Squeak-dev July 2004

squeak-dev@lists.squeakfoundation.org

146 participants
464 discussions

Re: [ENH] SetInspector (23112)
by Anthony Hannan 05 Aug '04

05 Aug '04

Hi Andrew, Sorry for the late response. Your changeset looks good. It completes the pattern I started with inspectorClass, thanks! The primitiveChangeClassTo:, I put in in Inspector>>inspect:, was only used by the debugger when inspecting a block closure (generated from the Closure Compiler). I now realize I should do this differently because I don't want Dictionaries, etc. to show as DictionaryInspectors but still as raw inspectors, so I will change this in my Closure Compiler (which makes a few changes to the debugger). So you may remove the primitiveChangeClassTo: if you want. I think it would only be useful if the same Inspector was used to inspect different objects at different times. However, it is never used like this. And even the debugger, which does reused the inspector does not want this behavior. Cheers, Anthony __________________________________ Do you Yahoo!? Yahoo! Mail is new and improved - Check it out! http://promotions.yahoo.com/new_mail

2 1

[ENH][MVC] Deferred UI actions moved into Controller, plus ProcessBrowser fix
by David T. Lewis 04 Aug '04

04 Aug '04

This change set merges the functions of DeferredActionStandardSystemController into Controller. Class DeferredActionStandardSystemController is retained as an stub subclass in order to protect external packages that refer to it (e.g. CommandShell). It should be removed one release cycle after this change set is incorporated. For both MVC and Morphic, deferred UI messages can now be scheduled by: Project addDeferredUIMessage: something The ProcessBrowser has been updated to use this idiom. In addition, some changes are made to ProcessBrowser to permit clean image restart for an active ProcessBrowser with auto-update enabled. Implementation note: The deferred action queue is moved up to Controller rather than StandardSystemController so that scheduling of deferred UI actions works regardless of which controller is currently active. This is important when, for example, deferred actions are scheduled during system startup and a ScreenController is the active controller (ProcessBrowser may do this during startup). ProcessBrowser note: For simplicity, any active instances of ProcessBrowser will have their auto-update processes stopped prior to this update. After the update is complete, new instances of ProcessBrowser will be able to maintain auto-update following an image restart; however, the current instances will still lack this ability. Just open new ProcessBrowsers and all should be well. Both the DeferredActionSystemController and ProcessBrowser are created by Ned Konz. This change set adds no new functionality except as described above. Dave

2 1

[FIX] TextMorphBuglet
by bparsia＠email.unc.edu 04 Aug '04

04 Aug '04

from preamble: "Change Set: TextMorphBuglet Date: 1 December 2003 Author: Bijan Parsia Another in a series of microfixes. This showed up when using the codeProvenance button to toggle colorPrinting from prettyPrinting. Since the actually characters didn't change, the display wouldn't update until you went to another method, or cycled through some other display mode. This fix follows the advice in Text>>=. I expect it'll slow somethings down, especially for complex texts (e.g., with embedded morphs). It may be better to change how this test logic works, but now, at least, the behavior is correct."!

2 1

[ENH] UsableSiblingInstanceFix-nk
by Ned Konz 04 Aug '04

04 Aug '04

from preamble: "Change Set: UsableSiblingInstanceFix-nk Date: 11 October 2003 Author: Ned Konz Morph>>usableSiblingInstance was always adding the created sibling to the World, even though many of its callers would turn around and immediately add the result to other morphs or delete them from the World. This change set makes #usableSiblingInstance not add the newly created Morph to the World. This makes several uses of #usableSiblingInstance more bug-free and efficient. "!

2 1

[BUG][FIX] Morphic: File Selection Dialog [was MVC: File Selectio n Dialog]
by Magistrello Alejandro (SFA) 04 Aug '04

04 Aug '04

>De: Boris Gaertner [SMTP:Boris.Gaertner@gmx.net] >Asunto: Re: [BUG][FIX] MVC: File Selection Dialog (shows another bug) > >I tried to execute > (PluggableFileList oldFileMenu: (FileDirectory on: 'c:\windows\fonts')) > startUpWithCaption: 'Select a File, please:' >in a Mrophic project and receive a strange effect, that I cannot quickly >fix: >I put that statement into a workspace that I place more or less centered >on the screen. When I execute the statement, the selection dialog is >displayed above the workspace. This looks good, but as soon as a >call up one of the menues of the dialog, the selected text of the workspace >comes into the foreground. >I think that in PluggableFileList>>open one should have something like > Smalltalk isMorphic > ifTrue: [<...>] > ifFalse: [<...>] > >I tried to do that, the result can be found in MorphicFSD.1.cs <<MorphicFSD.1.cs>> >In SystemWindow>>update: I added a reaction to the >message #close. >In PluggableFileList>>openAsMorphLasbel:inWorld: >I added an event loop. The condition of the loop uses the >fact that the model of a SystemWindow is niled when the >window is deleted. >I am not very experienced user of the Morphic interface, >so comments are highly wellcome. > > >Greetings, >Boris > if you use the closeButton you'll get a walkback. btw, this is pretty similar to FileList2 modalFileSelector regards, Alejandro

2 1

[SM] SM categories vs Debian debtags
by goran.krampe＠bluefish.se 04 Aug '04

04 Aug '04

Hi all! Just noticed an example where SM (our categories) has been ahead of Debian, until now: http://debtags.alioth.debian.org regards, Göran

2 3

Mac VM 3.7.1b2 (for testing)
by John M McIntosh 04 Aug '04

04 Aug '04

I've pushed a Mac VM 3.7.1b2 for testing to http://homepage.mac.com/johnmci/ Squeak 3.7.1Beta2.app.sit It contains: 3.7.1b2 a) JPEG READER WRITER Usage of Apple '-fast' uncovered a source code issue, initialization routines for internal jpeg plugin were not returning true or false, before this was kinda of considered as returning true? But when compiled with -fast it's zero, meaning the inialization failed, but that is bogus. b) Use of "-fast" apple gcc 3.3 compiler option, improves sends by significant % c) Broke high bit ascii macroman in 3.7.0bx. On EventKeyUp/EventKeyDown you get the UniCode On EventKeyChar you now get the UniCode->MacRoman translated character. For multi-byte character input you get EventKeyDown/EventKeyChar/EventKeyUp for each unicode character coming from the input sequence. Please note that I've a 3.7.1b3 pending due to window open issue in os-x 10.3.2, but please try this VM for speed/performance and correctness. OS-9 Users please note I did not rev the os-9 binary because these are os-x specific issues being fixed at this time. -- ======================================================================== === John M. McIntosh <johnmci(a)smalltalkconsulting.com> 1-800-477-2659 Corporate Smalltalk Consulting Ltd. http://www.smalltalkconsulting.com ======================================================================== ===

3 2

[BUG][FIX] WeakGCFix-wbk
by bryce＠kampjes.demon.co.uk 03 Aug '04

03 Aug '04

from preamble: "Change Set: WeakGCFix-wbk Date: 21 March 2004 Author: Bryce Kampjes Weak references can leave dangling pointers if the weak object is in old space and it points to a new space object. The problem occurs when the new space object is incrementally collected but the old space weak object remains. The weak object now has a dangling pointer to where the new space object was. This change set adds a little code to ObjectMemory>>sweepPhase so that it will finalise any weak rootTable objects. There is still the chance that an interpreter variable may contain a pointer to a weak object. However interpreter variables have very specific uses. I'm not sure if they should be checked for weak references. "!

5 14

[ENH] String and Character conversion methods
by Cancerbero 02 Aug '04

02 Aug '04

Hi. I don't know how formally make this kind of suggestions so excuse me if I'm doing it wrong. My situation is this: The user's input is a String representing a floating point number, so I want something that converts a String representing a floating point number to a Float. Example: Something like String>>asFloat so '3.14' asFloat returns the Float 3.14. So my suggesiont is to add a method String>>asFloat. For example: String>>asFloat "returns the float number represented by this String. For example '0.01' asFloat will return the Float 0.1e-1. The strings MUST be representing float numbers in the format 9.12345 and not in the format 0.1e-4. TO DO: the conversion when the represented float is in the format: 0.1e-4. " |x exp point| x:=0.0. point:=false. (self size) to: 1 by: -1 do: [: i| ((self at: i)=$.) ifTrue: [point:=true] ifFalse: [ point ifTrue: [exp:=self size - i - 1] ifFalse: [exp:=self size - i]. x:=x+(((self at: i)asDigit)*(10 raisedTo: exp)). ]. ]. (self size) to: 1 by: -1 do: [: i| ((self at: i)=$.) ifTrue: [x:=x * (10 raisedTo: (i-(self size)))] ]. ^x where Character>>asDigit is something like: Character>>asDigit "Example : $4 asDigit will return the SmallInteger 4" ^(self asInteger)-($0 asInteger) Bye -- Sebastián Gurin (Cancerbero) cancerbero_sgxARROBAusers.sourceforge.net http://www.geocities.com/cancerbero_sgx Un poco de sabiduría sí es posible; mas en todas las cosas he hallado esta feliz certidumbre: prefieren bailar sobre los pies del azar. --Nietzsche

4 3

About deprecating AlignmentMorph
by stéphane ducasse 02 Aug '04

02 Aug '04

Hi Alignment has been deprecated (and should not be used anymore) but there are a lot of ancient code still using it. Cleaning that would be nice, if somebody dares :). As a first step in that direction I would like to know where coudl we define the helper methods newVariableTransparentSpacer "Answer a space-filling instance of me of the given color." ^ self new hResizing: #spaceFill; vResizing: #spaceFill; layoutInset: 0; borderWidth: 0; extent: 1@1; color: Color transparent and if this is the same as Morph new layoutPolicy: TableLayout new; hResizing: #spaceFill; vResizing: #spaceFill; layoutInset: 0; borderWidth: 0; extent: 1@1; color: Color transparent Stef

3 3

← Newer
1
2
3
4
5
6
7
...
47
Older →

Jump to page: