Squeak-dev May 1999

squeak-dev@lists.squeakfoundation.org

126 participants
289 discussions

MessageTally (2.3)
by Andres Valloud 30 May '99

30 May '99

Hi. MessageTally doesn't tally a block that takes long to execute. It keeps running but the block does nothing and you can move the cursor and do things. While in this state, MessageTally new close takes a big while, and finally says that #terminate is not understood in the method #spyEvery:on:. The variable Timer is gone, now it's nil. I have exactly 100 alive instances of MessageTally (as informed by MessageTally allInstances inspect). Some of them have all their variables set to nil so #printOn: fails. None of them is working or doing something. At least one of those has no references to it when you ask for "objects referencing (or pointing to?) this value" in the inspector. Ideas? Andres.

1 0

Re: Event triggers, smart semaphores?
by Bob Arning 29 May '99

29 May '99

On Fri, 28 May 1999 10:17:46 +0930 "Peter Smet" <peter.smet(a)flinders.edu.au> wrote: >It would be good to have a "smart semaphore" that forced >outside callers to wait, but let self invoke multiple invocations >of a message. I would appreciate it if someone could point >out some code to identify if the invoker of the currently >executing message is self or not..... Hi Peter, You could try something like the code below. Cheers, Bob 'From Squeak 2.3 of January 14, 1999 on 29 May 1999 at 1:53:30 pm'! "Change Set: RecursiveSema Date: 29 May 1999 Author: Bob Arning I am a variation of the conventional Semaphore that allows consecutive waits without intervening signals to proceed as long as the process which issues the subsequent waits is the one which issued the first wait. This will permit recursive (or other) routines to exclude other processes from a block of code while allowing the current process to keep running. As soon as the number of signals equals the number of waits, another process may proceed. Test examples are included as class methods."! Object subclass: #RecursiveSemaphore instanceVariableNames: 'holdingProcess waitCount semaphore ' classVariableNames: '' poolDictionaries: '' category: 'Kernel-Processes'! !RecursiveSemaphore commentStamp: '<historical>' prior: 0! I am a variation of the conventional Semaphore that allows consecutive waits without intervening signals to proceed as long as the process which issues the subsequent waits is the one which issued the first wait. This will permit recursive (or other) routines to exclude other processes from a block of code while allowing the current process to keep running. As soon as the number of signals equals the number of waits, another process may proceed. Test examples are included as class methods.! !RecursiveSemaphore methodsFor: 'as yet unclassified' stamp: 'RAA 5/29/1999 13:27'! critical: mutuallyExcludedBlock "Evaluate mutuallyExcludedBlock only if the receiver is not currently in the process of running the critical: message. If the receiver is, evaluate mutuallyExcludedBlock after the other critical: message is finished." | blockValue | self wait. blockValue _ mutuallyExcludedBlock value. self signal. ^blockValue! ! !RecursiveSemaphore methodsFor: 'as yet unclassified' stamp: 'RAA 5/29/1999 13:14'! forMutualExclusion semaphore _ Semaphore forMutualExclusion. waitCount _ 0.! ! !RecursiveSemaphore methodsFor: 'as yet unclassified' stamp: 'RAA 5/29/1999 13:20'! signal waitCount _ waitCount - 1. waitCount <= 0 ifTrue: [ holdingProcess _ nil. waitCount _ 0. semaphore signal ]. ! ! !RecursiveSemaphore methodsFor: 'as yet unclassified' stamp: 'RAA 5/29/1999 13:12'! wait holdingProcess == Processor activeProcess ifFalse: [ semaphore wait. waitCount _ 0. holdingProcess _ Processor activeProcess. ]. waitCount _ waitCount + 1. ! ! !RecursiveSemaphore class reorganize! ('instance creation' forMutualExclusion) ('test example' reportAndWait: test test1: test2: test3: test4:) ! !RecursiveSemaphore class methodsFor: 'instance creation' stamp: 'RAA 5/29/1999 13:14'! forMutualExclusion ^self new forMutualExclusion! ! !RecursiveSemaphore class methodsFor: 'test example' stamp: 'RAA 5/29/1999 13:04'! reportAndWait: aString Transcript show: 'process = ',Processor activeProcess hash printString, ' method = ',aString; cr. (Delay forSeconds: 1) wait. ! ! !RecursiveSemaphore class methodsFor: 'test example' stamp: 'RAA 5/29/1999 13:34'! test " RecursiveSemaphore test " | semaphore | semaphore _ self forMutualExclusion. 3 timesRepeat: [self test1: semaphore].! ! !RecursiveSemaphore class methodsFor: 'test example' stamp: 'RAA 5/29/1999 13:34'! test1: semaphore [ self reportAndWait: 'test1'. self test2: semaphore. ] fork.! ! !RecursiveSemaphore class methodsFor: 'test example' stamp: 'RAA 5/29/1999 13:35'! test2: semaphore semaphore critical: [ self reportAndWait: 'test2'. self test3: semaphore. ].! ! !RecursiveSemaphore class methodsFor: 'test example' stamp: 'RAA 5/29/1999 13:35'! test3: semaphore semaphore critical: [ self reportAndWait: 'test3'. self test4: semaphore. ].! ! !RecursiveSemaphore class methodsFor: 'test example' stamp: 'RAA 5/29/1999 13:35'! test4: semaphore semaphore critical: [ self reportAndWait: 'test4'. ].! !

1 0

Re: [BUG] collection with: do: on uneven sizes problem...
by Peter Smet 29 May '99

29 May '99

Travis, The context of my problem was to find out if one collection is a prefix of another collection. eg #(1 2) prefix: #(1 2 3) I wanted it to return false for cases like: #(1 2 3) prefix: #(1 2). I suspect there are many cases where it would be more flexible to examine both collections until one or the other reaches its end. It's hard in this case to say what the correct behaviour is, since it depends on context and programmer intent. I'm not sure what you are doing with the ReadStream - you can only use a ReadStream on a sequencable collection, so using this method at the base of the Collection hierachy is not going to work. (offcourse you can do something like: self asOrderedCollection readStream) I find it frustrating that there are no external iterators for any of the unordered classes. I implemented a good workaround in Dolphin that can produce a readStream on any class that implements do: [ ..] The trick was to embed do: [ ] into a process, held by the readStream object. This process waits on a Mutex in the readStream object, which is only signalled when the readStream receives the #next message. ReadStream on: egDictionary | cursor mutex iterationProcess | iterationProcess := [egDictionary do: [ : each | mutex wait. cursor := each]] fork blah blah.. >> next mutex signal. ^cursor. This works with the usual proviso that you shouldn't modify a collection while iterating over it. If anyone is interested, I could try to port this to Squeak. It would be good if readStream was a message implemented by all collections to return an external cursor on themselves. This would lessen the needs for methods like with: do:, since it is notoriously difficult to use internal iterators to compare several collections. >Peter Smet wrote: > >> I notice that collection with:do: is ok if the >> first collection is smaller than the second collection, >> but fails if the first collection is the larger. >> Since the method is happy to run with >> unequal sizes one way, for the sake of >> consistency, it should do so the other way >> also. >> >> Here is a fix: >> >> with: otherCollection do: twoArgBlock >> "Evaluate twoArgBlock with corresponding elements from this collection and >> otherCollection." >> 1 to: (self size min: otherCollection size) do: >> [:index | >> twoArgBlock value: (self at: index) >> value: (otherCollection at: index)] > >IMO, if the two sizes do not match, an error should be raised. I'm doubting >that anyone is actually reliably relying on leveraging this (mis)feature. > >with: aCollection do: aBlock > | rs | > aCollection size = self size ifFalse: [self error: 'Collection sizes do not >match']. > rs := ReadStream on: aCollection. > self do: [:each | aBlock value: each value: rs next] > > >and reimplemented in SequenceableCollection as > >with: aCollection do: aBlock > | index | > aCollection size = self size ifFalse: [self error: 'Collection sizes do not >match']. > index := 0. > aCollection do: [:each | aBlock value: (self at: (index := index + 1)) >value: each] > > >-- >Travis Griggs (a.k.a. Lord of the Fries) >Member, Fraven Skreiggs Software Collective >Key Technology >P-P-P-Penguin Power! > > >

1 0

Speak speaks, 2nd release
by Hans-Martin Mosner 29 May '99

29 May '99

Hello, this week I've added quite a bit to the speech synth. The architecture has been modified quite a bit and now consists of a processing pipeline where you can switch the elements if you want to experiment. I've integrated number and character name reading. CMUDICT, the freely available pronounciation dictionary, is supported. You will need to get CMUDICT.0.6 from Carnegie-Mellon and scan it using the class CMUDict. The VoiceSynthesizer plugin is not affected, so you can use it as before. Before you file in the new stuff, you should remove the category "Speech Synthesis" because the structure of some classes has changed significantly. For download simplification, I've added my SqSpeech shared lib for the Mac (as a MacBinary file) and Andreas Raab's DLL for windows to the zip file. As before, it can be found at http://www.heeg.de/~hmm/squeak/SpeechSynthesis.zip Have fun, Hans-Martin

1 1

Event triggers, smart semaphores?
by Peter Smet 28 May '99

28 May '99

Mark, I just uploaded all the changes to 2.4b using the 'update from server' command, and saved the image, (I guess that makes it equivalent to 2.4c). I filed in my publish-subscribe change-set and everything appears to be fine - 'PostOffice start' begins testing in the background, and all the messages appear on the Transcript... When you got the error, were you working with a clean image? Does the error occur with 2.4b or 2.4c? One problem I have discovered is 'nested updates'. That is, if one trigger sets off another trigger by one of its subscribers, you will get a call to PostOffice incoming: msg from inside the PostOffice incoming: msg. Since this method uses a mutex critical: the two calls to the method will deadlock each other. (The simple fix is to remove the protected: block from PostOffice: incoming, which I have done) I have found this to be a problem with OO and Semaphores in general - there is so much cross-calling of messages it is difficult to predict when a locked message will call itself. It would be good to have a "smart semaphore" that forced outside callers to wait, but let self invoke multiple invocations of a message. I would appreciate it if someone could point out some code to identify if the invoker of the currently executing message is self or not..... Does this idea of having outside callers wait in line, and have self rudely pushing to the front of the queue make sense? Any other solutions? <irrelevant thought> I just realized that you can use the event system to uphold all the existing dependency relationships. Simply change addDependent: aDependent (in Object) to read: self when: #changed send: #update to: aDependent. plus its variant methods - you get the idea..... </irrelevant thought> Peter >When I try "PostOffice start" I get an error about error: having the wrong >number of parameters in WeakRegistry>>protected: It's a hard error to find >because (a) the error notifier shows up behind all other windows, since >it's running in a background thread (restore the display to see the error >notifier) and (b) the debugger for some reason doesn't show the backtrace >if you choose to debug. > >But maybe this is on purpose? I get a similar kind of error dialog when I >send #run to the Transcript (playing with variations on Peter's examples). > >Mark > >-------------------------- >Mark Guzdial : Georgia Tech : College of Computing : Atlanta, GA 30332-0280 >(404) 894-5618 : Fax (404) 894-0673 : guzdial(a)cc.gatech.edu >http://www.cc.gatech.edu/gvu/people/Faculty/Mark.Guzdial.html > > >

2 1

Re: [BUG] collection with: do: on uneven sizes problem...
by Travis Griggs 28 May '99

28 May '99

Peter Smet wrote: > I notice that collection with:do: is ok if the > first collection is smaller than the second collection, > but fails if the first collection is the larger. > Since the method is happy to run with > unequal sizes one way, for the sake of > consistency, it should do so the other way > also. > > Here is a fix: > > with: otherCollection do: twoArgBlock > "Evaluate twoArgBlock with corresponding elements from this collection and > otherCollection." > 1 to: (self size min: otherCollection size) do: > [:index | > twoArgBlock value: (self at: index) > value: (otherCollection at: index)] IMO, if the two sizes do not match, an error should be raised. I'm doubting that anyone is actually reliably relying on leveraging this (mis)feature. with: aCollection do: aBlock | rs | aCollection size = self size ifFalse: [self error: 'Collection sizes do not match']. rs := ReadStream on: aCollection. self do: [:each | aBlock value: each value: rs next] and reimplemented in SequenceableCollection as with: aCollection do: aBlock | index | aCollection size = self size ifFalse: [self error: 'Collection sizes do not match']. index := 0. aCollection do: [:each | aBlock value: (self at: (index := index + 1)) value: each] -- Travis Griggs (a.k.a. Lord of the Fries) Member, Fraven Skreiggs Software Collective Key Technology P-P-P-Penguin Power!

1 0

How to select a GIF file?
by Craig H. Anderson 28 May '99

28 May '99

I'm working on an app where I want a browser to select a GIF file. I'd like some advice and pointers to examples to look at. This is with Squeak 2.4 on Linux. I'm likely to have a number of beginner questions. Is this the best for forum for them? Thanks

2 1

Re: Folding editor?
by Marc Nijdam 28 May '99

28 May '99

Edward P Luwish <eluwish(a)uswest.com> writes: > This looks like a very good place to start. I had not planned to go > "morphic" but there's always a first time (also, this is supposed to > run on a PDA - I don't know yet how much Squeak will fit on it). > Thank you very much for the quick response. I've also been wondering how much of the current Squeak would still fit onto a PDA. It sure looks like Morphic is currently to "heavy" in memory and cpu requirements to be workable. Marc

2 1

RE: Better solution to Random>>initializa
by agree＠carltonfields.com 28 May '99

28 May '99

> PS: I have translated to Squeak the C code to generate random numbers > with a normal distribution published in "Numerical Recipes in > C", second > edition, and implemented the class NormalRandom. Is it legal > to publish such code? As a lawyer, I can offer a definitive answer: It depends. :-) In practice, I seem to recall that Numerical Recipes had a specific license set forth somewhere in the preamble or at the end of the book, stating how and for what you can use the code; which as I recall was a liberal license. I would start there, and if it lets you do what you want, you're all set. In the alternative, just ask the authors for their OK. (Consent is the single best defense there is to a copyright action.) Saul Teulkowsky and his wife, Roslyn, are old bridge teammates of mine, so I'd be pleased to make an introduction. Otherwise, we get into tricky questions of fair use, which while it doesn't ordinarily get in the way of taking a code snippet from a book, the answer I'd have to give as a lawyer is, "it depends."

1 0

Better solution to Random>>initializa
by Juan Cires Martinez 28 May '99

28 May '99

----- Forwarded message from jcm ----- >From jcm Thu May 27 16:09:17 1999 Subject: Better solution to Random>>initializa To: squeak(a)cs.uiuc.edu (Squeak mailing list) Date: Thu, 27 May 1999 16:09:17 +0200 (METDST) MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 8bit Content-Length: 765 The problem with the initialization of a Random is in the statement: seed := Time millisecondClockValue * self hash. which has a problem when self hash evaluates to zero. The solution I propose is to modify that line: initialize " Set a reasonable Park-Miller starting seed " seed := Time millisecondClockValue * (self hash + 1). a := 16r000041A7 asFloat. " magic constant = 16807 " m := 16r7FFFFFFF asFloat. " magic constant = 2147483647 " q := (m quo: a) asFloat. r := (m \\ a) asFloat. Greetings, Juan. PS: I have translated to Squeak the C code to generate random numbers with a normal distribution published in "Numerical Recipes in C", second edition, and implemented the class NormalRandom. Is it legal to publish such code? ----- End of forwarded message from jcm -----

1 0

Jump to page: