27 Jul
2007
27 Jul
'07
5:36 p.m.
Hey Sara,
the issue is fixed, now you can say
SCG.Moose asMetaDescription classes
and get all meta-descriptions of all classes without the annoying
initialize loop.
PS: next week I am in Berlin, I might not have time to fix bugs, but
I am usually available via mail.
cheers,
AA
On 27 Jul 2007, at 15:32 , sellossa(a)ensieta.fr wrote:
hello,
why, when I do: SCG.Moose asMetaDescription, I get sthing different
from when I do: SCG.Moose allClasses do: #asMetaDescription.
^SCG.Moose asMetaDescription
Stef find it terrible!
sara
13 Aug
13 Aug
3:40 p.m.
New subject: metamodel from Smalltalk Code
Hi Adrian
This is stef typing. I'm looking at the code of sara and I see the
following behavior:
generateEcoreString:aNamespace
|package |
"aNamespace asMetaDescription classes."
package:=aNamespace asMetaDescription.
package acceptMetamodelVisitor: self.
^ rootXmlElement printString.
If we remove the comments the code works, but if we run it like it is
ie package is empty (otherwise it contains the right meta descriptions).
I think that the design is not good. The programmer should not guess that
a query is doing something magic in his back. Could you change that?
At the minimum call classes initializeSomething
Stef
Hey Sara,
the issue is fixed, now you can say
SCG.Moose asMetaDescription classes
and get all meta-descriptions of all classes without the annoying
initialize loop.
PS: next week I am in Berlin, I might not have time to fix bugs, but
I am usually available via mail.
cheers,
AA
On 27 Jul 2007, at 15:32 , sellossa(a)ensieta.fr wrote:
hello,
why, when I do: SCG.Moose asMetaDescription, I get sthing different
from when I do: SCG.Moose allClasses do: #asMetaDescription.
^SCG.Moose asMetaDescription
Stef find it terrible!
sara
14 Aug
14 Aug
7:29 p.m.
New subject: metamodel from Smalltalk Code
Hey Stef and Sara,
sound like a heisenbug :) how do you check if the package is empty or
not later on? because the preferred way to do that is by calling
#classes, which would initialize the package.
AA
On 13 Aug 2007, at 15:40 , sellossa(a)ensieta.fr wrote:
Hi Adrian
This is stef typing. I'm looking at the code of sara and I see the
following behavior:
generateEcoreString:aNamespace
|package |
"aNamespace asMetaDescription classes."
package:=aNamespace asMetaDescription.
package acceptMetamodelVisitor: self.
^ rootXmlElement printString.
If we remove the comments the code works, but if we run it like it is
ie package is empty (otherwise it contains the right meta
descriptions).
I think that the design is not good. The programmer should not
guess that
a query is doing something magic in his back. Could you change that?
At the minimum call classes initializeSomething
Stef
Hey Sara,
the issue is fixed, now you can say
SCG.Moose asMetaDescription classes
and get all meta-descriptions of all classes without the annoying
initialize loop.
PS: next week I am in Berlin, I might not have time to fix bugs, but
I am usually available via mail.
cheers,
AA
On 27 Jul 2007, at 15:32 , sellossa(a)ensieta.fr wrote:
_______________________________________________
Moose-dev mailing list
Moose-dev(a)iam.unibe.ch
https://www.iam.unibe.ch/mailman/listinfo/moose-dev hello,
why, when I do: SCG.Moose asMetaDescription, I get sthing different
from when I do: SCG.Moose allClasses do: #asMetaDescription.
^SCG.Moose asMetaDescription
Stef find it terrible!
sara
9 p.m.
New subject: metamodel from Smalltalk Code
Hi Adrian!
I get something weird when I use it for SCG.Moose.
When I do:
SCG.Moose asMetaDescription classes.
SCG.Moose asMetaDescription inspect.
Hey Stef and Sara,
sound like a heisenbug :) how do you check if the package is empty or
not later on? because the preferred way to do that is by calling
#classes, which would initialize the package.
AA
On 13 Aug 2007, at 15:40 , sellossa(a)ensieta.fr wrote:
Hi Adrian
This is stef typing. I'm looking at the code of sara and I see the
following behavior:
generateEcoreString:aNamespace
|package |
"aNamespace asMetaDescription classes."
package:=aNamespace asMetaDescription.
package acceptMetamodelVisitor: self.
^ rootXmlElement printString.
If we remove the comments the code works, but if we run it like it is
ie package is empty (otherwise it contains the right meta
descriptions).
I think that the design is not good. The programmer should not
guess that
a query is doing something magic in his back. Could you change that?
At the minimum call classes initializeSomething
Stef
Hey Sara,
the issue is fixed, now you can say
SCG.Moose asMetaDescription classes
and get all meta-descriptions of all classes without the annoying
initialize loop.
PS: next week I am in Berlin, I might not have time to fix bugs, but
I am usually available via mail.
cheers,
AA
On 27 Jul 2007, at 15:32 , sellossa(a)ensieta.fr wrote:
_______________________________________________
Moose-dev mailing list
Moose-dev(a)iam.unibe.ch
https://www.iam.unibe.ch/mailman/listinfo/moose-dev hello,
why, when I do: SCG.Moose asMetaDescription, I get sthing different
from when I do: SCG.Moose allClasses do: #asMetaDescription.
^SCG.Moose asMetaDescription
Stef find it terrible!
sara
9:04 p.m.
New subject: metamodel from Smalltalk Code
Hi Adrian!
I get something weird when I use it for SCG.Moose.
When I do:
|package|
SCG.Moose asMetaDescription classes.
package:=SCG.Moose asMetaDescription inspect.
package doesn't contain the entites FAMIXClass, FAMIXAttribute and so on...
However, FAMIX entites are defined in the namespace SCG.Moose, isn't it?
Thanks you for your help
Sara
Hey Stef and Sara,
sound like a heisenbug :) how do you check if the package is empty or
not later on? because the preferred way to do that is by calling
#classes, which would initialize the package.
AA
On 13 Aug 2007, at 15:40 , sellossa(a)ensieta.fr wrote:
Hi Adrian
This is stef typing. I'm looking at the code of sara and I see the
following behavior:
generateEcoreString:aNamespace
|package |
"aNamespace asMetaDescription classes."
package:=aNamespace asMetaDescription.
package acceptMetamodelVisitor: self.
^ rootXmlElement printString.
If we remove the comments the code works, but if we run it like it is
ie package is empty (otherwise it contains the right meta
descriptions).
I think that the design is not good. The programmer should not
guess that
a query is doing something magic in his back. Could you change that?
At the minimum call classes initializeSomething
Stef
Hey Sara,
the issue is fixed, now you can say
SCG.Moose asMetaDescription classes
and get all meta-descriptions of all classes without the annoying
initialize loop.
PS: next week I am in Berlin, I might not have time to fix bugs, but
I am usually available via mail.
cheers,
AA
On 27 Jul 2007, at 15:32 , sellossa(a)ensieta.fr wrote:
> hello,
>
> why, when I do: SCG.Moose asMetaDescription, I get sthing different
> from when I do: SCG.Moose allClasses do: #asMetaDescription.
> ^SCG.Moose asMetaDescription
>
> Stef find it terrible!
>
> sara
_______________________________________________
Moose-dev mailing list
Moose-dev(a)iam.unibe.ch
https://www.iam.unibe.ch/mailman/listinfo/moose-dev 
15 Aug
15 Aug
9:31 p.m.
New subject: metamodel from Smalltalk Code
On 14 août 07, at 19:29, Adrian Kuhn wrote:
Hey Stef and Sara,
sound like a heisenbug :) how do you check if the package is empty or
not later on? because the preferred way to do that is by calling
#classes, which would initialize the package.
Sorry I do not understand what you mean.
My point is that either you need to call classes and you use the resutl
cl := aNamespace asMetaDescription classes.
or you do not need it
or you change the selector name
> generateEcoreString:aNamespace
> |package |
> aNamespace initializeMetaDescription.
> package:=aNamespace asMetaDescription.
> package acceptMetamodelVisitor: self.
> ^ rootXmlElement printString.
Is my point clearer.
Stef
AA
On 13 Aug 2007, at 15:40 , sellossa(a)ensieta.fr wrote:
Hi Adrian
This is stef typing. I'm looking at the code of sara and I see the
following behavior:
generateEcoreString:aNamespace
|package |
"aNamespace asMetaDescription classes."
package:=aNamespace asMetaDescription.
package acceptMetamodelVisitor: self.
^ rootXmlElement printString.
If we remove the comments the code works, but if we run it like it is
ie package is empty (otherwise it contains the right meta
descriptions).
I think that the design is not good. The programmer should not
guess that
a query is doing something magic in his back. Could you change that?
At the minimum call classes initializeSomething
Stef
_______________________________________________
Moose-dev mailing list
Moose-dev(a)iam.unibe.ch
https://www.iam.unibe.ch/mailman/listinfo/moose-dev Hey Sara,
the issue is fixed, now you can say
SCG.Moose asMetaDescription classes
and get all meta-descriptions of all classes without the annoying
initialize loop.
PS: next week I am in Berlin, I might not have time to fix bugs, but
I am usually available via mail.
cheers,
AA
On 27 Jul 2007, at 15:32 , sellossa(a)ensieta.fr wrote:
_______________________________________________
Moose-dev mailing list
Moose-dev(a)iam.unibe.ch
https://www.iam.unibe.ch/mailman/listinfo/moose-dev hello,
why, when I do: SCG.Moose asMetaDescription, I get sthing different
from when I do: SCG.Moose allClasses do: #asMetaDescription.
^SCG.Moose asMetaDescription
Stef find it terrible!
sara
6191
days inactive
6210
days old
5 comments
3 participants
participants (3)
-
Adrian Kuhn -
sellossa＠ensieta.fr -
Stéphane Ducasse