[Pharo-dev] Re: Confusing Refactoring Pull Up dialog

2025-06-28 Thread stephane ducasse
HI daniel You are right. Now a bit more context of our current work. We are working on this. The point is that the old refactoring code hardcoded in precondition the possible actions in an non user friendly way. Now in Pharo 13 and 14 we have introduced Drivers. A drivers is - using

[Pharo-dev] Re: Does not understand message inconsistency

2025-06-28 Thread Daniel Slomovits
There is a bit of confusion around the word "understand" here. It is sensible to think of an object as "understanding" a message or not, but in fact the selector that tells you whether an object understand a message is #respondsTo:, while the equivalent on the class which tells you whether instance

[Pharo-dev] Re: Confusing Refactoring Pull Up dialog

2025-06-28 Thread Daniel Slomovits
Yes, I've run into this before. I think the reasoning is that, if you do not copy down the superclass implementation, this is not truly a refactoring—it will change the behavior of a selector that was already understood by the siblings. But I agree that in many cases, this change of behavior is the

[Pharo-dev] Confusing Refactoring Pull Up dialog

2025-06-28 Thread Aik-Siong Koh
There are two questions, but only response accepted. I just want a pull up with nothing extra done. If I click yes, I get the extra methods. If I click no, nothing at all happens. Please advise, Thanks, Aik-Siong Koh

[Pharo-dev] Re: Variable references, readers, writers

2025-06-28 Thread Marcus Denker
This is the code that uses assignmentNodes: ((Point slotNamed: #x) assignmentNodes collect: [ :each | each methodNode compiledMethod ] ) asSet But as a downside, it has to create the AST. So maybe better is to use the API of the Variable: Point methods select: [ :each | (Point