Squeak-dev August 2011

squeak-dev@lists.squeakfoundation.org

55 participants
171 discussions

The Trunk: ST80-nice.131.mcz
by commits＠source.squeak.org 03 Aug '11

03 Aug '11

Nicolas Cellier uploaded a new version of ST80 to project The Trunk: http://source.squeak.org/trunk/ST80-nice.131.mcz ==================== Summary ==================== Name: ST80-nice.131 Author: nice Time: 3 August 2011, 5:38:51.551 pm UUID: a95fbf37-a396-48dd-951d-e64e664a0d29 Ancestors: ST80-cmm.130 cosmetic: use self normalCharacter: rather than self perform: #normalCharacter: =============== Diff against ST80-cmm.130 =============== Item was changed: ----- Method: ParagraphEditor>>dispatchOnCharacter:with: (in category 'parenblinking') ----- dispatchOnCharacter: char with: typeAheadStream "Carry out the action associated with this character, if any. Type-ahead is passed so some routines can flush or use it." | honorCommandKeys result | self clearParens. "mikki 1/3/2005 21:31 Preference for auto-indent on return added." char asciiValue = 13 ifTrue: [ ^Preferences autoIndent ifTrue: [ sensor controlKeyPressed ifTrue: [self normalCharacter: typeAheadStream] ifFalse: [self crWithIndent: typeAheadStream]] ifFalse: [ sensor controlKeyPressed ifTrue: [self crWithIndent: typeAheadStream] ifFalse: [self normalCharacter: typeAheadStream]]]. ((honorCommandKeys := Preferences cmdKeysInText) and: [char = Character enter]) ifTrue: [^ self dispatchOnEnterWith: typeAheadStream]. "Special keys overwrite crtl+key combinations - at least on Windows. To resolve this conflict, assume that keys other than cursor keys aren't used together with Crtl." ((self class specialShiftCmdKeys includes: char asciiValue) and: [char asciiValue < 27]) ifTrue: [^ sensor controlKeyPressed ifTrue: [self perform: (ShiftCmdActions at: char asciiValue + 1) with: typeAheadStream] ifFalse: [self perform: (CmdActions at: char asciiValue + 1) with: typeAheadStream]]. "backspace, and escape keys (ascii 8 and 27) are command keys" ((honorCommandKeys and: [sensor commandKeyPressed]) or: [self class specialShiftCmdKeys includes: char asciiValue]) ifTrue: [^ sensor leftShiftDown ifTrue: [self perform: (ShiftCmdActions at: char asciiValue + 1 ifAbsent: [#noop:]) with: typeAheadStream] ifFalse: [self perform: (CmdActions at: char asciiValue + 1 ifAbsent: [#noop:]) with: typeAheadStream]]. "the control key can be used to invoke shift-cmd shortcuts" (honorCommandKeys and: [sensor controlKeyPressed]) ifTrue: [^ self perform: (ShiftCmdActions at: char asciiValue + 1 ifAbsent: [#noop:]) with: typeAheadStream]. + result := self normalCharacter: typeAheadStream. - - result := self perform: #normalCharacter: with: typeAheadStream. (')]}' includes: char) ifTrue: [self blinkPrevParen: char ]. ^result!

1 0

Squeak Oversight Board minutes – 8/2/11
by Chris Cunnington 03 Aug '11

03 Aug '11

Squeak Oversight Board minutes – 8/2/11 Attending: Chris Muller, Bert Freudenberg, Colin Putney, Randal Schwartz, Levente Uzonyi, Jecel Assumpacao Jr., Chris Cunnington - Jecel is now the contact person for SFC and will look into paying Bert what he is owed for his ESUG 2010 trip and Squeak Germany for hosting costs - the nature of relationships for the management of the box and web site was tentatively explored - a light review of yearly Squeak costs was explored - discussion about the possibility of removing unused vm sources from immediate visibility on the squeakvm.org svn repository, so as not to confuse newcomers. It was agreed that if a branch uses autotools, then it’s old, and should be removed from immediate sight - agenda item for next time: Spoon first impressions

1 0

The Trunk: Tools-fbs.372.mcz
by commits＠source.squeak.org 03 Aug '11

03 Aug '11

Levente Uzonyi uploaded a new version of Tools to project The Trunk: http://source.squeak.org/trunk/Tools-fbs.372.mcz ==================== Summary ==================== Name: Tools-fbs.372 Author: fbs Time: 3 August 2011, 8:26:08.804 am UUID: cda1854b-3aff-43e6-b3a3-bd88c3b9e5de Ancestors: Tools-bf.371 When you rename a system category, you should have the newly-named system category selected after the rename. =============== Diff against Tools-bf.371 =============== Item was changed: ----- Method: Browser>>renameSystemCategory (in category 'system category functions') ----- renameSystemCategory "Prompt for a new category name and add it before the current selection, or at the end if no current selection" | oldSelection newName | oldSelection := self selectedSystemCategory. oldSelection isNil ifTrue: [^ self]. "no selection" self okToChange ifFalse: [^ self]. newName := self request: 'Please type new category name' initialAnswer: oldSelection. newName isEmpty ifTrue: [^ self] ifFalse: [newName := newName asSymbol]. oldSelection = newName ifTrue: [^ self]. systemOrganizer renameCategory: oldSelection toBe: newName. + self selectSystemCategory: newName. - self selectSystemCategory: oldSelection. self changed: #systemCategoryList.!

1 0

ESUG 2011 Innovation Technology Awards: Reminder
by jdelgado 03 Aug '11

03 Aug '11

Hi, Again, a reminder: ESUG 2011 Innovation Technology Awards http://esug.org/Conferences/2011/InnovationTechnologyAwards New Deadline: August 15. Have you written innovative Smalltalk? The developers of any Smalltalk-based software can enter by submitting a 3-5min video of their software. Entrants demo their systems in a session at the end of Monday August 22nd 2011, after which the conference attendees vote. (A vote consists in providing a sorted list of 3 preferred pieces of software.) The top 3 teams with the most innovative software will receive, respectively, 500 Euros, 300 Euros and 200 Euros. The results are announced in the Awards ceremony that is held during the social event. Applicants should provide the following information on the conference registration server. Once you have registered your personal info, an "Awards" menu allows submitting your software. You can provide this information when you first register, or login again later and update your details. Info to provide: Name of the software Licence information (free, or shareware, or a commercial) Name of the Smalltalk dialect used Names, Affiliation and Country of developers URL for a video (3-5 min) showing a screencast / demo of the software URL for downloading the software or at least a runnable demo Deadline: August 15 Any Smalltalker can apply (students, companies, ...). The presented piece of code/software should be written in Smalltalk or directly support Smalltalk (e.g. Smalltalk VM). All Smalltalk dialects are accepted. The applicant should own the copyright/copyleft of presented code or at least be the official representative of the copyright/copyleft owner. The software should be available on-line for evaluation (time-limited evaluation version, demo version, free for non-commercial use ...). Bests Jordi -- View this message in context: http://forum.world.st/ESUG-2011-Innovation-Technology-Awards-Reminder-tp371… Sent from the Squeak - Dev mailing list archive at Nabble.com.

1 0

something we lost with events [was Optimizing RunArray]
by Nicolas Cellier 03 Aug '11

03 Aug '11

I've got the feeling that we've lost one feature: When you're typing super fast, characters would be inserted several at once in a st80 ParagraphEditor That typically happens when the displaying or more probably the paragraph re-composition is slow. The composition/display cycle would occur after first key - giving the impression of a non responsive UI - but then consecutive key strokes would flow all in once thanks to the while loops in #readKeyboard and the buffered Queue of events. Not the most interactive UI ever, but now things are worse with current event implementation. There is a one by one keystroke / composition / display, and the result is a less responsive UI than polling ! If you edit a Smalltalk method, no problem, but try editing your 45 MByte change log just to see how deadly it feels, super fast means 0.2 Hz on my machine. I wonder how we could restore this feature... Nicolas

3 2

The Inbox: Tools-fbs.372.mcz
by commits＠source.squeak.org 03 Aug '11

03 Aug '11

A new version of Tools was added to project The Inbox: http://source.squeak.org/inbox/Tools-fbs.372.mcz ==================== Summary ==================== Name: Tools-fbs.372 Author: fbs Time: 3 August 2011, 8:26:08.804 am UUID: cda1854b-3aff-43e6-b3a3-bd88c3b9e5de Ancestors: Tools-bf.371 When you rename a system category, you should have the newly-named system category selected after the rename. =============== Diff against Tools-bf.371 =============== Item was changed: ----- Method: Browser>>renameSystemCategory (in category 'system category functions') ----- renameSystemCategory "Prompt for a new category name and add it before the current selection, or at the end if no current selection" | oldSelection newName | oldSelection := self selectedSystemCategory. oldSelection isNil ifTrue: [^ self]. "no selection" self okToChange ifFalse: [^ self]. newName := self request: 'Please type new category name' initialAnswer: oldSelection. newName isEmpty ifTrue: [^ self] ifFalse: [newName := newName asSymbol]. oldSelection = newName ifTrue: [^ self]. systemOrganizer renameCategory: oldSelection toBe: newName. + self selectSystemCategory: newName. - self selectSystemCategory: oldSelection. self changed: #systemCategoryList.!

1 0

Re: [Pharo-project] [squeak-dev] Re: Optimizing RunArray
by Nicolas Cellier 03 Aug '11

03 Aug '11

I have tried again with huge text (my 45 MByte change log). Main contributor seems to be Text composition. A marginal gain (2 to 3% when resizing the window) is to remove the indirection in TextStopConditions, and replace instance variable stops with direct integer slots (variableSubclass) No guaranty the fileIn of attachment will work, modifying Text processing is dangerous... Nicolas 2011/8/2 Stéphane Ducasse <stephane.ducasse(a)inria.fr>: > Yes I was wondering about the same (for athens too :)) > > Stef > > On Aug 2, 2011, at 2:16 PM, Igor Stasenko wrote: > >> Does it impacts the text rendering/processing speed? >> >> >> On 2 August 2011 12:11, Nicolas Cellier >> <nicolas.cellier.aka.nice(a)gmail.com> wrote: >>> To complete myself, the fast #collect: already exists and is named >>> #mapValues: except that it modifies the RunArray in place and also >>> won't coalesce... >>> >>> I also gain a huge factor for #collect:as: be defining this method: >>> >>> RunArray>>fillFrom: aCollection with: aBlock >>> "Evaluate aBlock with each of aCollections's elements as the argument. >>> Collect the resulting values into self. Answer self." >>> | newRuns newValues lastLength lastValue | >>> newRuns := (Array new: aCollection size) writeStream. >>> newValues := (Array new: aCollection size) writeStream. >>> lastLength := 0. >>> lastValue := Object new. >>> aCollection do: [:each | >>> | value | >>> value := aBlock value: each. >>> lastValue = value >>> ifTrue: [lastLength := lastLength + 1] >>> ifFalse: >>> [lastLength > 0 >>> ifTrue: >>> [newRuns nextPut: lastLength. >>> newValues nextPut: lastValue]. >>> lastLength := 1. >>> lastValue := value]]. >>> lastLength > 0 >>> ifTrue: >>> [newRuns nextPut: lastLength. >>> newValues nextPut: lastValue]. >>> self setRuns: newRuns contents setValues: newValues contents >>> >>> [ (Array new: 1000) collect: [:e | 4 atRandom] as: RunArray] bench. >>> BEFORE: '25.1 per second.' >>> AFTER: '1,080 per second.' >>> >>> It's worth a few lines of code. >>> >>> Nicolas >>> >>> 2011/8/2 Nicolas Cellier <nicolas.cellier.aka.nice(a)gmail.com>: >>>> I played a bit with RunArray, and found some un-optimized features. >>>> First, I don't know why RunArray is an ArrayedCollection. It cannot >>>> #add: but it can #addFirst: and #addLast:. >>>> It cannot #add:withOccurrences: but it can #addLast:times:. Why >>>> inventing new selectors for old behaviours ? >>>> These operations will cost a realloc it the last value is different, >>>> so the underlying runs/values could better be an OrderedCollection if >>>> these operations are used often. >>>> A RunArray cannot remove at all. >>>> Very weird collection species, I don't like the implementation too much. >>>> >>>> Then, #do: loops could be far faster. They rely on ArrayedCollection >>>> which inlines do: loops with #to:do: and #at: >>>> But #at: is not that fast. Scanning the runs and counting elements >>>> would result in a n^2 cost. >>>> Fortunately there is a cache lastIndex,lastRun,lastOffset to keep a cost n. >>>> Nonetheless, all the tests cost, and the loop is suboptimal. >>>> Let use see: >>>> >>>> version 1: >>>> RunArray>>fastDo: aBlock >>>> runs with: values do: [:r :v | >>>> r timesRepeat: [aBlock value: v]]. >>>> >>>> | tmp | >>>> tmp := ((Array new: 1000) collect: [:e | 4 atRandom]) as: RunArray. >>>> { >>>> [ tmp do: [:e |]] bench. >>>> [ tmp fastDo: [:e |]] bench. >>>> } >>>> #('3,220 per second.' '6,290 per second.') >>>> >>>> But timesRepeat: is slow, it is unoptimized by the compiler and costs >>>> a message send. >>>> I think we should implement BlockClosure>>repeat: and optimize that >>>> call in Compiler. >>>> But let's not do it, and rather inline by ourself: >>>> >>>> version 2: >>>> runs with: values do: [:r :v | >>>> 1 to: r do: [:i | aBlock value: v]]. >>>> >>>> | tmp | >>>> tmp := ((Array new: 1000) collect: [:e | 4 atRandom]) as: RunArray. >>>> { >>>> [ tmp do: [:e |]] bench. >>>> [ tmp do2: [:e |]] bench. >>>> } >>>> #('3,070 per second.' '25,500 per second.') >>>> >>>> We can even inline the with:do: loop itself: >>>> version 3: >>>> 1 to: runs size do: [:i | >>>> | r v | >>>> v := values at: i. >>>> r := runs at: i. >>>> [( r := r - 1) >= 0] >>>> whileTrue: [aBlock value: v]]. >>>> >>>> | tmp | >>>> tmp := ((Array new: 1000) collect: [:e | 4 atRandom]) as: RunArray. >>>> { >>>> [ tmp do: [:e |]] bench. >>>> [ tmp do2: [:e |]] bench. >>>> } >>>> #('3,370 per second.' '32,200 per second.') >>>> >>>> Now the operation I wanted to use was reverseDo: so I implemented: >>>> RunArray>>fastReverseDo: aBlock >>>> | i | >>>> i := runs size. >>>> [i > 0] >>>> whileTrue: >>>> [ | r v | >>>> v := values at: i. >>>> r := runs at: i. >>>> i := i - 1. >>>> [( r := r - 1) >= 0] >>>> whileTrue: [aBlock value: v]]. >>>> | tmp | >>>> tmp := ((Array new: 1000) collect: [:e | 4 atRandom]) as: RunArray. >>>> { >>>> [ tmp reverseDo: [:e |]] bench. >>>> [ tmp reverseDo2: [:e |]] bench. >>>> } >>>> #('83.9 per second.' '32,600 per second.') >>>> >>>> Ouch! The cache is missing a lot of indices, and our loop turns into a n^2 cost. >>>> I know, premature optimization bla bla bla, but a factor x400 is worth >>>> some inlining no? >>>> >>>> I guess these features are never used. >>>> By now RunArray is kind of private utility for Text implementation. >>>> But it could / should be generic. >>>> >>>> I also have proposals for count: / select: / collect:. etc... >>>> It would be to evaluate the block only once per group of values. >>>> For example >>>> RunArray>>collect: aBlock >>>> "Beware, the block will be evaluated only once per group of values." >>>> ^(self class runs: (runs collect: aBlock) contents values: values >>>> copy) coalesce >>>> But that's controversial, it would make the RunArray behave >>>> differently if the block has side effects... >>>> >>>> | i tmp tmp2 tmp3 | >>>> tmp := ((Array new: 1000) collect: [:e | 4 atRandom]). >>>> i := 0. >>>> tmp2 := tmp collect: [:e | i := i + 1]. >>>> i := 0. >>>> tmp3 := (tmp as: RunArray) collect: [:e | i := i + 1]. >>>> tmp2 = tmp3 asArray >>>> ==> false >>>> >>>> Nicolas >>>> >>> >>> >> >> >> >> -- >> Best regards, >> Igor Stasenko AKA sig. >> > > >

3 5

Optimizing RunArray
by Nicolas Cellier 02 Aug '11

02 Aug '11

I played a bit with RunArray, and found some un-optimized features. First, I don't know why RunArray is an ArrayedCollection. It cannot #add: but it can #addFirst: and #addLast:. It cannot #add:withOccurrences: but it can #addLast:times:. Why inventing new selectors for old behaviours ? These operations will cost a realloc it the last value is different, so the underlying runs/values could better be an OrderedCollection if these operations are used often. A RunArray cannot remove at all. Very weird collection species, I don't like the implementation too much. Then, #do: loops could be far faster. They rely on ArrayedCollection which inlines do: loops with #to:do: and #at: But #at: is not that fast. Scanning the runs and counting elements would result in a n^2 cost. Fortunately there is a cache lastIndex,lastRun,lastOffset to keep a cost n. Nonetheless, all the tests cost, and the loop is suboptimal. Let use see: version 1: RunArray>>fastDo: aBlock runs with: values do: [:r :v | r timesRepeat: [aBlock value: v]]. | tmp | tmp := ((Array new: 1000) collect: [:e | 4 atRandom]) as: RunArray. { [ tmp do: [:e |]] bench. [ tmp fastDo: [:e |]] bench. } #('3,220 per second.' '6,290 per second.') But timesRepeat: is slow, it is unoptimized by the compiler and costs a message send. I think we should implement BlockClosure>>repeat: and optimize that call in Compiler. But let's not do it, and rather inline by ourself: version 2: runs with: values do: [:r :v | 1 to: r do: [:i | aBlock value: v]]. | tmp | tmp := ((Array new: 1000) collect: [:e | 4 atRandom]) as: RunArray. { [ tmp do: [:e |]] bench. [ tmp do2: [:e |]] bench. } #('3,070 per second.' '25,500 per second.') We can even inline the with:do: loop itself: version 3: 1 to: runs size do: [:i | | r v | v := values at: i. r := runs at: i. [( r := r - 1) >= 0] whileTrue: [aBlock value: v]]. | tmp | tmp := ((Array new: 1000) collect: [:e | 4 atRandom]) as: RunArray. { [ tmp do: [:e |]] bench. [ tmp do2: [:e |]] bench. } #('3,370 per second.' '32,200 per second.') Now the operation I wanted to use was reverseDo: so I implemented: RunArray>>fastReverseDo: aBlock | i | i := runs size. [i > 0] whileTrue: [ | r v | v := values at: i. r := runs at: i. i := i - 1. [( r := r - 1) >= 0] whileTrue: [aBlock value: v]]. | tmp | tmp := ((Array new: 1000) collect: [:e | 4 atRandom]) as: RunArray. { [ tmp reverseDo: [:e |]] bench. [ tmp reverseDo2: [:e |]] bench. } #('83.9 per second.' '32,600 per second.') Ouch! The cache is missing a lot of indices, and our loop turns into a n^2 cost. I know, premature optimization bla bla bla, but a factor x400 is worth some inlining no? I guess these features are never used. By now RunArray is kind of private utility for Text implementation. But it could / should be generic. I also have proposals for count: / select: / collect:. etc... It would be to evaluate the block only once per group of values. For example RunArray>>collect: aBlock "Beware, the block will be evaluated only once per group of values." ^(self class runs: (runs collect: aBlock) contents values: values copy) coalesce But that's controversial, it would make the RunArray behave differently if the block has side effects... | i tmp tmp2 tmp3 | tmp := ((Array new: 1000) collect: [:e | 4 atRandom]). i := 0. tmp2 := tmp collect: [:e | i := i + 1]. i := 0. tmp3 := (tmp as: RunArray) collect: [:e | i := i + 1]. tmp2 = tmp3 asArray ==> false Nicolas

4 6

SS down?
by Alexander Lazarević 02 Aug '11

02 Aug '11

http://www.downforeveryoneorjustme.com/squeaksource.com Alex

3 2

Contexts must only be created with newForMethod: ...
by Chris Muller 01 Aug '11

01 Aug '11

.. but #newForMethod: requires more in the API than it really uses. It only checks the methods #frameSize to determine what basicSized MethodContext to create - it doesn't actually set its method from the one passed in. newForMethod: aMethod --- (comment snipped for brevity) --- ^ super basicNew: aMethod frameSize So, would it be ok to relax MethodContext class>>basicNew: to something like: basicNew: size ^ (size = CompiledMethod smallFrameSize or: [ size = CompiledMethod fullframeSize ]) ifTrue: [ super basicNew: size ] ifFalse: [ self error: 'Contexts must be ', CompiledMethod smallFrameSize, ' or ', CompiledMethod fullframeSize, ' bytes.' ]

2 1

Jump to page: