Dear List,
While programming with Simon, we bumped into a sensible issue. Consider the two following methods:
SetupStorage>>selectAllWithType: aSmalltalkClass ^ self select: [ :each | "-- select returns an ArrayList" each isKindOf: aSmalltalkClass ]
RuntimeStorage>>selectAllWithType: aSmalltalkClass ^ byGroup at: aSmalltalkClass ifAbsent: [ OrderedCollection new ]
These two methods do not have the same behavior. We could have something starting with "(byGroup keys select: [:cls | cls isKindOf: aSmalltalkClass]) ..." But that would defeat the whole idea of having different storages...
Cheers, Alexandre
Hi Alex,
Why do you say that they do not have the same behavior? They both return a collection with all the instances of the given Class from the group.
Is just that the SetupStorage stores elements in an ArrayList, which makes it faster to add elements, while the RuntimeStorage has a more elaborate caching mechanism (for example byGroup) for retrieving elements faster.
Cheers, Doru
On 27 Nov 2009, at 18:39, Alexandre Bergel wrote:
Dear List,
While programming with Simon, we bumped into a sensible issue. Consider the two following methods:
SetupStorage>>selectAllWithType: aSmalltalkClass ^ self select: [ :each | "-- select returns an ArrayList" each isKindOf: aSmalltalkClass ]
RuntimeStorage>>selectAllWithType: aSmalltalkClass ^ byGroup at: aSmalltalkClass ifAbsent: [ OrderedCollection new ]
These two methods do not have the same behavior. We could have something starting with "(byGroup keys select: [:cls | cls isKindOf: aSmalltalkClass]) ..." But that would defeat the whole idea of having different storages...
Cheers, Alexandre -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- www.tudorgirba.com
"Be rather willing to give than demanding to get."
On 27 nov. 2009, at 17:06, Tudor Girba wrote:
Hi Alex,
Why do you say that they do not have the same behavior? They both return a collection with all the instances of the given Class from the group.
Alex forgot to explain the use case.
Take some elements of ClassA, some elements of ClassB, and both ClassA and ClassB are subclasses of ClassC.
Now ask aMooseModel allWithType: ClassC
It works with a SetupStorage which performs a #isKindOf: test
but does not work with a RuntimeStorage for which there is no byGroup at: ClassC Instead, you have one group ClassA -> {a1, a2} and one group ClassB -> {b1, b2}, which are initialized when filling up the model.
Now I havent really thought through this, but I guess it will be ok to do some lazy computation for such a case.
Is just that the SetupStorage stores elements in an ArrayList, which makes it faster to add elements, while the RuntimeStorage has a more elaborate caching mechanism (for example byGroup) for retrieving elements faster.
Cheers, Doru
On 27 Nov 2009, at 18:39, Alexandre Bergel wrote:
Dear List,
While programming with Simon, we bumped into a sensible issue. Consider the two following methods:
SetupStorage>>selectAllWithType: aSmalltalkClass ^ self select: [ :each | "-- select returns an ArrayList" each isKindOf: aSmalltalkClass ]
RuntimeStorage>>selectAllWithType: aSmalltalkClass ^ byGroup at: aSmalltalkClass ifAbsent: [ OrderedCollection new ]
These two methods do not have the same behavior. We could have something starting with "(byGroup keys select: [:cls | cls isKindOf: aSmalltalkClass]) ..." But that would defeat the whole idea of having different storages...
Cheers, Alexandre -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- www.tudorgirba.com
"Be rather willing to give than demanding to get."
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- Simon
Hi Simon,
Thanks for the explanations. Indeed, from this perspective it is not correct, but in this case, I will replace "each isKindOf:" with "each class =" because this is the intention.
We can add another method with allWithKind: that looks for the superclasses as well.
Cheers, Doru
On 27 Nov 2009, at 21:27, Simon Denier wrote:
On 27 nov. 2009, at 17:06, Tudor Girba wrote:
Hi Alex,
Why do you say that they do not have the same behavior? They both return a collection with all the instances of the given Class from the group.
Alex forgot to explain the use case.
Take some elements of ClassA, some elements of ClassB, and both ClassA and ClassB are subclasses of ClassC.
Now ask aMooseModel allWithType: ClassC
It works with a SetupStorage which performs a #isKindOf: test
but does not work with a RuntimeStorage for which there is no byGroup at: ClassC Instead, you have one group ClassA -> {a1, a2} and one group ClassB -
{b1, b2}, which are initialized when filling up the model.
Now I havent really thought through this, but I guess it will be ok to do some lazy computation for such a case.
Is just that the SetupStorage stores elements in an ArrayList, which makes it faster to add elements, while the RuntimeStorage has a more elaborate caching mechanism (for example byGroup) for retrieving elements faster.
Cheers, Doru
On 27 Nov 2009, at 18:39, Alexandre Bergel wrote:
Dear List,
While programming with Simon, we bumped into a sensible issue. Consider the two following methods:
SetupStorage>>selectAllWithType: aSmalltalkClass ^ self select: [ :each | "-- select returns an ArrayList" each isKindOf: aSmalltalkClass ]
RuntimeStorage>>selectAllWithType: aSmalltalkClass ^ byGroup at: aSmalltalkClass ifAbsent: [ OrderedCollection new ]
These two methods do not have the same behavior. We could have something starting with "(byGroup keys select: [:cls | cls isKindOf: aSmalltalkClass]) ..." But that would defeat the whole idea of having different storages...
Cheers, Alexandre -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- www.tudorgirba.com
"Be rather willing to give than demanding to get."
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- Simon
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- www.tudorgirba.com
"Problem solving should be concentrated on describing the problem in a way that is relevant for the solution."
On 27 nov. 2009, at 17:34, Tudor Girba wrote:
Hi Simon,
Thanks for the explanations. Indeed, from this perspective it is not correct, but in this case, I will replace "each isKindOf:" with "each class =" because this is the intention.
ok but no so ok. Be sure to run the tests because it may break a few things.
We can add another method with allWithKind: that looks for the superclasses as well.
Cheers, Doru
On 27 Nov 2009, at 21:27, Simon Denier wrote:
On 27 nov. 2009, at 17:06, Tudor Girba wrote:
Hi Alex,
Why do you say that they do not have the same behavior? They both return a collection with all the instances of the given Class from the group.
Alex forgot to explain the use case.
Take some elements of ClassA, some elements of ClassB, and both ClassA and ClassB are subclasses of ClassC.
Now ask aMooseModel allWithType: ClassC
It works with a SetupStorage which performs a #isKindOf: test
but does not work with a RuntimeStorage for which there is no byGroup at: ClassC Instead, you have one group ClassA -> {a1, a2} and one group ClassB -> {b1, b2}, which are initialized when filling up the model.
Now I havent really thought through this, but I guess it will be ok to do some lazy computation for such a case.
Is just that the SetupStorage stores elements in an ArrayList, which makes it faster to add elements, while the RuntimeStorage has a more elaborate caching mechanism (for example byGroup) for retrieving elements faster.
Cheers, Doru
On 27 Nov 2009, at 18:39, Alexandre Bergel wrote:
Dear List,
While programming with Simon, we bumped into a sensible issue. Consider the two following methods:
SetupStorage>>selectAllWithType: aSmalltalkClass ^ self select: [ :each | "-- select returns an ArrayList" each isKindOf: aSmalltalkClass ]
RuntimeStorage>>selectAllWithType: aSmalltalkClass ^ byGroup at: aSmalltalkClass ifAbsent: [ OrderedCollection new ]
These two methods do not have the same behavior. We could have something starting with "(byGroup keys select: [:cls | cls isKindOf: aSmalltalkClass]) ..." But that would defeat the whole idea of having different storages...
Cheers, Alexandre -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- www.tudorgirba.com
"Be rather willing to give than demanding to get."
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- Simon
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- www.tudorgirba.com
"Problem solving should be concentrated on describing the problem in a way that is relevant for the solution."
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- Simon
Hi,
I fixed the issue and added two tests for it. All the other tests seem to be fine.
Cheers, Doru
On 27 Nov 2009, at 23:57, Simon Denier wrote:
On 27 nov. 2009, at 17:34, Tudor Girba wrote:
Hi Simon,
Thanks for the explanations. Indeed, from this perspective it is not correct, but in this case, I will replace "each isKindOf:" with "each class =" because this is the intention.
ok but no so ok. Be sure to run the tests because it may break a few things.
We can add another method with allWithKind: that looks for the superclasses as well.
Cheers, Doru
On 27 Nov 2009, at 21:27, Simon Denier wrote:
On 27 nov. 2009, at 17:06, Tudor Girba wrote:
Hi Alex,
Why do you say that they do not have the same behavior? They both return a collection with all the instances of the given Class from the group.
Alex forgot to explain the use case.
Take some elements of ClassA, some elements of ClassB, and both ClassA and ClassB are subclasses of ClassC.
Now ask aMooseModel allWithType: ClassC
It works with a SetupStorage which performs a #isKindOf: test
but does not work with a RuntimeStorage for which there is no byGroup at: ClassC Instead, you have one group ClassA -> {a1, a2} and one group ClassB -> {b1, b2}, which are initialized when filling up the model.
Now I havent really thought through this, but I guess it will be ok to do some lazy computation for such a case.
Is just that the SetupStorage stores elements in an ArrayList, which makes it faster to add elements, while the RuntimeStorage has a more elaborate caching mechanism (for example byGroup) for retrieving elements faster.
Cheers, Doru
On 27 Nov 2009, at 18:39, Alexandre Bergel wrote:
Dear List,
While programming with Simon, we bumped into a sensible issue. Consider the two following methods:
SetupStorage>>selectAllWithType: aSmalltalkClass ^ self select: [ :each | "-- select returns an ArrayList" each isKindOf: aSmalltalkClass ]
RuntimeStorage>>selectAllWithType: aSmalltalkClass ^ byGroup at: aSmalltalkClass ifAbsent: [ OrderedCollection new ]
These two methods do not have the same behavior. We could have something starting with "(byGroup keys select: [:cls | cls isKindOf: aSmalltalkClass]) ..." But that would defeat the whole idea of having different storages...
Cheers, Alexandre -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- www.tudorgirba.com
"Be rather willing to give than demanding to get."
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- Simon
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- www.tudorgirba.com
"Problem solving should be concentrated on describing the problem in a way that is relevant for the solution."
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- Simon
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- www.tudorgirba.com
"Every now and then stop and ask yourself if the war you're fighting is the right one."
Thanks Doru,
Alexandre
On 28 Nov 2009, at 13:55, Tudor Girba wrote:
Hi,
I fixed the issue and added two tests for it. All the other tests seem to be fine.
Cheers, Doru
On 27 Nov 2009, at 23:57, Simon Denier wrote:
On 27 nov. 2009, at 17:34, Tudor Girba wrote:
Hi Simon,
Thanks for the explanations. Indeed, from this perspective it is not correct, but in this case, I will replace "each isKindOf:" with "each class =" because this is the intention.
ok but no so ok. Be sure to run the tests because it may break a few things.
We can add another method with allWithKind: that looks for the superclasses as well.
Cheers, Doru
On 27 Nov 2009, at 21:27, Simon Denier wrote:
On 27 nov. 2009, at 17:06, Tudor Girba wrote:
Hi Alex,
Why do you say that they do not have the same behavior? They both return a collection with all the instances of the given Class from the group.
Alex forgot to explain the use case.
Take some elements of ClassA, some elements of ClassB, and both ClassA and ClassB are subclasses of ClassC.
Now ask aMooseModel allWithType: ClassC
It works with a SetupStorage which performs a #isKindOf: test
but does not work with a RuntimeStorage for which there is no byGroup at: ClassC Instead, you have one group ClassA -> {a1, a2} and one group ClassB -> {b1, b2}, which are initialized when filling up the model.
Now I havent really thought through this, but I guess it will be ok to do some lazy computation for such a case.
Is just that the SetupStorage stores elements in an ArrayList, which makes it faster to add elements, while the RuntimeStorage has a more elaborate caching mechanism (for example byGroup) for retrieving elements faster.
Cheers, Doru
On 27 Nov 2009, at 18:39, Alexandre Bergel wrote:
Dear List,
While programming with Simon, we bumped into a sensible issue. Consider the two following methods:
SetupStorage>>selectAllWithType: aSmalltalkClass ^ self select: [ :each | "-- select returns an ArrayList" each isKindOf: aSmalltalkClass ]
RuntimeStorage>>selectAllWithType: aSmalltalkClass ^ byGroup at: aSmalltalkClass ifAbsent: [ OrderedCollection new ]
These two methods do not have the same behavior. We could have something starting with "(byGroup keys select: [:cls | cls isKindOf: aSmalltalkClass]) ..." But that would defeat the whole idea of having different storages...
Cheers, Alexandre -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- www.tudorgirba.com
"Be rather willing to give than demanding to get."
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- Simon
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- www.tudorgirba.com
"Problem solving should be concentrated on describing the problem in a way that is relevant for the solution."
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- Simon
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-- www.tudorgirba.com
"Every now and then stop and ask yourself if the war you're fighting is the right one."
Moose-dev mailing list Moose-dev@iam.unibe.ch https://www.iam.unibe.ch/mailman/listinfo/moose-dev
-
Alexandre Bergel -
Simon Denier -
Tudor Girba