Some references to transcript should be changed into File stdout
Hi I have the impression that some references to Transcript should be rewritten as File stdout For example DummyUIManager>>unhandledErrorDefaultAction: anException Transcript show: '*** EXCEPTION ***'; cr. Transcript show: anException; cr. Transcript show: (thisContext stack first: (20 min: thisContext stack size)); cr. anException isResumable ifTrue: [ anException resume ] Stef
NonInteractiveTranscript is usually the class then, so no need to File stdout anything I guess. NonInteractiveTranscript install is "de rigueur" for CLI style apps. Phil On Wed, Sep 23, 2015 at 1:16 PM, stepharo <stepharo@free.fr> wrote:
Hi
I have the impression that some references to Transcript should be rewritten as File stdout
For example
DummyUIManager>>unhandledErrorDefaultAction: anException
Transcript show: '*** EXCEPTION ***'; cr. Transcript show: anException; cr. Transcript show: (thisContext stack first: (20 min: thisContext stack size)); cr. anException isResumable ifTrue: [ anException resume ]
Stef
Hi Stef, Transcript in Smalltalk-80 systems and in Squeak and earlier Pharo versions is a variable. So when the system is running headless or on the command line Transcript can be assigned something that writes to stdout (eg an instance of NonInteractiveTranscript). This is nice and simple. Eliminating the variable and replacing it by a class might be a mistake because it obscures the idea that Transcript may be different in different system states. _,,,^..^,,,_ (phone)
On Sep 23, 2015, at 4:16 AM, stepharo <stepharo@free.fr> wrote:
Hi
I have the impression that some references to Transcript should be rewritten as File stdout
For example
DummyUIManager>>unhandledErrorDefaultAction: anException
Transcript show: '*** EXCEPTION ***'; cr. Transcript show: anException; cr. Transcript show: (thisContext stack first: (20 min: thisContext stack size)); cr. anException isResumable ifTrue: [ anException resume ]
Stef
On 24 Sep 2015, at 15:38, Eliot Miranda <eliot.miranda@gmail.com> wrote:
Hi Stef,
Transcript in Smalltalk-80 systems and in Squeak and earlier Pharo versions is a variable. So when the system is running headless or on the command line Transcript can be assigned something that writes to stdout (eg an instance of NonInteractiveTranscript). This is nice and simple. Eliminating the variable and replacing it by a class might be a mistake because it obscures the idea that Transcript may be different in different system states.
Exactly, and that is what we have with NonInteractiveTranscript (and in Pharo 5, DailyNonInteractiveTranscript and TTranscript).
_,,,^..^,,,_ (phone)
On Sep 23, 2015, at 4:16 AM, stepharo <stepharo@free.fr> wrote:
Hi
I have the impression that some references to Transcript should be rewritten as File stdout
For example
DummyUIManager>>unhandledErrorDefaultAction: anException
Transcript show: '*** EXCEPTION ***'; cr. Transcript show: anException; cr. Transcript show: (thisContext stack first: (20 min: thisContext stack size)); cr. anException isResumable ifTrue: [ anException resume ]
Stef
Guys may be you do not understand or I'm too stupid. My point is what are the preconditions (tools loaded in the system) that DummyUIManager should have. I thought that DummyUIManager (which should not be morphic else I do not understand what is MorphicUIManager) was for a system without a Morphic UI. So how it is possible that a system without UI could report error using a UI tool? Stef Le 23/9/15 11:16, stepharo a écrit :
Hi
I have the impression that some references to Transcript should be rewritten as File stdout
For example
DummyUIManager>>unhandledErrorDefaultAction: anException
Transcript show: '*** EXCEPTION ***'; cr. Transcript show: anException; cr. Transcript show: (thisContext stack first: (20 min: thisContext stack size)); cr. anException isResumable ifTrue: [ anException resume ]
Stef
Was teaching 8 hours per day this week and dead. Ok I read the other answers. I do not like this ide of global variable. Le 24/9/15 18:45, stepharo a écrit :
Guys
may be you do not understand or I'm too stupid. My point is what are the preconditions (tools loaded in the system) that DummyUIManager should have. I thought that DummyUIManager (which should not be morphic else I do not understand what is MorphicUIManager) was for a system without a Morphic UI. So how it is possible that a system without UI could report error using a UI tool?
Stef
Le 23/9/15 11:16, stepharo a écrit :
Hi
I have the impression that some references to Transcript should be rewritten as File stdout
For example
DummyUIManager>>unhandledErrorDefaultAction: anException
Transcript show: '*** EXCEPTION ***'; cr. Transcript show: anException; cr. Transcript show: (thisContext stack first: (20 min: thisContext stack size)); cr. anException isResumable ifTrue: [ anException resume ]
Stef
Hu Stef, On Thu, Sep 24, 2015 at 11:45 AM, stepharo <stepharo@free.fr> wrote:
Guys
may be you do not understand or I'm too stupid. My point is what are the preconditions (tools loaded in the system) that DummyUIManager should have. I thought that DummyUIManager (which should not be morphic else I do not understand what is MorphicUIManager) was for a system without a Morphic UI. So how it is possible that a system without UI could report error using a UI tool?
1. The object referred to by the Transcript variable *is not necessarily* a UI tool. It may be headless and have no UI component; it may send output ti stdout or a file or... There is no assumption in DummyUIManager's use that Transcript is a UI component. 2. In testing DummyUIManager it may indeed be useful to have DummyUIManager output to a GUI transcript window. HTH
Stef
Le 23/9/15 11:16, stepharo a écrit :
Hi
I have the impression that some references to Transcript should be rewritten as File stdout
For example
DummyUIManager>>unhandledErrorDefaultAction: anException
Transcript show: '*** EXCEPTION ***'; cr. Transcript show: anException; cr. Transcript show: (thisContext stack first: (20 min: thisContext stack size)); cr. anException isResumable ifTrue: [ anException resume ]
Stef
-- _,,,^..^,,,_ best, Eliot
participants (4)
-
Eliot Miranda -
phil＠highoctane.be -
stepharo -
Sven Van Caekenberghe