Squeak-dev November 2009

squeak-dev@lists.squeakfoundation.org

89 participants
351 discussions

The Trunk: Monticello-ul.331.mcz
by commits＠source.squeak.org 05 Nov '09

05 Nov '09

Andreas Raab uploaded a new version of Monticello to project The Trunk: http://source.squeak.org/trunk/Monticello-ul.331.mcz ==================== Summary ==================== Name: Monticello-ul.331 Author: ul Time: 4 November 2009, 10:55:34 am UUID: d5d3402e-d4f0-4c46-8a5f-0a81f2d790ae Ancestors: Monticello-nice.330 - faster hashing with String >> #hashWithInitialHash:. Load Collections-ul.179 before this package. =============== Diff against Monticello-nice.330 =============== Item was changed: ----- Method: MCClassTraitDefinition>>hash (in category 'accessing') ----- hash + | hash | + hash := baseTrait hashWithInitialHash: 0. + hash := self classTraitCompositionString hashWithInitialHash: hash. - hash := String stringHash: baseTrait initialHash: 0. - hash := String stringHash: self classTraitCompositionString initialHash: hash. ^hash ! Item was changed: ----- Method: MCClassDefinition>>hash (in category 'comparing') ----- hash + | hash | + hash := name hashWithInitialHash: 0. + hash := superclassName hashWithInitialHash: hash. + hash := self traitCompositionString hashWithInitialHash: hash. + hash := self classTraitComposition asString hashWithInitialHash: hash. + hash := (category ifNil: ['']) hashWithInitialHash: hash. + hash := type hashWithInitialHash: hash. + variables do: [ :v | + hash := v name hashWithInitialHash: hash ]. + ^hash! - hash := String stringHash: name initialHash: 0. - hash := String stringHash: superclassName initialHash: hash. - hash := String stringHash: self traitCompositionString initialHash: hash. - hash := String stringHash: self classTraitComposition asString initialHash: hash. - hash := String stringHash: (category ifNil: ['']) initialHash: hash. - hash := String stringHash: type initialHash: hash. - variables do: [ - :v | - hash := String stringHash: v name initialHash: hash. - ]. - ^ hash! Item was changed: ----- Method: MCTraitDefinition>>hash (in category 'comparing') ----- hash + | hash | + hash := name hashWithInitialHash: 0. + hash := self traitCompositionString hashWithInitialHash: hash. + hash := (category ifNil: ['']) hashWithInitialHash: hash. + ^hash - hash := String stringHash: name initialHash: 0. - hash := String stringHash: self traitCompositionString initialHash: hash. - hash := String stringHash: (category ifNil: ['']) initialHash: hash. - ^ hash ! Item was changed: ----- Method: MCMethodDefinition>>hash (in category 'comparing') ----- hash | hash | + hash := classIsMeta asString hashWithInitialHash: 0. + hash := source hashWithInitialHash: hash. + hash := category hashWithInitialHash: hash. + hash := className hashWithInitialHash: hash. - hash := String stringHash: classIsMeta asString initialHash: 0. - hash := String stringHash: source initialHash: hash. - hash := String stringHash: category initialHash: hash. - hash := String stringHash: className initialHash: hash. ^ hash!

1 0

The Trunk: Kernel-ul.287.mcz
by commits＠source.squeak.org 05 Nov '09

05 Nov '09

Andreas Raab uploaded a new version of Kernel to project The Trunk: http://source.squeak.org/trunk/Kernel-ul.287.mcz ==================== Summary ==================== Name: Kernel-ul.287 Author: ul Time: 4 November 2009, 12:18:09 pm UUID: 8dcd7ba0-4578-4d4b-836f-b5050896431a Ancestors: Kernel-nice.286 - simplified ClassDescription >> #organization =============== Diff against Kernel-nice.286 =============== Item was changed: ----- Method: ClassDescription>>organization (in category 'organization') ----- organization "Answer the instance of ClassOrganizer that represents the organization of the messages of the receiver." + organization ifNil: [ + self organization: (ClassOrganizer defaultList: self methodDict keys sort) ]. + (organization isMemberOf: Array) ifTrue: [ + self recoverFromMDFaultWithTrace ]. - organization ifNil: - [self organization: (ClassOrganizer defaultList: self methodDict keys asSortedCollection asArray)]. - (organization isMemberOf: Array) ifTrue: - [self recoverFromMDFaultWithTrace]. "Making sure that subject is set correctly. It should not be necessary." + organization ifNotNil: [ organization setSubject: self ]. + ^organization! - organization ifNotNil: [organization setSubject: self]. - ^ organization!

1 0

The Trunk: Collections-ar.183.mcz
by commits＠source.squeak.org 05 Nov '09

05 Nov '09

Andreas Raab uploaded a new version of Collections to project The Trunk: http://source.squeak.org/trunk/Collections-ar.183.mcz ==================== Summary ==================== Name: Collections-ar.183 Author: ar Time: 4 November 2009, 11:14:39 am UUID: 2e07a36b-f72b-6a42-a58b-72a620a3d515 Ancestors: Collections-ul.179, Collections-ul.180, Collections-ul.181, Collections-ul.182 Merging Collections-ul.179, Collections-ul.180, Collections-ul.181, Collections-ul.182: - added: String >> #hashWithInitialHash: which lets the string to decide which hash implementation to use, this way ByteStrings can use the primitive implementation instead of the general one - speed up mergesort for the general case (using <= for comparison), sacrificing < 1% performance for other cases - allow symbols to be evaluated as comparison blocks. For example: sorting an Array in descening order can be as simple as: array sort: #>=, it's also faster than using a block. - simplified Dictionary >> #keysSortedSafely - simplified WeakRegistry >> #keys - save a few bytecodes in #scanFor: of identity based collections when their size is between 4097 and 8191 - fix: WeakKeyDictionary >> #finalizeValues: (increment the tally, and make sure that we don't lose the key) - collect the objects which should be finalized into an IdentitySet instead of an OrderedCollection in WeakRegistry >> #finalizeValues this should give better performance in WeakKeyDictionary >> #finalizeValues: =============== Diff against Collections-nice.178 =============== Item was changed: ----- Method: ArrayedCollection>>sort: (in category 'sorting') ----- sort: aSortBlock "Sort this array using aSortBlock. The block should take two arguments + and return true if the first element should preceed the second one. + If aSortBlock is nil then <= is used for comparison." - and return true if the first element should preceed the second one." self mergeSortFrom: 1 to: self size by: aSortBlock! Item was changed: ----- Method: IdentityDictionary>>scanFor: (in category 'private') ----- scanFor: anObject "Scan the key array for the first slot containing either a nil (indicating an empty slot) or an element that matches anObject. Answer the index of that slot or raise an error if no slot is found. This method will be overridden in various subclasses that have different interpretations for matching elements." | index start hash | + array size >= 8192 - array size > 4096 ifTrue: [ hash := anObject identityHash * (array size // 4096) ] ifFalse: [ hash := anObject identityHash ]. index := start := hash \\ array size + 1. [ | element | ((element := array at: index) == nil or: [ element key == anObject ]) ifTrue: [ ^index ]. (index := index \\ array size + 1) = start ] whileFalse. self errorNoFreeSpace! Item was changed: ----- Method: ArrayedCollection>>mergeFirst:middle:last:into:by: (in category 'sorting') ----- mergeFirst: first middle: middle last: last into: dst by: aBlock "Private. Merge the sorted ranges [first..middle] and [middle+1..last] of the receiver into the range [first..last] of dst." | i1 i2 val1 val2 out | i1 := first. i2 := middle + 1. val1 := self at: i1. val2 := self at: i2. out := first - 1. "will be pre-incremented" "select 'lower' half of the elements based on comparator" + [ (i1 <= middle) and: [ i2 <= last ] ] whileTrue: [ + (aBlock + ifNil: [ val1 <= val2 ] + ifNotNil: [ aBlock value: val1 value: val2 ]) + ifTrue: [ + dst at: (out := out + 1) put: val1. - [(i1 <= middle) and: [i2 <= last]] whileTrue: - [(aBlock value: val1 value: val2) - ifTrue: [dst at: (out := out + 1) put: val1. val1 := self at: (i1 := i1 + 1)] + ifFalse: [ + dst at: (out := out + 1) put: val2. + (i2 := i2 + 1) <= last ifTrue: [ + val2 := self at: i2 ] ] ]. - ifFalse: [dst at: (out := out + 1) put: val2. - i2 := i2 + 1. - i2 <= last ifTrue: [val2 := self at: i2]]]. "copy the remaining elements" i1 <= middle ifTrue: [dst replaceFrom: out + 1 to: last with: self startingAt: i1] ifFalse: [dst replaceFrom: out + 1 to: last with: self startingAt: i2]! Item was changed: ----- Method: WeakRegistry>>finalizeValues (in category 'finalization') ----- finalizeValues "Some of our elements may have gone away. Look for those and activate the associated executors." | finiObjects | finiObjects := nil. "First collect the objects." + self protected: [ + valueDictionary associationsDo: [ :assoc | + assoc key ifNil: [ + (finiObjects ifNil: [ finiObjects := IdentitySet new ]) + add: assoc value ] ]. + finiObjects ifNotNil: [ valueDictionary finalizeValues: finiObjects ] ]. - self protected:[ - valueDictionary associationsDo:[:assoc| - assoc key isNil ifTrue:[ - finiObjects isNil - ifTrue:[finiObjects := OrderedCollection with: assoc value] - ifFalse:[finiObjects add: assoc value]] - ]. - finiObjects isNil ifFalse:[valueDictionary finalizeValues: finiObjects asArray]. - ]. "Then do the finalization" + finiObjects ifNil: [ ^self ]. + finiObjects do: [ :each | each finalize ]! - finiObjects isNil ifTrue:[^self]. - finiObjects do:[:each| each finalize]. - ! Item was changed: ----- Method: WeakIdentityKeyDictionary>>scanFor: (in category 'private') ----- scanFor: anObject "Scan the key array for the first slot containing either a nil (indicating an empty slot) or an element that matches anObject. Answer the index of that slot or raise an error if no slot is found. This method will be overridden in various subclasses that have different interpretations for matching elements." | index start hash | + array size >= 8192 - array size > 4096 ifTrue: [ hash := anObject identityHash * (array size // 4096) ] ifFalse: [ hash := anObject identityHash ]. index := start := hash \\ array size + 1. [ | element | ((element := array at: index) == nil or: [ element key == anObject ]) ifTrue: [ ^index ]. (index := index \\ array size + 1) = start ] whileFalse. self errorNoFreeSpace! Item was added: + ----- Method: Symbol>>value:value: (in category 'evaluating') ----- + value: anObject value: anotherObject + + ^anObject perform: self with: anotherObject! Item was changed: ----- Method: ArrayedCollection>>sort (in category 'sorting') ----- sort "Sort this array into ascending order using the '<=' operator." + self sort: nil! - self sort: [:a :b | a <= b]! Item was changed: ----- Method: OrderedCollection>>sort: (in category 'sorting') ----- sort: aSortBlock + "Sort this collection using aSortBlock. The block should take two arguments + and return true if the first element should preceed the second one. + If aSortBlock is nil then <= is used for comparison." - "Sort this array using aSortBlock. The block should take two arguments - and return true if the first element should preceed the second one." self ifNotEmpty: [ array mergeSortFrom: firstIndex to: lastIndex by: aSortBlock ]! Item was added: + ----- Method: String>>hashWithInitialHash: (in category 'comparing') ----- + hashWithInitialHash: initialHash + + ^ self class stringHash: self initialHash: initialHash! Item was changed: ----- Method: Dictionary>>keysSortedSafely (in category 'accessing') ----- keysSortedSafely + "Answer a sorted Array containing the receiver's keys." - "Answer an Array containing the receiver's keys." + ^self keys sort: [ :x :y | - | sortedKeys | - sortedKeys := Array new: self size streamContents: [ :stream | - self keysDo: [ :each | stream nextPut: each ] ]. - sortedKeys sort: [ :x :y | "Should really be use <obj, string, num> compareSafely..." ((x isString and: [ y isString ]) or: [ x isNumber and: [ y isNumber ] ]) ifTrue: [ x < y ] ifFalse: [ x class == y class ifTrue: [ x printString < y printString ] ifFalse: [ x class name < y class name ] ] ]. + ! - ^sortedKeys! Item was changed: ----- Method: WeakRegistry>>keys (in category 'accessing') ----- keys + + ^self protected: [ valueDictionary keys ]! - ^self protected:[ - Array streamContents:[:s| valueDictionary keysDo:[:key| s nextPut: key]]].! Item was changed: ----- Method: WeakKeyDictionary>>finalizeValues: (in category 'finalization') ----- finalizeValues: finiObjects "Remove all associations with key == nil and value is in finiObjects. This method is folded with #rehash for efficiency." | oldArray | oldArray := array. array := Array new: oldArray size. tally := 0. 1 to: array size do:[ :i | | association | (association := oldArray at: i) ifNotNil: [ + | key | + ((key := association key) == nil and: [ "Don't let the key go away" + finiObjects includes: association value ]) + ifFalse: [ + array + at: (self scanForEmptySlotFor: key) + put: association. + tally := tally + 1 ] ] ]! - (association key == nil and: [ finiObjects includes: association value ]) - ifFalse:[ - array - at: (self scanForEmptySlotFor: association key) - put: association ] ] ]! Item was changed: ----- Method: OrderedCollection>>sort (in category 'sorting') ----- sort "Sort this array into ascending order using the '<=' operator." + self sort: nil! - self sort: [:a :b | a <= b]! Item was changed: ----- Method: IdentitySet>>scanFor: (in category 'private') ----- scanFor: anObject "Scan the key array for the first slot containing either a nil (indicating an empty slot) or an element that matches anObject. Answer the index of that slot or raise an error if no slot is found. This method will be overridden in various subclasses that have different interpretations for matching elements." | index start hash | + array size >= 8192 - array size > 4096 ifTrue: [ hash := anObject identityHash * (array size // 4096) ] ifFalse: [ hash := anObject identityHash ]. index := start := hash \\ array size + 1. [ | element | ((element := array at: index) == nil or: [ element == anObject ]) ifTrue: [ ^index ]. (index := index \\ array size + 1) = start ] whileFalse. self errorNoFreeSpace! Item was changed: ----- Method: KeyedIdentitySet>>scanFor: (in category 'private') ----- scanFor: anObject "Scan the key array for the first slot containing either a nil (indicating an empty slot) or an element that matches anObject. Answer the index of that slot or raise an error if no slot is found. This method will be overridden in various subclasses that have different interpretations for matching elements." | index start hash | + array size >= 8192 - array size > 4096 ifTrue: [ hash := anObject identityHash * (array size // 4096) ] ifFalse: [ hash := anObject identityHash ]. index := start := hash \\ array size + 1. [ | element | ((element := array at: index) == nil or: [ (keyBlock value: element) == anObject ]) ifTrue: [ ^index ]. (index := index \\ array size + 1) = start ] whileFalse. self errorNoFreeSpace!

1 0

Sophie 2.0 ported to Java. Why?
by alesch 04 Nov '09

04 Nov '09

Hi List: Does anyone know why Sophie is being ported to Java? In http://sophie2.org/trac/wiki/AboutPage this page , they mention Astea Solutions and their contributed Java codebase... "Sophie is currently being significantly revised and improved, thanks to a third grant from the Mellon Foundation in October, 2008 and to a Java codebase contributed by Astea Solutions, a company formed to design and develop proprietary and open source electronic publishing products. With this funding, Sophie is not only being rewritten in Java to enhance its stability, but is being transformed from its initial iteration as a powerful multimedia authoring too into a reading and authoring environment that incorporates the recent cultural shifts instigated by social networking and software." I have not used Sophie myself, but was it unstable? Thanks! -- View this message in context: http://n4.nabble.com/Sophie-2-0-ported-to-Java-Why-tp443676p443676.html Sent from the Squeak - Dev mailing list archive at Nabble.com.

1 0

Box-Admins Team Report for October, 2009
by Ken Causey 04 Nov '09

04 Nov '09

The box-admins team is responsible for seeing that the community shared server is maintained and setting up any new features to be hosted on that server. For October, 2009: This has been the most tumultuous month since October of 2008, in fact there was very strong sense of living a rerun, with multiple server outages requiring server resets. I posted multiple status notices: http://squeakboard.wordpress.com/2009/10/18/squeak-orgsqueakfoundation-org-… http://squeakboard.wordpress.com/2009/10/25/squeak-org-server-down-again/ http://squeakboard.wordpress.com/2009/10/28/server-down-once-again/ http://squeakboard.wordpress.com/2009/10/30/latest-on-server-outages/ The status remains the same as the last notice: good. The server has now been up for a recent record of 5 days, 14 hours, and 48 minutes as of the moment of writing this note. In any case this server is decidedly aging and the Squeak Oversight Board along with this team will be investigating options to replace and upgrade/improve our hosting capabilities in the coming months. http://www.squeak.org/Foundation/SupportSqueak/ and purchasing a Squeak badge. Note that a donation here may be used for other Squeak community services including but not limited to development, development infrastructure, and promotion of Squeak. The services being hosted have continued to grow and include (but are not limited to): http://www.squeak.org/ SqueakMap - http://map.squeak.org/ http://ftp.squeak.org/ http://lists.squeakfoundation.org/ http://source.squeak.org/ http://map.squeak.org/ https://svn.squeak.org/ http://update.squeakfoundation.org/ http://wiki.squeak.org/ http://bugs.squeak.org/ and DNS services for squeak.org and squeakfoundation.org domain names. Ken

1 0

printing collections
by Nicolas Cellier 04 Nov '09

04 Nov '09

While at isSelfEvaluating, I wonder why we do not abuse {..} notation a bit more. Instead of printing 'aSet( 0@0 1@2 )' we could just make it evaluate proof '{0@0. 1@2} asSet' Of course, with limited stream, we take the risk of loosing trailing ' asSet' information. So we can also print using a less nice but still evaluating 'Set newFrom: {0@0. 1@2}'. As a bonus, generalizing this behaviour might also eliminate a few bytecodes and methods from the Kernel. Same for storeOn: What do you think ?

4 13

a pair of suggestions
by Chris Cunnington 04 Nov '09

04 Nov '09

Hi all, I have a pair of suggestions. The first is about the Linux download from the squeak.org homepage. The second is about the RFB package from Squeak Map Package Loader. When a person downloads the Croquet SDK they only need to double click on a shell script to start the image for Linux. This one click solution is the same as the Mac and PC. At the moment if you hit the Linux link, you get a VM and some instructions. The person then needs to hunt for what they need and then install it. Quelle drag. I've created a linux.zip package that is a one click solution, hit a shell script to start the image in Linux Squeak. It's pretty easy to bundle what they need and create a shell script like (I suppose) Andreas did. The other thing is whether there is a RFB maintainer. RFB is really, really fun. Downloading the package from the SMPL (and I suppose this problem is the same in Squeaksource) and finding it doesn't work is a big let down. Seaside comes with a WAVNCController app ready to register as an application, too. To make Ian P.'s RFB work online, I found a forum with some code, copied it into a file, and used it as a patch. It seems to me that this is the kind of thing a maintainer is/ought to be doing. I take it that the RFB package worked in 2006 and then broke with new Squeaks, and that's why I needed this hard to find piece of code. (Which I haven't bothered to try and understand, by the way). If nobody is maintaining RFB, I'm willing to be the maintainer. If you need a maintainer that understands RFB top to bottom, then I'm not qualified, nor will I bone up on the topic just for the job. But if being a maintainer means that people point the news to me and I do what's necessary to see that the latest, working version is available, then I'd like to help. Chris

3 2

JPEG plugin.
by John M McIntosh 04 Nov '09

04 Nov '09

I'm looking at the JPEG plugin and wondering? (a) Should I convert that to using the os-x platforms JPEG implementation, versus our usage of the source from the "Independent JPEG Group's software" ? (b) Should we consider upgrading the VMs from version "6b 27- Mar-1998" to the more current "release 7 of 27-Jun-2009" since we store the source in the VM SVN tree at the moment. I've not looked but (a) Apple might have a private implementation that could be faster since it would likely use hardware vector services. (b) Relying on the system libraries for conversion means fixes for security issues get prompt attention, versus our non-attention now. -- = = = ======================================================================== John M. McIntosh <johnmci(a)smalltalkconsulting.com> Twitter: squeaker68882 Corporate Smalltalk Consulting Ltd. http://www.smalltalkconsulting.com = = = ========================================================================

2 3

Daily Commit Log
by commits＠source.squeak.org 04 Nov '09

04 Nov '09

Changes to Trunk (http://source.squeak.org/trunk.html) in the last 24 hours: http://lists.squeakfoundation.org/pipermail/packages/2009-November/001729.h… Name: CollectionsTests-nice.107 Ancestors: CollectionsTests-nice.106 Add a test denying a WeakArray isLiteral =============================================

1 0

The Trunk: CollectionsTests-nice.107.mcz
by commits＠source.squeak.org 03 Nov '09

03 Nov '09

Nicolas Cellier uploaded a new version of CollectionsTests to project The Trunk: http://source.squeak.org/trunk/CollectionsTests-nice.107.mcz ==================== Summary ==================== Name: CollectionsTests-nice.107 Author: nice Time: 3 November 2009, 10:24:26 am UUID: 065a79ad-d5bf-4ec4-a882-683d2d307f73 Ancestors: CollectionsTests-nice.106 Add a test denying a WeakArray isLiteral =============== Diff against CollectionsTests-nice.106 =============== Item was changed: ----- Method: ArrayTest>>testIsLiteral (in category 'testing') ----- testIsLiteral "We work with a copy of literalArray, to avoid corrupting the code." + + | aLiteralArray | + aLiteralArray := literalArray copy. + self assert: aLiteralArray isLiteral. + aLiteralArray at: 1 put: self class. + self deny: aLiteralArray isLiteral. + self deny: (literalArray as: WeakArray) isLiteral description: 'instances of Array subclasses are not literal'.! - | l | - l := literalArray copy. - self assert: l isLiteral. - l at: 1 put: self class. - self deny: l isLiteral!

1 0

Jump to page: