always do Smalltalk at: #MethodContext put: ActivationContext to get older packages to load. I don't know what's best, but for once I like the clean-up :)
I see :)
I was more concerned about the fact that VM people may rely on MethodContext (name) while ContextPart as a superclass should not be used (or has a lower probably to be used).
Right. My point is that MethodContext is now a bad name since MethodContext represents both method and block activations.
Stef
-- best, Eliot
participants (5)
-
Eliot Miranda -
Marcus Denker -
Mariano Martinez Peck -
Nicolas Cellier -
Stéphane Ducasse