Squeak-dev September 2017

squeak-dev@lists.squeakfoundation.org

30 participants
85 discussions

The Trunk: Compression-dtl.54.mcz
by commits＠source.squeak.org 28 Sep '17

28 Sep '17

David T. Lewis uploaded a new version of Compression to project The Trunk: http://source.squeak.org/trunk/Compression-dtl.54.mcz ==================== Summary ==================== Name: Compression-dtl.54 Author: dtl Time: 28 September 2017, 8:00:37.70922 am UUID: 97cfd1cd-0c07-40a7-92e9-4ea35b82fd5e Ancestors: Compression-nice.53 In an archive file is too short, or more commonly if it is empty, identify its name in the error message. This is helpful in the case of a corrupt file in a package-cache, which may have been created due to some earlier system or network problem. =============== Diff against Compression-nice.53 =============== Item was changed: ----- Method: ZipArchive class>>findEndOfCentralDirectoryFrom: (in category 'constants') ----- findEndOfCentralDirectoryFrom: stream "Seek in the given stream to the end, then read backwards until we find the signature of the central directory record. Leave the file positioned right before the signature. Answers the file position of the EOCD, or 0 if not found." | data fileLength seekOffset pos maxOffset | stream setToEnd. fileLength := stream position. "If the file length is less than 18 for the EOCD length plus 4 for the signature, we have a problem" + fileLength < 22 ifTrue: [^ self error: 'file is too short: ', stream name]. - fileLength < 22 ifTrue: [^ self error: 'file is too short']. seekOffset := 0. pos := 0. data := ByteArray new: 4100. maxOffset := 40960 min: fileLength. "limit search range to 40K" [ seekOffset := (seekOffset + 4096) min: fileLength. stream position: fileLength - seekOffset. data := stream next: (4100 min: seekOffset) into: data startingAt: 1. pos := self lastIndexOfPKSignature: EndOfCentralDirectorySignature in: data. pos = 0 and: [seekOffset < maxOffset] ] whileTrue. ^ pos > 0 ifTrue: [ | newPos | stream position: (newPos := (stream position + pos - seekOffset - 1)). newPos] ifFalse: [0]!

1 0

The Inbox: Morphic-hjh.1351.mcz
by commits＠source.squeak.org 28 Sep '17

28 Sep '17

A new version of Morphic was added to project The Inbox: http://source.squeak.org/inbox/Morphic-hjh.1351.mcz ==================== Summary ==================== Name: Morphic-hjh.1351 Author: hjh Time: 28 September 2017, 11:39:54.321636 am UUID: 8eb78880-4b1d-4e8b-b961-d4d5c49d1c47 Ancestors: Morphic-mt.1350 Removed comment from Editor referring to SimpleEditor and CellStyleEditor which are not in the image. =============== Diff against Morphic-mt.1350 =============== Item was changed: Object subclass: #Editor instanceVariableNames: 'morph selectionShowing' classVariableNames: 'BlinkingCursor DestructiveBackWord DumbbellCursor KeystrokeActions SelectionsMayShrink' poolDictionaries: '' category: 'Morphic-Text Support'! + !Editor commentStamp: 'hjh 9/28/2017 11:37' prior: 0! - !Editor commentStamp: '<historical>' prior: 0! New text editors. TextEditor provides most of the functionality that used to be in TextMorphEditor. SmalltalkEditor is has Smalltalk code specific features. + ! - SimpleEditor provides basic functionality for single line text editing. It does not handle fonts and styles, aligning and Smalltalk utilities. It handles one single line. - CellStyleEditor allows entering alphabetic characters using only number keys, like many cell phones do.!

1 0

SystemWindow positiioning question
by David T. Lewis 28 Sep '17

28 Sep '17

When I use the mouse to move a SystemWindow to a new position, the mouse down produces a square frame that moves with the hand position, and the mouse up makes the frame go away with the SystemWindow being relocated to the new position. In most cases, the new position of the SystemWindow is the position to which it was moved with the mouse. However, if the new position is at the top of the world PasteUpMorph, possibly conflicting with the world main docking bar or the Sugar navigator bar, then the new SystemWindow position is modified to avoid the top region. Where is the code that figures out the new position in the case of a SystemWindow being dragged to the top of the world PasteUpMorph? I am trying to debug a window positioning problem in one of my images, and I cannot figure out where the magic happens. Thanks, Dave

2 2

Metacello load error 'File is too short' in Squeak6.0alpha latest update: #17402 ?
by H. Hirzel 28 Sep '17

28 Sep '17

Hello Following the Metacello installation instructions [1] the following step A runs fine but the step B afterwards gives an error message 'File too short' [2], marked with '"ERROR FILE TOO SHORT' "A: Get the Metacello configuration (for Squeak users)" Installer gemsource project: 'metacello'; addPackage: 'ConfigurationOfMetacello'; install. "B: Bootstrap Metacello Preview, using mcz files (#'previewBootstrap' symbolic version" ((Smalltalk at: #ConfigurationOfMetacello) project version: #'previewBootstrap') load. Any suggestion how to fix this? Kind regards Hannes [1] https://github.com/dalehenrich/metacello-work [2] ZipArchive class findEndOfCentralDirectoryFrom: stream "Seek in the given stream to the end, then read backwards until we find the signature of the central directory record. Leave the file positioned right before the signature. Answers the file position of the EOCD, or 0 if not found." | data fileLength seekOffset pos maxOffset | stream setToEnd. fileLength := stream position. "If the file length is less than 18 for the EOCD length plus 4 for the signature, we have a problem" fileLength < 22 ifTrue: [^ self error: 'file is too short']. "ERROR FILE TOO SHORT" seekOffset := 0. pos := 0. data := ByteArray new: 4100. maxOffset := 40960 min: fileLength. "limit search range to 40K" [ seekOffset := (seekOffset + 4096) min: fileLength. stream position: fileLength - seekOffset. data := stream next: (4100 min: seekOffset) into: data startingAt: 1. pos := self lastIndexOfPKSignature: EndOfCentralDirectorySignature in: data. pos = 0 and: [seekOffset < maxOffset] ] whileTrue. ^ pos > 0 ifTrue: [ | newPos | stream position: (newPos := (stream position + pos - seekOffset - 1)). newPos] ifFalse: [0]

2 3

Abusing #userInterfaceTheme
by Stéphane Rollandin 27 Sep '17

27 Sep '17

Hello, I would like to offer some technical feedback here, please do not take it as a complaint directed against anyone's good work. In particular, thanks again to all involved in designing the theming system. Only, there are some worrisome issues in my opinion. Here we go: 1) The code below raises a DNU for #hardShadowOffset in SqueakTheme, because #userInterfaceTheme is not used as intented (after #softShadowOffset is sent, m gets popped out of the theme scope and is not there anymore when #hardShadowOffset is sent): | m uit | m := Morph new. uit := m userInterfaceTheme. {uit softShadowOffset . uit hardShadowOffset} However, for someone not aware of the intricacies of UserInterfaceTheme, the code seems fine. In my opinion this is not a good state of affair. Maybe the error raised here could be more explicit and in line with what is really happening in the DNU, such as "There is no object in scope at the moment". 2) Now consider the following code: | m1 uit1 m2 uit2 | m2 := BalloonMorph new. uit2 := m2 userInterfaceTheme . m1 := Morph new. uit1 := m1 userInterfaceTheme. {uit1 color . uit1 color} When I evaluate it I get {nil . Color lightYellow1} This is because while the first "uit1 color" does redirect to m1, the second "uit1 color" redirects to m2 now that m1 has been popped out of scope. This time there is no error raised, but the result is very misleading. 3) For the record let me recall that doing 100 timesRepeat: [Morph new userInterfaceTheme] results in having the current theme referencing 100 morphs that will never be automatically garbage-collected. Overall the three above points make me think that the method #userInterfaceTheme is much too easy to unknowingly abuse of, and when this happens undesirable behaviors can be rather difficult to make sense of. At the moment the method comment is: "Call this to conveniently access properties from the current user interface theme." I believe it needs some explicit warnings about how not to use it. Just my 2 cents... Stef

3 10

logging service?
by Alan Pinch 26 Sep '17

26 Sep '17

Is there a standard logging service I could use in squeak? Much of my activity I want to log is on non-UI background processes. -- Thank you for your consideration, Alan

2 2

UserInterfaceTheme testing & TestRunner
by tim Rowledge 26 Sep '17

26 Sep '17

I tried running the tests on UserInterfaceThem to check I hadn’t broken too much and as is so often the case on slower machines some of the tests timed out. It’s not a simply replicatable thing either, with each run producing different tests that ‘fail’ this way. So the obvious thing to day was a) add a timeout: pragma to the tests that seem to be slow b) add an option to profile an individual test While running these to play with them I managed to occasionally lock up the input process, which I’d make a wild guess might be something to do with the debugging etc in the testrunner code. It’s not reliable enough to get a good handle on it. It did, however manage to completely lock up the Pi on one occasion - not merely Squeak but the entire Pi. Attempts to login remotely had no effect. Naughty, naughty somewhere in the stack. Whilst writing this little extension I noticed some code of a form I’m not keen on. Perhaps there is a good solid reason for it to be like this that I’m unaware of - runFailures self result instVarNamed: 'failures' put: Set new. self runSuite: self suiteFailures. (similar in runErrors) There’s not often a good reason to use #instVarNamed:put: and it’s sufficiently odd that I have to wonder if there is a hidden subtlety making it appropriate. tim -- tim Rowledge; tim(a)rowledge.org; http://www.rowledge.org/tim "Bother!" said Pooh, as Piglet pressed <START> on the Microwave...

3 3

The Trunk: SUnitGUI-tpr.68.mcz
by commits＠source.squeak.org 25 Sep '17

25 Sep '17

tim Rowledge uploaded a new version of SUnitGUI to project The Trunk: http://source.squeak.org/trunk/SUnitGUI-tpr.68.mcz ==================== Summary ==================== Name: SUnitGUI-tpr.68 Author: tpr Time: 25 September 2017, 2:40:47.67289 pm UUID: 5d33c773-e628-446f-88e8-47bfdbf2dc22 Ancestors: SUnitGUI-ul.67 Add a simple way to run a single failed method in a MessageTally. =============== Diff against SUnitGUI-ul.67 =============== Item was changed: ----- Method: TestRunner>>failureMenu: (in category 'accessing-menu') ----- failureMenu: aMenu + ^ aMenu + add: 'Profile this test' action: #runProfileOnFailure; + yourself! - ^ aMenu! Item was added: + ----- Method: TestRunner>>runProfileOnFailure (in category 'actions') ----- + runProfileOnFailure + "if we have selected a failure case, run it within a MessageTAlly and see what is taking time" + + failedSelected ifNotNil: [self + basicRunSuite: (TestSuite new addTest: failedSelected; + yourself) + do: [:each | MessageTally + spyOn: [self runTest: each]]]!

1 0

The Trunk: System-tpr.964.mcz
by commits＠source.squeak.org 25 Sep '17

25 Sep '17

tim Rowledge uploaded a new version of System to project The Trunk: http://source.squeak.org/trunk/System-tpr.964.mcz ==================== Summary ==================== Name: System-tpr.964 Author: tpr Time: 25 September 2017, 12:03:05.473661 pm UUID: ee9898d1-8279-4c25-b78c-f423cf7eb408 Ancestors: System-mt.963 A proposed improvement to the class comment to reflect recent changes, along with adding the most basic property setting message to the request instances. I suggest that there are a number of other messages that ought to be passed along in a similar manner but others are likely to have a better idea of the right ones. =============== Diff against System-mt.963 =============== Item was changed: Object subclass: #UserInterfaceTheme instanceVariableNames: 'properties name next ignoreApply lastScaleFactor' classVariableNames: 'All Current Default' poolDictionaries: '' category: 'System-Support'! + !UserInterfaceTheme commentStamp: 'tpr 9/25/2017 11:59' prior: 0! - !UserInterfaceTheme commentStamp: '<historical>' prior: 0! A UserInterfaceTheme is a dictionary of preferred visual-properties; colors, borderStyles, borderWidths, fonts, forms, etc. used to color and style the IDE. Accessing The Theme + To access the proper UserInterfaceTheme instance for an object, send it #userInterfaceTheme. The default implementation on Object provides the an instance of UserInterfaceThemeRequest that provides a lightweight, clean proxy of the actual theme in-use by the IDE at the current time. To do anything more sophisticated than basic query and setting of properties you must ask the proxy for the actual theme by sending #theme. - To access the proper UserInterfaceTheme instance for an object, send it #userInterfaceTheme. The default implementation on Object provides the one instance of that is in-use by the IDE at the current time. Customizing The Theme We can ask the userInterfaceTheme for the value of any visual-property, by name: mySystemWindow userInterfaceTheme closeBoxImage + Initially that would answer nil, which causes the legacy code to use whatever default it's always used. To override various visual-properties of any kind of object, the #set: for: to: message can be used. For example, - Initially, the above answers nil, which causes the legacy code to use whatever default it's always used. To override various visual-properties of any kind of object, the #set: onAny: to: message can be used. For example, myUserInterfaceTheme set: #closeBoxImage for: SystemWindow to: MenuIcons smallCancelIcon + Now the closeBoxImage message will answer the MenuIcons icon instead of nil. + Alternatively, values may be derived based on other values in the theme, as in: myUserInterfaceTheme set: #color for: FillInTheBlankMorph to: { MenuMorph->#color. #twiceDarker } + This would make FillInTheBlankMorph use the same color as a MenuMorph but twiceDarker, providing a clean way to build coherent seets of colors within a theme. SystemWindow's code can be changed to use the expression above to access elements of the theme. - Now, the accessing expression, above, will answer will answer MenuIcons' smallCancelIcon instead of nil. SystemWindow's code can be changed to use the expression above to access elements of the theme. Upgrading Legacy Code Following the introduction of this class, various client code all around the system must be modified to access it. This variety of legacy code uses a variety of methods to specify their visual properties: 1) a hard-coded values. 2) a values derived from some other value. 3) providing local storage for a settable value which can be nil. 4) providing local storage for a settable value which is expected to always have a particular valid value (never nil). The requirement, for each case, is to let the value be overridden. The solution for each of the above should be handled respectively to the above list, as follows: 1) Check the userInterfaceTheme, if that property returns nil, use the legacy hard-coded value. (see example: SystemWindow>>#createCloseBox). 2) Nothing to do -- simply perform the same derivation on the result of (1). 3) Check the local storage, if present, use it. If nil, then check the userInterfaceTheme, if it has this property present, use it, else return nil. 4) Check the userInterfaceTheme, if the property is not nil, use it, otherwise use the local value. Tool Support If a new access to #userInterfaceTheme is added to the code, be sure to add the property and its description to the #themeSettings for that class. See implementors of #themeSettings for examples.! Item was added: + ----- Method: UserInterfaceThemeRequest>>set:for:to: (in category 'accessing') ----- + set: propertySymbol for: aClassOrSymbol to: valueObject + "Where aClass asks its userInterfaceTheme for propertySymbol, provide valueObject. Pass this on to the actual theme" + + ^self theme set: propertySymbol for: aClassOrSymbol to: valueObject + + !

1 0

Cannot rename class vars via Refactoring Tools menu
by sst 25 Sep '17

25 Sep '17

Hi everyone, I just installed the Refactoring Tools in Squeak 5.1 like this: Metacello new configuration: 'RefactoringTools'; version: #stable; load. and I am trying to rename some class variables (not instance variables) but when I select the option from the menu shown below it prompts me for a new class name, which is unexpected. I'm expecting it to give me the list of class vars (which do exist) to select one to rename (the same way that the instance vars rename feature works). If I provide a class name it actually renames my class. Seems that the wrong code is attached to this menu item? Anyone else had this issue? <http://forum.world.st/file/t266054/refactor.png> I'm trying to figure out where the code is that attaches a method to this menu. I suspect the wrong method is being called from this menu as the RBRenameClassVariableTest passes in TestRunner which to me means the code to rename class variables is working as expected, it's just not linked to this menu item as expected? Any help appreciated, Steve. -- Sent from: http://forum.world.st/Squeak-Dev-f45488.html

2 3

← Newer
1
2
3
4
5
6
7
8
9
Older →