Squeak-dev September 2009

squeak-dev@lists.squeakfoundation.org

119 participants
370 discussions

Question about FFI and constants
by Mariano Martinez Peck 01 Oct '09

01 Oct '09

Hi folks: I am calling to the OpenDBX library and this library uses C constants. Suppose something like this: enum odbxerr { ODBX_ERR_SUCCESS, #define ODBX_ERR_SUCCESS ODBX_ERR_SUCCESS ODBX_ERR_BACKEND, #define ODBX_ERR_BACKEND ODBX_ERR_BACKEND ODBX_ERR_NOCAP, #define ODBX_ERR_NOCAP ODBX_ERR_NOCAP ODBX_ERR_PARAM, ......... Then I have a function like this: char* odbx_error( odbx_t* handle, int error ) In that case the int error es the index of the array or I can use the constant. Right now, I am using int from 0 to N. But I would like to invoke calls to that function using Smalltalk Strings that represent C constants. For example, now I have this: OpenDBXUnix>>apiError: handle number: err "long odbx_error(odbx_t*, int)" <cdecl: char* 'odbx_error' (ulong long) module: 'opendbx' > ^self externalCallFailed And I call them this way for example: OpenDBXUnix apiError: handle number: 1. But, I would love to do: OpenDBXUnix apiError: handle number: 'ODBX_ERR_BACKEND'. Of course this fails because it cannot coerce and String to a long. So, that someone know how can I do this (if I can) ? Thanks Mariano

3 2

How do I use Monticello Configurations?
by Joshua Gargus 30 Sep '09

30 Sep '09

I've never used Monticello Configurations before, and would like some guidance before I screw up the trunk (there's your motivation to answer quickly, hah! :-). Here are my questions: I don't feel like adding each package individually, since this seems error-prone and is definitely tedious. Is it safe to add packages using match '*'? This seems a bit dangerous, because I might inadvertently add a package that doesn't belong (eg: if I'm also working on a package that isn't in trunk/). Is there a way to start from a previous configuration? In other words, can I populate a new configuration using the latest versions of all packages in a previous configuration? If not, how do others go about this? Can you specify multiple version of the same package in a single configuration? For example, if I want to load Compiler-jcg.85 before Compiler-jcg.86, can I add them in that order to a Configuration, or do I need to create two Configurations? Thanks, Josh

3 2

can't save home project in Damein C's web-based image 3.10.2
by Prashanth Hebbar 30 Sep '09

30 Sep '09

Hello: When I try to save the home project of image 3.10.2-7179 (Damien Cassou's web based image), I get a DNU. Here is the last four context from the debugger: Trait(Object)>>doesNotUnderstand: #isSystemDefined Receiver: TWriteStreamTest Arguments and temporary variables: aMessage: isSystemDefined This happens on both: VM versions 3.10.6 (2007) and 3.11.5 (Sep 2009). I have reported this on mantis (issue id: 0007399) Can someone please help me with this. Regards, -- Prashanth Hebbar

2 1

Re: [squeak-dev] The Trunk: Monticello-ar.323.mcz
by Miguel Enrique Cobá Martinez 30 Sep '09

30 Sep '09

El mié, 30-09-2009 a las 04:39 +0000, commits(a)source.squeak.org escribió: > Andreas Raab uploaded a new version of Monticello to project The Trunk: > http://source.squeak.org/trunk/Monticello-ar.323.mcz Doesn't Monticello has its master repository on the wiresong site? > > ==================== Summary ==================== > > Name: Monticello-ar.323 > Author: ar > Time: 29 September 2009, 9:38:50 am > UUID: 927bef84-8215-0443-afe4-c8575b3a300c > Ancestors: Monticello-cwp.322 > > A small improvement to MC loading. New rules are: > - Pass 1: Load class changes first > - Pass 2: Compile new / changed methods > - Pass 3: Install new / changed methods > - Pass 4: Remove old methods > This should provide a bit more stability to the loading process. > > =============== Diff against Monticello-cwp.322 =============== > > Item was added: > + ----- Method: MCMethodDefinition>>asMethodAddition (in category 'accessing') ----- > + asMethodAddition > + ^MethodAddition new > + compile: source > + classified: category > + withStamp: timeStamp > + notifying: (SyntaxError new category: category) > + logSource: true > + inClass: self actualClass.! > > Item was changed: > ----- Method: MCPackageLoader>>basicLoad (in category 'private') ----- > basicLoad > errorDefinitions := OrderedCollection new. > + [["Pass 1: Load everything but the methods, which are collected in methodAdditions." > + additions do: [:ea | > + [ea isMethodDefinition > + ifTrue:[methodAdditions add: ea asMethodAddition] > + ifFalse:[ea load]]on: Error do: [errorDefinitions add: ea]. > + ] displayingProgress: 'Reshaping classes...'. > + > + "Pass 2: We compile new / changed methods" > + methodAdditions do:[:ea| ea createCompiledMethod] displayingProgress: 'Compiling...'. > - [[ > - > - "FIXME. Do a separate pass on loading class definitions as the very first thing. > - This is a workaround for a problem with the so-called 'atomic' loading (you wish!!) > - which isn't atomic at all but mixes compilation of methods with reshapes of classes. > - > - Since the method is not installed until later, any class reshape in the middle *will* > - affect methods in subclasses that have been compiled before. There is probably > - a better way of dealing with this by ensuring that the sort order of the definition lists > - superclass definitions before methods for subclasses but I need this NOW, and adding > - an extra pass ensures that methods are compiled against their new class definitions." > > + "Pass 3: Install the new / changed methods > + (this is a separate pass to allow compiler changes to be loaded)" > + methodAdditions do:[:ea| ea installMethod] displayingProgress: 'Installing...'. > + > + "Pass 4: Remove the obsolete methods" > + removals do:[:ea| ea unload] displayingProgress: 'Cleaning up ...'. > + > + "Try again any delayed definitions" > - additions do: [:ea | self loadClassDefinition: ea] displayingProgress: 'Loading classes...'. > - > - additions do: [:ea | self tryToLoad: ea] displayingProgress: 'Compiling methods...'. > - removals do: [:ea | ea unload] displayingProgress: 'Cleaning up...'. > self shouldWarnAboutErrors ifTrue: [self warnAboutErrors]. > + errorDefinitions do: [:ea | ea load] displayingProgress: 'Reloading...'. > + > + "Finally, notify observers for the method additions" > - errorDefinitions do: [:ea | ea addMethodAdditionTo: methodAdditions] displayingProgress: 'Reloading...'. > - methodAdditions do: [:each | each installMethod]. > methodAdditions do: [:each | each notifyObservers]. > + > + additions do: [:ea | ea postloadOver: (self obsoletionFor: ea)] displayingProgress: 'Initializing...' > + ] on: InMidstOfFileinNotification do: [:n | n resume: true] > + ] ensure: [self flushChangesFile]! > - additions do: [:ea | ea postloadOver: (self obsoletionFor: ea)] displayingProgress: 'Initializing...'] > - on: InMidstOfFileinNotification > - do: [:n | n resume: true]] > - ensure: [self flushChangesFile]! > > -- Miguel Cobá http://miguel.leugim.com.mx

3 2

New Collections modifications in the inbox
by Levente Uzonyi 30 Sep '09

30 Sep '09

Hi! I uploaded my latest modifications to collections to the inbox. The suggested load/merge order is: Collections-ul.144 Kernel-ul.251 Collections-ul.145 To load/merge these packages, you should have both Collections-ul.140 and Collections-ul.143 loaded/merged. Since the trunk doesn't have enough tests for the Collections package I decided to load these changes into the latest pharo image and run the tests. All of them pass, except for the tests for WeakKeyToCollectionDictionary, but that's because of a bug in the tests (The tests add SmallIntegers as values instead of Collections to the dictionary. Unfortunately they all pass with the original implementation, because #rehash is never sent to the dictionary). To see if it's worth (or not) to use these changes I wrote a small benchmark for Dictionary which can be found here: http://leves.web.elte.hu/collections/ . Cheers, Levente

3 6

Problem with ProportionalSplitters
by Alexander Lazarević 30 Sep '09

30 Sep '09

Hi! I guess not too many use ProportionalSplitters to change the layout of windows these days, but they are broken in the latest trunk image. This is because the Method containingWindow in Morph has changed. The containing Window is now searched based on the model of the component, but ProportionalSplitters do have no model. A quick fix is to look up the owner chain of a ProportionalSplitter to find an owner that has a model. Most of the time (always?) the owner will be the containing window, so I don't know if the ifFalse: Branch is needed. I will publish this fix so ProportionalSplitters should work again until someone with better insights has a better solution. Alex

4 6

Last call for flatten vs. flattened
by Andreas Raab 30 Sep '09

30 Sep '09

Folks - After having popped the stack of stuff in the inbox, I'm intent on clearing it out for good. As such I need either people to voice opinions (and write code) about the issue of #flatten vs. #flattened or if noone cares it'll be moved out of the inbox and into treated. Either way is fine by me. Cheers, - Andreas

3 2

The Trunk: Compiler-jcg.85.mcz
by commits＠source.squeak.org 30 Sep '09

30 Sep '09

Joshua Gargus uploaded a new version of Compiler to project The Trunk: http://source.squeak.org/trunk/Compiler-jcg.85.mcz ==================== Summary ==================== Name: Compiler-jcg.85 Author: jcg Time: 30 September 2009, 1:14:16 am UUID: 21a5f88d-4107-48c3-b41c-d5ea735319a7 Ancestors: Compiler-tfel.84 Revert to lazy-initialization of method properties, which is used by the externally-maintained Prolog package (which loads cleanly into 3.10.2, but not trunk). Part 1 of 2... need to first re-add the #properties method to Parser before changing methods to send it. =============== Diff against Compiler-tfel.84 =============== Item was added: + ----- Method: Parser>>properties (in category 'pragmas') ----- + properties + ^ properties ifNil: [ properties := AdditionalMethodState new ]!

1 0

The Trunk: Collections-ar.147.mcz
by commits＠source.squeak.org 30 Sep '09

30 Sep '09

Andreas Raab uploaded a new version of Collections to project The Trunk: http://source.squeak.org/trunk/Collections-ar.147.mcz ==================== Summary ==================== Name: Collections-ar.147 Author: ar Time: 29 September 2009, 11:39:57 am UUID: bef1189d-525b-c540-a331-e19062c1ecbf Ancestors: Collections-ar.146, Collections-ul.136 Merging Collections-ul.136 - primitive 60 won't help with Array >> #at:ifAbsent: because it accepts only one parameter, so changed SequenceableCollection >> #at:ifAbsent: to get some improvement (assumes that indexes greater than the upper bound are more common than indexes smaller than the lower bound) =============== Diff against Collections-ar.146 =============== Item was changed: ----- Method: SequenceableCollection>>at:ifAbsent: (in category 'accessing') ----- at: index ifAbsent: exceptionBlock "Answer the element at my position index. If I do not contain an element at index, answer the result of evaluating the argument, exceptionBlock." + (index <= self size and: [ 1 <= index ]) ifTrue: [ ^self at: index ]. + ^exceptionBlock value! - (index between: 1 and: self size) ifTrue: [^ self at: index]. - ^ exceptionBlock value!

1 0

The Trunk: Collections-ar.146.mcz
by commits＠source.squeak.org 30 Sep '09

30 Sep '09

Andreas Raab uploaded a new version of Collections to project The Trunk: http://source.squeak.org/trunk/Collections-ar.146.mcz ==================== Summary ==================== Name: Collections-ar.146 Author: ar Time: 29 September 2009, 11:13:32 am UUID: 92161bb4-bec3-7a43-b30d-9d2d9da56bb8 Ancestors: Collections-ar.145 Pseudo-merging Collections-ul.145: deprecated #findElementOrNil: and #noCheckAdd: =============== Diff against Collections-ar.145 =============== Item was changed: ----- Method: KeyedSet>>noCheckAdd: (in category 'private') ----- noCheckAdd: anObject + + self deprecated: 'This method should not be used anymore.'. array at: (self scanFor: (keyBlock value: anObject)) put: anObject. tally := tally + 1! Item was changed: ----- Method: Set>>findElementOrNil: (in category 'private') ----- findElementOrNil: anObject "Answer the index of a first slot containing either a nil (indicating an empty slot) or an element that matches the given object. Answer the index of that slot or zero. Fail if neither a match nor an empty slot is found." | index | + self deprecated: 'Use #scanFor:.'. - index := self scanFor: anObject. index > 0 ifTrue: [^index]. "Bad scene. Neither have we found a matching element nor even an empty slot. No hashed set is ever supposed to get completely full." self error: 'There is no free space in this set!!'.! Item was changed: ----- Method: Set>>noCheckAdd: (in category 'private') ----- noCheckAdd: anObject + + self deprecated: 'This method should not be used anymore.'. array at: (self scanFor: anObject) put: anObject. tally := tally + 1! Item was changed: ----- Method: Dictionary>>noCheckAdd: (in category 'private') ----- noCheckAdd: anObject "Must be defined separately for Dictionary because (self scanFor:) expects a key, not an association. 9/7/96 tk" + self deprecated: 'This method should not be used anymore.'. array at: (self scanFor: anObject key) put: anObject. tally := tally + 1!

1 0

Jump to page: