Squeak-dev February 2015

squeak-dev@lists.squeakfoundation.org

45 participants
152 discussions

Voting System Down - Should be back up tomorrow. Voted today? Pease check your votes tomorrow!
by Ron Teitelbaum 18 Feb '15

18 Feb '15

Hi All, I heard back from Cornell. The voting system is down and has been since 4pm EDT (9pm UTC). It is possible you will need to revote if you voted near that time. I will check the system tomorrow and report back or if you see the system is up and running please post to squeak-dev. When the system is up and you voted today please check your vote. Revote if necessary. Sorry about the trouble. Thank you for your participation in the Squeak Oversight Board 2015 election. All the best, Ron Teitelbaum >From Cornell: Hi Ron, We had a hard disk failure but we have backups. Unfortunately the replacement disk also went bad immediately (Go Seagate!). Hopefully it will be back up tomorrow. A small number of people might need to vote again if they voted just before the system went down around 4pm (the system will be able to tell them if it has registered their votes). Apologies for the inconvenience.

2 2

The Trunk: KernelTests-eem.289.mcz
by commits＠source.squeak.org 18 Feb '15

18 Feb '15

Eliot Miranda uploaded a new version of KernelTests to project The Trunk: http://source.squeak.org/trunk/KernelTests-eem.289.mcz ==================== Summary ==================== Name: KernelTests-eem.289 Author: eem Time: 17 February 2015, 4:12:02.897 pm UUID: 64e0435e-3bad-413e-9523-d0d35493ee61 Ancestors: KernelTests-eem.288 Fix the 64-bitized testCreationFromBytes tests. The read bytes need to be reversed. =============== Diff against KernelTests-eem.288 =============== Item was changed: ----- Method: IntegerTest>>testCreationFromBytes1 (in category 'tests - instance creation') ----- testCreationFromBytes1 "self run: #testCreationFromBytes1" "it is illegal for a LargeInteger to be less than SmallInteger maxVal." "here we test that Integer>>byte!!byte2:byte3:byte4: resconstructs SmallInteger maxVal as an instance of SmallInteger. " | maxSmallInt hexString builtInteger bytes | maxSmallInt := SmallInteger maxVal. hexString := maxSmallInt printStringHex. hexString size odd ifTrue: [hexString := '0', hexString]. self assert: hexString size / 2 = maxSmallInt digitLength. + bytes := ((1 to: hexString size by: 2) collect: + [:i| Number readFrom: (hexString copyFrom: i to: i + 1) base: 16]) reversed. - bytes := (1 to: hexString size by: 2) collect: - [:i| Number readFrom: (hexString copyFrom: i to: i + 1) base: 16]. builtInteger := bytes size > 4 ifTrue: [Integer byte1: (bytes at: 1) byte2: (bytes at: 2) byte3: (bytes at: 3) byte4: (bytes at: 4) byte5: (bytes at: 5) byte6: (bytes at: 6) byte7: (bytes at: 7) byte8: (bytes at: 8)] ifFalse: [Integer byte1: (bytes at: 1) byte2: (bytes at: 2) byte3: (bytes at: 3) byte4: (bytes at: 4)]. self assert: builtInteger = maxSmallInt. self assert: builtInteger class = SmallInteger ! Item was changed: ----- Method: IntegerTest>>testCreationFromBytes2 (in category 'tests - instance creation') ----- testCreationFromBytes2 "self run: #testCreationFromBytes2" "it is illegal for a LargeInteger to be less than SmallInteger maxVal." "here we test that Integer>>byte!!byte2:byte3:byte4: resconstructs (SmallInteger maxVal + 1) as an instance of LargePositiveInteger. " | maxSmallInt hexString builtInteger bytes | maxSmallInt := SmallInteger maxVal. hexString := (maxSmallInt + 1) printStringHex. hexString size odd ifTrue: [hexString := '0', hexString]. self assert: hexString size / 2 = maxSmallInt digitLength. + bytes := ((1 to: hexString size by: 2) collect: + [:i| Number readFrom: (hexString copyFrom: i to: i + 1) base: 16]) reversed. - bytes := (1 to: hexString size by: 2) collect: - [:i| Number readFrom: (hexString copyFrom: i to: i + 1) base: 16]. builtInteger := bytes size > 4 ifTrue: [Integer byte1: (bytes at: 1) byte2: (bytes at: 2) byte3: (bytes at: 3) byte4: (bytes at: 4) byte5: (bytes at: 5) byte6: (bytes at: 6) byte7: (bytes at: 7) byte8: (bytes at: 8)] ifFalse: [Integer byte1: (bytes at: 1) byte2: (bytes at: 2) byte3: (bytes at: 3) byte4: (bytes at: 4)]. self assert: builtInteger = (maxSmallInt + 1). self deny: builtInteger class = SmallInteger ! Item was changed: ----- Method: IntegerTest>>testCreationFromBytes3 (in category 'tests - instance creation') ----- testCreationFromBytes3 "self run: #testCreationFromBytes3" "it is illegal for a LargeInteger to be less than SmallInteger maxVal." "here we test that Integer>>byte!!byte2:byte3:byte4: resconstructs (SmallInteger maxVal - 1) as an instance of SmallInteger. " | maxSmallInt hexString builtInteger bytes | maxSmallInt := SmallInteger maxVal. hexString := (maxSmallInt - 1) printStringHex. hexString size odd ifTrue: [hexString := '0', hexString]. self assert: hexString size / 2 = maxSmallInt digitLength. + bytes := ((1 to: hexString size by: 2) collect: + [:i| Number readFrom: (hexString copyFrom: i to: i + 1) base: 16]) reversed. - bytes := (1 to: hexString size by: 2) collect: - [:i| Number readFrom: (hexString copyFrom: i to: i + 1) base: 16]. builtInteger := bytes size > 4 ifTrue: [Integer byte1: (bytes at: 1) byte2: (bytes at: 2) byte3: (bytes at: 3) byte4: (bytes at: 4) byte5: (bytes at: 5) byte6: (bytes at: 6) byte7: (bytes at: 7) byte8: (bytes at: 8)] ifFalse: [Integer byte1: (bytes at: 1) byte2: (bytes at: 2) byte3: (bytes at: 3) byte4: (bytes at: 4)]. self assert: builtInteger = (maxSmallInt - 1). self assert: builtInteger class = SmallInteger !

1 0

The Trunk: Kernel-ul.904.mcz
by commits＠source.squeak.org 17 Feb '15

17 Feb '15

Levente Uzonyi uploaded a new version of Kernel to project The Trunk: http://source.squeak.org/trunk/Kernel-ul.904.mcz ==================== Summary ==================== Name: Kernel-ul.904 Author: ul Time: 17 February 2015, 1:40:49.811 pm UUID: 7e873979-7e22-4abb-bd69-fa6d72f8c7fc Ancestors: Kernel-nice.902 Avoid the frequent time checks, and the clock rollover bug in BlockClosure >> #bench. Introduced BlockClosure >> #benchFor:, a variant of #bench, which takes the duration of the benchmark as its parameter. Added more information about the running times to the answer. E.g.: [ 10 factorial ] benchFor: 0.1 seconds. "'7,990,000 per second. 125 nanoseconds per run.'" [ 100 factorial ] benchFor: 10 seconds. "'19,900 per second. 50.3 microseconds per run.'" [ 1000 factorial ] bench. "'378 per second. 2.65 milliseconds per run.'" [ 20000 factorial ] bench. "'0.843 per second. 1.19 seconds per run.'" =============== Diff against Kernel-nice.902 =============== Item was changed: ----- Method: BlockClosure>>bench (in category 'evaluating') ----- bench "See how many times I can value in 5 seconds. I'll answer a meaningful description." + ^self benchFor: 5 seconds! - | startTime endTime count roundTo3Digits | - roundTo3Digits := [:num | - | rounded lowDigit | - rounded := (num * 1000) rounded. "round to 1/1000" - lowDigit := (rounded numberOfDigitsInBase: 10) - 3. "keep only first 3 digits" - rounded := rounded roundTo:(10 raisedTo: lowDigit). - (lowDigit >= 3 or: [rounded \\ 1000 = 0]) "display fractional part only when needed" - ifTrue: [(rounded // 1000) asStringWithCommas] - ifFalse: [(rounded / 1000.0) printString]]. - count := 0. - endTime := Time millisecondClockValue + 5000. - startTime := Time millisecondClockValue. - [ Time millisecondClockValue > endTime ] whileFalse: [ self value. count := count + 1 ]. - endTime := Time millisecondClockValue. - ^count = 1 - ifTrue: [ (roundTo3Digits value: (endTime - startTime) / 1000) , ' seconds.' ] - ifFalse: - [ (roundTo3Digits value: (count * 1000) / (endTime - startTime)) , ' per second.' ]! Item was added: + ----- Method: BlockClosure>>benchFor: (in category 'evaluating') ----- + benchFor: aDuration + "See how many times I can value within the given duration. I'll answer a meaningful description." + + | startTime shouldRun count elapsedTime roundTo3Digits delay | + roundTo3Digits := [:num | + | rounded lowDigit | + rounded := (num * 1000) rounded. "round to 1/1000" + lowDigit := (rounded numberOfDigitsInBase: 10) - 3. "keep only first 3 digits" + rounded := rounded roundTo:(10 raisedTo: lowDigit). + (lowDigit >= 3 or: [rounded \\ 1000 = 0]) "display fractional part only when needed" + ifTrue: [(rounded // 1000) asStringWithCommas] + ifFalse: [(rounded / 1000.0) printString]]. + delay := aDuration asDelay. + count := 0. + shouldRun := true. + [ delay wait. shouldRun := false ] forkAt: Processor timingPriority - 1. + startTime := Time millisecondClockValue. + [ shouldRun ] whileTrue: [ + self value. + count := count + 1 ]. + elapsedTime := Time millisecondsSince: startTime. + ^(roundTo3Digits value: count * 1000 / elapsedTime) , ' per second.', (( + #( + (1e-3 'seconds') + (1 'milliseconds') + (1e3 'microseconds') + (1e6 'nanoseconds') + ) + detect: [ :pair | elapsedTime * pair first >= count ] + ifNone: [ #(1e9 'picoseconds') ]) + in: [ :pair | + ' {1} {2} per run.' format: { + (roundTo3Digits value: elapsedTime * pair first / count). + pair second } ])!

1 0

CAD, Electroics, and my interests
by Kirk Fraser 17 Feb '15

17 Feb '15

Casey, Thanks for your replies. On 3D CAD I recommend taking a course in SolidWorks - it is a very simple way of dealing with graphic objects compared to others and is popular even at $4,,000 per seat - I don't suggest purchase but seeing the simple user interface would help duplicate it and improve on it. The hardest part to me would be writing the file translators such as .dxf. On chip design, one tip is to get a free education starting at Parallax.com which released their Propeller micro-controller chip design into public domain. There is a community of people on their forum which can help. Then move up to a CPU core design like Raspberry Pi uses. If you can make it this far, I suggest a chip with 0.1" pin-outs is way easier to hand solder than smaller surface mount designs. While I'm interested in all this, I have other interests with more leverage that I choose to spend time on. One is Bible analysis which provides understanding of deep text analysis beyond the skills of most Artificial Intelligence (AI) folks. I am working on my own AI after I finish an AI language on top of Smalltalk, with a small object database. I'm also working on a robot which will drive a tractor around a field and more if I can get the garbage collection problem solved. I suspect tag team GC requests on cooperating cores MIGHT do it but I'd rather be sure than end up dead. The first driverless car was programmed in Python which has garbage collection control. My robot work is shown at www.reliablerobots.com Of primary concern at present is a result of my Bible studies at http://jesusgospelchurch.com/Timeline.html which indicates two possible limits on my lifespan, the Rev. 12:5 ascension aka. pre-trib rapture I'm guessing will be in 2024 for good Christians and the one as Jesus returns in 2031 which will take the majority of currently bad Christians including Pope, Pastors, etc. who will have been humbled by the Rev. 16 world earthquake, and having to look to God to help them survive. As it appears I'm the only guy speculating these years. leaving days and hours alone, I think it may become important for me to spend more time preaching after the Rev. 8 asteroid hits I'm guessing in 2020. So I have only 5-6 years to spend on technology and trying to make more than sustenance income. Thus I must prioritize. I'm making great progress on my AI language, it's half done, executing the first example. So I'm thinking I need to spend time on software. Robot hardware is expensive and I cannot afford it except for prior investments in a period of life where I had money. So I dribble toward it as I can, say $250 a month plus or minus. I realized a humanoid robot is virtually worthless without machine vision so I plan to work on the higher levels of that in my AI language after getting camera access in Squeak. All this is to say I won't become a Virtual Machine expert for Linux because I don't have time at present. If my AI is wildly successful in being the first to pass a GED then a copy could be assigned that or any job except being part of the image of the Beast robot in Rev. 13 - which fear has hindered me since college. My industrial farm robot won't be pretty enough to fulfill that prophecy. Sometimes I make bad investment choices like anyone, but it hurts more when poor. For example I hired an offshore programmer in India to work on Squeak for a month and the result was basically nothing with no refund - later I did the first half of the job in 2 weeks. So in any case my time is limited. Kirk W. Fraser w <http://freetom.info/TrueChurch>ww.JesusGospelChurch.com <http://www.JesusGospelChurch.com> - Replace the fraud churches with the true church. http://freetom.info - Example of False Justice common in America

1 0

Voting has started! Please VOTE!
by Ron Teitelbaum 16 Feb '15

16 Feb '15

Hi all, It's time to vote. Please do find the voting email in your inbox and vote! Look for an email from: andru(a)cs.cornell.edu Subject: Poll: Squeak Oversight Board 2015 Your voice matters. Voting for None of the Below Voting for "None of the Below" allows you to vote for the people you want most without having to vote for candidates that you do not prefer. Candidates that get less votes than "None of the Below" are not elected to the board. In the case that "None of the Below" is chosen above the 7 candidates needed, the board will have fewer than 7 members for the year. When voting starts you will be asked to rank the candidates in the order you wish to have them added to the board. If you don't have a preference between two candidates rank the users with the same rank. Ranking "None of the Below" above other members indicates your desire to vote against the remaining candidates. DO NOT select "No Opinion" as this does not rank the remaining candidates at all. Select a lower rank then "None of the Below" to vote against them. If you have questions please contact voters(a)squeak.org for more information. Your question and answer will be posted to squeak-dev. All the best, Ron Teitelbaum

1 0

Fwd: [Newbies] How do I start Squeak on Raspberry Pi ?
by Kirk Fraser 16 Feb '15

16 Feb '15

Please update Squeak 4.5 All-In-One so it works on Raspberry Pi with the default OS (which should be about as popular as OLPC). In Terminal I entered ./squeak.sh and got a popup: Error This Squeak version does not support Linux-armv6| [ok] Also, for beginners, the download webpage should be updated. 1) Mention double-clicking on squeak.sh and other ways of starting Squeak in Linux, which are all different from double-clicking on Squeak.exe 2) Update or remove the FunSqueak Demo Image which requires beginners to do brain surgery to make it work. Replace it with a one click download. (Personally I saw no value in FunSqueak other than the name. So how about replacing it with a Squeak 4.5 preloaded to do all the things shown on the screenshots on www.squeak.org) Thanks. Kirk Fraser

2 1

Squeak Election Starts on Monday
by Ron Teitelbaum 16 Feb '15

16 Feb '15

Hi All, This is a reminder that the election for the Squeak Oversight Board starts soon. If you have not announced your candidacy please do it now. If you are not on the voters list please get yourself added now. Time is running out. You have until Monday 16th of February 3PM (20.00 UTC). All the best, Ron Teitelbaum

1 0

Re: [squeak-dev] The Trunk: Kernel-bf.899.mcz
by Bert Freudenberg 15 Feb '15

15 Feb '15

On 10.02.2015, at 15:37, commits(a)source.squeak.org wrote: > > Bert Freudenberg uploaded a new version of Kernel to project The Trunk: > http://source.squeak.org/trunk/Kernel-bf.899.mcz > > ==================== Summary ==================== > > Name: Kernel-bf.899 > Author: bf > Time: 10 February 2015, 4:37:05.988 pm > UUID: 1cc7d0c6-0a3d-457f-9609-fa508d11310e > Ancestors: Kernel-eem.898 > > Fix random for Really Large Integers. > > =============== Diff against Kernel-eem.898 =============== > > Item was changed: > ----- Method: Random>>nextInt: (in category 'accessing') ----- > nextInt: anInteger > " Answer a random integer in the interval [1, anInteger]. anInteger should be less than 16r80000000. " > > anInteger strictlyPositive ifFalse: [ self error: 'Range must be positive' ]. > + "avoid Float arithmetic in #next to work with LargeInts" > + ^ ((seed := self nextValue) asInteger * anInteger // M asInteger) + 1! > - ^ (self next * anInteger) truncated + 1! We might want to do something better if anInteger > 16r80000000, because that's the maximum number of different random integers we can produce. We would need to call nextValue twice (or more times) to produce enough randomness. My fix just solves the immediate problem of running into infinite Floats. - Bert -

4 9

The Inbox: Kernel-ul.903.mcz
by commits＠source.squeak.org 15 Feb '15

15 Feb '15

Levente Uzonyi uploaded a new version of Kernel to project The Inbox: http://source.squeak.org/inbox/Kernel-ul.903.mcz ==================== Summary ==================== Name: Kernel-ul.903 Author: ul Time: 15 February 2015, 9:57:57.503 pm UUID: f1522d00-538e-45d7-a988-e541ea56db56 Ancestors: Kernel-nice.902 Introduced MTRandom, an implementation of the 32-bit Mersenne Twister PRNG. It includes a version optimized for Squeak, which is used by default. It's a possible replacement for Random. Removed LargePositiveInteger >> #atRandom:, since it's something that Random >> #nextInt: should do. Alternatively it can do something different later (eg send #nextLargeInt:). This is for review-only. If we decide to use this generator, then Random has to be replaced with this. After loading, the generator will only used by ThreadSafeRandom class >> #default. If you want to replace all existing generators in the image, then you can use the following snippet: | old new | old := Random allInstances. new := old collect: [ :each | MTRandom new ]. old elementsForwardIdentityTo: new =============== Diff against Kernel-nice.902 =============== Item was removed: - ----- Method: LargePositiveInteger>>atRandom: (in category 'truncation and round off') ----- - atRandom: aGenerator - "Answer a random integer from 1 to self picked from aGenerator." - - | chunkByteLength chunkBitLength chunkCount chunkMax bigRandomInteger | - chunkByteLength := 3. - chunkBitLength := chunkByteLength * 8. - chunkCount := - self highBitOfMagnitude + chunkBitLength - 1 // chunkBitLength "self would fit in that many chunks..." - + 2. "and two more chunks (48 bits) so as to have a pretty fair distribution" - chunkMax := 1<<chunkBitLength-1. - - "fill a big random integer by chunks of 3 bytes (24 bits)" - bigRandomInteger := self class new: chunkCount*chunkByteLength neg: false. - 0 to: chunkCount*chunkByteLength - 1 by: chunkByteLength do: [:byteOffset | - | chunk | - chunk := (aGenerator nextInt: chunkMax) - 1. - 1 to: chunkByteLength do: [:byteIndex | - bigRandomInteger digitAt: byteOffset + byteIndex put: (chunk digitAt: byteIndex)]]. - - ^self * bigRandomInteger >> (chunkCount * chunkBitLength) + 1! Item was added: + Object subclass: #MTRandom + instanceVariableNames: 'states index parameters' + classVariableNames: '' + poolDictionaries: '' + category: 'Kernel-Numbers'! + + !MTRandom commentStamp: 'ul 2/15/2015 15:02' prior: 0! + I implement the 32-bit version of the Mersenne Twister PRNG, based on http://www.math.sci.hiroshima-u.ac.jp/~m-mat/MT/ARTICLES/mt.pdf . Alternatively the w parameter (number of bits per word) can be less than 32, but it should be at least 27. + + Instance Variables + index: <Integer> + parameters: <MTRandomParameters> + states: <Array> + + index + - the index of the state, which should be used to generate the next random integer value + + parameters + - an MTRandomParameters instance which defines the parameters to be used with this genetator. Must be set before the seed is defined. + + states + - a WordArray holding the internal state of the generator + ! Item was added: + ----- Method: MTRandom class>>new (in category 'instance creation') ----- + new + + | seed | + seed := Time primUTCMicrosecondClock. + seed = 0 ifTrue: [ seed := Time millisecondClockValue ]. + ^self seed: seed! Item was added: + ----- Method: MTRandom class>>parameters:seed: (in category 'instance creation') ----- + parameters: anMTRandomParameters seed: anInteger + + ^self basicNew + parameters: anMTRandomParameters seed: anInteger; + yourself! Item was added: + ----- Method: MTRandom class>>seed: (in category 'instance creation') ----- + seed: anInteger + + ^self + parameters: MTRandomParameters mt521b30 + seed: anInteger! Item was added: + ----- Method: MTRandom>>check: (in category 'compatibility - die rolling') ----- + check: nDice + "Roll some dice, WoD-style." + + ^ self check: nDice difficulty: 6! Item was added: + ----- Method: MTRandom>>check:against: (in category 'compatibility - die rolling') ----- + check: nAttack against: nDefend + "Roll some dice, WoD-style." + + ^ self check: nAttack against: nDefend difficulty: 6! Item was added: + ----- Method: MTRandom>>check:against:difficulty: (in category 'compatibility - die rolling') ----- + check: nAttack against: nDefend difficulty: diff + "Roll some dice, WoD-style." + + | attacks defends | + attacks := self check: nAttack difficulty: diff. + attacks < 0 ifTrue: [^ attacks]. + defends := self check: nDefend difficulty: diff. + ^ attacks - defends min: 0! Item was added: + ----- Method: MTRandom>>check:difficulty: (in category 'compatibility - die rolling') ----- + check: nDice difficulty: diff + "Roll some dice, WoD-style." + + | result | + result := 0. + nDice timesRepeat: + [ | die | + (die := self nextInt: 10) = 1 + ifTrue: [result := result - 1] + ifFalse: [die >= diff ifTrue: [result := result + 1]]]. + ^ result! Item was added: + ----- Method: MTRandom>>diceToken: (in category 'compatibility - die rolling') ----- + diceToken: stream + "Private. Mini scanner, see #roll:" + + stream atEnd ifTrue: [^ nil]. + stream peek isDigit ifTrue: [^ Number readFrom: stream]. + ^ stream next asLowercase! Item was added: + ----- Method: MTRandom>>generate (in category 'private') ----- + generate + "Generate untempered numbers into the states variable, and reset the index to 1." + + parameters generate: states. + index := 1! Item was added: + ----- Method: MTRandom>>next (in category 'accessing') ----- + next + "Answer a random 53-bit Float from the [0, 1) interval. The implementation assumes that the w parameter is between 27 and 53." + + | multiplier | + multiplier := parameters wFloatMultiplier. + ^self nextValue asFloat * multiplier + (self nextValue bitAnd: multiplier - 1) / 9.007199254740992e15 "(1 << 53) asFloat"! Item was added: + ----- Method: MTRandom>>next: (in category 'compatibility') ----- + next: anInteger + "To be exchangable with Random." + + ^self next: anInteger into: (Array new: anInteger)! Item was added: + ----- Method: MTRandom>>next:into: (in category 'compatibility') ----- + next: anInteger into: anArray + "To be exchangable with Random." + + 1 to: anInteger do: [ :index | anArray at: index put: self next ]. + ^anArray! Item was added: + ----- Method: MTRandom>>nextInt: (in category 'accessing') ----- + nextInt: anInteger + "Answer a random integer value from the interval [1, anInteger]" + + anInteger strictlyPositive ifFalse: [ self error: 'Range must be positive' ]. + anInteger isLarge ifTrue: [ ^self nextLargeInt: anInteger ]. + "Here we assume that the w parameter is at least 30. Using any less is a waste." + anInteger isPowerOfTwo ifTrue: [ ^(self nextValue bitAnd: anInteger - 1) + 1 ]. + ^(self next * anInteger) truncated + 1! Item was added: + ----- Method: MTRandom>>nextLargeInt: (in category 'private') ----- + nextLargeInt: anInteger + + | byteCount bigRandom remainder remainingBits i | + byteCount := anInteger digitLength + 4. + bigRandom := LargePositiveInteger new: byteCount. + remainder := remainingBits := 0. + i := 1. + [ i <= byteCount ] whileTrue: [ + remainingBits >= 8 + ifTrue: [ + bigRandom at: i put: (remainder bitAnd: 16rFF). + remainder := remainder bitShift: -8. + remainingBits := remainingBits - 8. + i := i + 1 ] + ifFalse: [ + remainingBits = 0 + ifTrue: [ remainder := self nextValue ] + ifFalse: [ + | newRandom | + newRandom := self nextValue. "Assume that at least 8 new bits are generated" + bigRandom at: i put: (remainder bitShift: 8 - remainingBits) + + (newRandom bitAnd: (1 bitShift: 8 - remainingBits) - 1). + i := i + 1. + remainder := newRandom bitShift: 0 - remainingBits ]. + remainingBits := parameters w - remainingBits ] ]. + ^(anInteger * bigRandom bitShift: -8 * byteCount) + 1 + ! Item was added: + ----- Method: MTRandom>>nextValue (in category 'private') ----- + nextValue + "Answer a random integer between 0 and wordMask, and increment the index." + + | result | + result := states at: index. + result isLarge + ifTrue: [ result := parameters temperLarge: result ] + ifFalse: [ result := parameters temperSmall: result ]. + (index := index + 1) > states size ifTrue: [ self generate ]. + ^result! Item was added: + ----- Method: MTRandom>>parameters:seed: (in category 'initialization') ----- + parameters: anMTRandomParameters seed: anInteger + "Initialize the generator for the given parameters, and seed." + + parameters := anMTRandomParameters. + self seed: anInteger! Item was added: + ----- Method: MTRandom>>roll: (in category 'compatibility - die rolling') ----- + roll: diceString + "Roll some dice, DnD-style, according to this mini-grammar: + dice := epxr {pm expr} + pm := '+' | '-' + expr := num | num dD | dD numP | num dD numP + dD := 'd' | 'D' + num := digit+ + numP := num | '%'" + + | stream op result | + stream := diceString readStream. + result := 0. + op := #+. + [ | res range dice token | + token := self diceToken: stream. + token isNumber + ifTrue: [dice := token. + token := self diceToken: stream] + ifFalse: [token == $d + ifTrue: [dice := 1] + ifFalse: [res := 0]]. + token == $d + ifTrue: [token := self diceToken: stream. + token isNumber + ifTrue: [range := token. + token := self diceToken: stream] + ifFalse: [token == $% + ifTrue: [range := 100. + token := self diceToken: stream] + ifFalse: [range := 6]]. + res := 0. + dice timesRepeat: [res := res + (self nextInt: range)]]. + result := result perform: op with: res. + token ifNil: [^ result]. + (token == $+ or: [token == $-]) + ifFalse: [self error: 'unknown token ' , token]. + op := token asSymbol] repeat! Item was added: + ----- Method: MTRandom>>seed: (in category 'initialization') ----- + seed: anInteger + "Initialize the generator for the given seed." + + states := parameters createStatesArrayForSeed: anInteger. + self generate! Item was added: + Object subclass: #MTRandomParameters + instanceVariableNames: 'p w a m b c uDiv sMul tMul lDiv upperMask lowerMask wordMask n l u s t bShifted cShifted r wFloatMultiplier' + classVariableNames: 'MT19937b32 MT521b30 MT521b31 MT521b32' + poolDictionaries: '' + category: 'Kernel-Numbers'! + + !MTRandomParameters commentStamp: 'ul 2/15/2015 03:13' prior: 0! + I provide the parameters for MTRandom instances. Once initialized, my values must not be changed, because they are shared among multiple MTRandom instances. + + Instance Variables + a: <Integer> + b: <Integer> + bShifted: <Integer> + c: <Integer> + cShifted: <Integer> + l: <Integer> + lDiv: <Integer> + lowerMask: <Integer> + m: <Integer> + n: <Integer> + p: <Integer> + r: <Integer> + s: <Integer> + sMul: <Integer> + t: <Integer> + tMul: <ObjeIntegerct> + u: <Integer> + uDiv: <Integer> + upperMask: <Integer> + w: <Integer> + wFloatMultiplier: <Integer> + wordMask: <Integer> + + a + - the last row of matrix A + + b + - tempering mask b + + bShifted + - tempering mask b shifted by s. This is a computed value. + + c + - tempering mask c + + cShifted + - tempering mask c shifted by t. This is a computed value. + + l + - the bits used for lower shifts. Its value must be negative. + + lDiv + - this number can be used to replace shifting by l with a division. This is a computed value. + + lowerMask + - the mask for the low r bits + + m + - the middle term + + n + - the number of words used in the state array. This is a computed value. + + p + - the exponent of the period of the generator. 2^p-1 must be a mersenne prime, therefore p must be a prime. + + r + - the number of set bits in the lowerMask. This is a computed value. + + s + - the shifts used after applying the tempering mask b + + sMul + - this number can be used to replace shifting by s with a multiplication. This is a computed value. + + t + - the shifts used after applying the tempering mask c + + tMul + - this number can be used to replace shifting by t with a multiplication. This is a computed value. + + u + - the bits used for upper shifts. Its value must be negative. + + uDiv + - this number can be used to replace shifting by u with a division. This is a computed value. + + upperMask + - a mask with w - r set bits in the upper region, and r zero bits in the lower region. This is a computed value. + + w + - the number of bits in a word + + wFloatMultiplier + - a multiplier which can create an 53 bit number from a word. This is a computed value. + + wordMask + - a mask with w bits + ! Item was added: + ----- Method: MTRandomParameters class>>initialize (in category 'class initialization') ----- + initialize + " self initialize " + + MT19937b32 := self new initializeForMT19937. + MT521b32 := self new initializeForMT521. + MT521b31 := self new initializeFor31BitsMT521. + MT521b30 := self new initializeFor30BitsMT521.! Item was added: + ----- Method: MTRandomParameters class>>mt19937b32 (in category 'accessing') ----- + mt19937b32 + + ^MT19937b32! Item was added: + ----- Method: MTRandomParameters class>>mt521b30 (in category 'accessing') ----- + mt521b30 + + ^MT521b30! Item was added: + ----- Method: MTRandomParameters class>>mt521b31 (in category 'accessing') ----- + mt521b31 + + ^MT521b31! Item was added: + ----- Method: MTRandomParameters class>>mt521b32 (in category 'accessing') ----- + mt521b32 + + ^MT521b32! Item was added: + ----- Method: MTRandomParameters>>a (in category 'accessing') ----- + a + + ^ a! Item was added: + ----- Method: MTRandomParameters>>b (in category 'accessing') ----- + b + + ^ b! Item was added: + ----- Method: MTRandomParameters>>bShifted (in category 'accessing') ----- + bShifted + + ^ bShifted! Item was added: + ----- Method: MTRandomParameters>>c (in category 'accessing') ----- + c + + ^ c! Item was added: + ----- Method: MTRandomParameters>>cShifted (in category 'accessing') ----- + cShifted + + ^ cShifted! Item was added: + ----- Method: MTRandomParameters>>calculateValues (in category 'private') ----- + calculateValues + " Calculate multipliers, divisors and masks used by the actual implementation. " + + self + assert: p isPrime; + assert: (w between: 27 and: 32); + assert: u < 0; + assert: l < 0. + n := (p / w) ceiling. "The size of the array." + r := n * w - p. "The number of bits in lowerMask" + lDiv := 1 bitShift: l negated. + uDiv := 1 bitShift: u negated. + sMul := 1 bitShift: s. + tMul := 1 bitShift: t. + bShifted := b bitShift: s negated. + cShifted := c bitShift: t negated. + wordMask := 1 bitShift: w - 1. + lowerMask := 1 bitShift: r - 1. + upperMask := wordMask - lowerMask. + wFloatMultiplier := 1 bitShift: (53 - w)! Item was added: + ----- Method: MTRandomParameters>>createStatesArrayForSeed: (in category 'generatring') ----- + createStatesArrayForSeed: anInteger + "Create an array for these parameters and the given seed. Hash the seed if it's not a 32-bit unsigned integer." + + | seed states | + seed := (anInteger between: 0 and: 16rFFFFFFFF) + ifTrue: [ anInteger ] + ifFalse: [ self hashSeed: anInteger ]. + states := Array new: n. + 1 to: n do: [ :i | + states at: i put: (wordMask bitAnd: seed). + seed := 16rFFFFFFFF bitAnd: (16r6C078965 * (seed bitXor: seed // 1073741824 "bitShift: -30") + i) ]. + ^states! Item was added: + ----- Method: MTRandomParameters>>generate: (in category 'generatring') ----- + generate: states + "Generate untempered numbers into the states variable. Choose the best method based on word size." + + wordMask isLarge ifTrue: [ ^self generateLarge: states ]. + ^self generateSmall: states! Item was added: + ----- Method: MTRandomParameters>>generateLarge: (in category 'generatring') ----- + generateLarge: states + "Generate untempered numbers into the states variable. Split up the operation into three parts to avoid the use of #atWrap: for accessing the states array. Optimized for LargeInteger operations." + + | i limit y offset | + "Part 1" + i := 0. + offset := m. + limit := n - offset. + [ (i := i + 1) <= limit ] whileTrue: [ + y := (upperMask bitAnd: (states at: i)) bitOr: (lowerMask bitAnd: (states at: i + 1)). + states + at: i + put: (y even + ifTrue: [ (states at: i + offset) bitXor: y // 2 ] + ifFalse: [ a bitXor: ((states at: i + offset) bitXor: y // 2) ]) ]. + "Part 2" + limit := n - 1. + offset := m - n + 1. + [ (i := i + 1) <= limit ] whileTrue: [ + y := (upperMask bitAnd: (states at: i)) bitOr: (lowerMask bitAnd: (states at: i + 1)). + states + at: i + put: (y even + ifTrue: [ (states at: i + offset) bitXor: y // 2 ] + ifFalse: [ a bitXor: ((states at: i + offset) bitXor: y // 2) ]) ]. + "Part 3" + y := (upperMask bitAnd: (states at: i)) bitOr: (lowerMask bitAnd: (states at: 1)). + states + at: i + put: (y even + ifTrue: [ (states at: m) bitXor: y // 2 ] + ifFalse: [ a bitXor: ((states at: m) bitXor: y // 2) ])! Item was added: + ----- Method: MTRandomParameters>>generateSmall: (in category 'generatring') ----- + generateSmall: states + "Generate untempered numbers into the states variable. Split up the operation into three parts to avoid the use of #atWrap: for accessing the states array. Optimized for SmallInteger operations." + + | i limit y offset | + "Part 1" + i := 0. + offset := m. + limit := n - offset. + [ (i := i + 1) <= limit ] whileTrue: [ + y := (upperMask bitAnd: (states at: i)) bitOr: (lowerMask bitAnd: (states at: i + 1)). + states + at: i + put: ((y bitAnd: 1) = 0 + ifTrue: [ (states at: i + offset) bitXor: (y bitShift: -1) ] + ifFalse: [ a bitXor: ((states at: i + offset) bitXor: (y bitShift: -1)) ]) ]. + "Part 2" + limit := n - 1. + offset := m - n + 1. + [ (i := i + 1) <= limit ] whileTrue: [ + y := (upperMask bitAnd: (states at: i)) bitOr: (lowerMask bitAnd: (states at: i + 1)). + states + at: i + put: ((y bitAnd: 1) = 0 + ifTrue: [ (states at: i + offset) bitXor: (y bitShift: -1) ] + ifFalse: [ a bitXor: ((states at: i + offset) bitXor: (y bitShift: -1)) ]) ]. + "Part 3" + y := (upperMask bitAnd: (states at: i)) bitOr: (lowerMask bitAnd: (states at: 1)). + states + at: i + put: ((y bitAnd: 1) = 0 + ifTrue: [ (states at: m) bitXor: (y bitShift: -1) ] + ifFalse: [ a bitXor: ((states at: m) bitXor: (y bitShift: -1)) ])! Item was added: + ----- Method: MTRandomParameters>>hashSeed: (in category 'private') ----- + hashSeed: anInteger + "Use the 32-bit version of the FNV-1a algorithm to hash the seed, and produce a 32-bit unsigned integer." + + | fnvPrime hash | + fnvPrime := 16777619 " 32-bit FVN prime ". + hash := anInteger negative + ifTrue: [ 3490449840 " mix in the sign as (2166136261 bitXor: 2r1010101) * 16777619 bitAnd: 16rFFFFFFFF "] + ifFalse: [ 2166136261 " 32-bit FVN offset basis "]. + 1 to: anInteger digitLength do: [ :index | + hash := 16rFFFFFFFF bitAnd: (hash bitXor: (anInteger digitAt: index)) * fnvPrime ]. + ^hash! Item was added: + ----- Method: MTRandomParameters>>initializeFor30BitsMT521 (in category 'initialize-release') ----- + initializeFor30BitsMT521 + "30-bit MT521 parameters. generated with a modified version of dcmt, ID=4172. http://www.math.sci.hiroshima-u.ac.jp/~m-mat/MT/DC/dc.html" + + p := 521. + w := 30. + m := 9. + a := 16r2c0b19af. + u := -12. + l := -18. + s := 7. + t := 15. + b := 16r1caf6480. + c := 16r3fcd0000. + self calculateValues + ! Item was added: + ----- Method: MTRandomParameters>>initializeFor31BitsMT521 (in category 'initialize-release') ----- + initializeFor31BitsMT521 + "31-bit MT521 parameters, generated with dcmt, ID=4172. http://www.math.sci.hiroshima-u.ac.jp/~m-mat/MT/DC/dc.html" + + p := 521. + w := 31. + m := 8. + a := 16r54dc9020. + u := -12. + l := -18. + s := 7. + t := 15. + b := 16r4bbadf80. + c := 16r3dee8000. + self calculateValues! Item was added: + ----- Method: MTRandomParameters>>initializeForMT19937 (in category 'initialize-release') ----- + initializeForMT19937 + "32-bit MT19937 parameters as in http://www.math.sci.hiroshima-u.ac.jp/~m-mat/MT/ARTICLES/mt.pdf on page 8." + + p := 19937. + w := 32. + m := 397. + a := 16r9908B0DF. + u := -11. + l := -18. + s := 7. + t := 15. + b := 16r9D2C5680. + c := 16rEFC60000. + self calculateValues! Item was added: + ----- Method: MTRandomParameters>>initializeForMT521 (in category 'initialize-release') ----- + initializeForMT521 + "32-bit MT521 parameters, generated with dcmt, ID=4172. http://www.math.sci.hiroshima-u.ac.jp/~m-mat/MT/DC/dc.html" + + p := 521. + w := 32. + m := 8. + a := 16rF0C1E1B8. + u := -12. + l := -18. + s := 7. + t := 15. + b := 16r74BD5f80. + c := 16rFFD50000. + self calculateValues! Item was added: + ----- Method: MTRandomParameters>>l (in category 'accessing') ----- + l + + ^ l! Item was added: + ----- Method: MTRandomParameters>>lDiv (in category 'accessing') ----- + lDiv + + ^ lDiv! Item was added: + ----- Method: MTRandomParameters>>lowerMask (in category 'accessing') ----- + lowerMask + + ^ lowerMask! Item was added: + ----- Method: MTRandomParameters>>m (in category 'accessing') ----- + m + + ^ m! Item was added: + ----- Method: MTRandomParameters>>n (in category 'accessing') ----- + n + + ^ n! Item was added: + ----- Method: MTRandomParameters>>r (in category 'accessing') ----- + r + + ^ r! Item was added: + ----- Method: MTRandomParameters>>s (in category 'accessing') ----- + s + + ^ s! Item was added: + ----- Method: MTRandomParameters>>sMul (in category 'accessing') ----- + sMul + + ^ sMul! Item was added: + ----- Method: MTRandomParameters>>t (in category 'accessing') ----- + t + + ^ t! Item was added: + ----- Method: MTRandomParameters>>tMul (in category 'accessing') ----- + tMul + + ^ tMul! Item was added: + ----- Method: MTRandomParameters>>temperLarge: (in category 'tempering') ----- + temperLarge: anInteger + " Optimized for 32-bit LargeInteger operations. " + + | y | + y := anInteger bitXor: anInteger // uDiv. + y := y bitXor: ((y bitAnd: bShifted) * sMul). + y := y bitXor: ((y bitAnd: cShifted) * tMul). + ^y bitXor: y // lDiv! Item was added: + ----- Method: MTRandomParameters>>temperSmall: (in category 'tempering') ----- + temperSmall: anInteger + " Optimized for SmallInteger operations. " + + | y | + y := anInteger bitXor: (anInteger bitShift: u). + y := ((y bitAnd: bShifted) bitShift: s) bitXor: y. + y := ((y bitAnd: cShifted) bitShift: t) bitXor: y. + ^(y bitShift: l) bitXor: y! Item was added: + ----- Method: MTRandomParameters>>u (in category 'accessing') ----- + u + + ^ u! Item was added: + ----- Method: MTRandomParameters>>uDiv (in category 'accessing') ----- + uDiv + + ^ uDiv! Item was added: + ----- Method: MTRandomParameters>>upperMask (in category 'accessing') ----- + upperMask + + ^ upperMask! Item was added: + ----- Method: MTRandomParameters>>w (in category 'accessing') ----- + w + + ^ w! Item was added: + ----- Method: MTRandomParameters>>wFloatMultiplier (in category 'accessing') ----- + wFloatMultiplier + + ^ wFloatMultiplier! Item was added: + ----- Method: MTRandomParameters>>wordMask (in category 'accessing') ----- + wordMask + + ^ wordMask! Item was changed: ----- Method: ThreadSafeRandom class>>default (in category 'accessing') ----- default + ^self value: MTRandom new! - ^self value: Random new!

1 0

Squeak Oversight Board 2015 Elections
by Ron Teitelbaum 14 Feb '15

14 Feb '15

Hi All, Time goes fast. It's Friday again and there is only the weekend left to run for the 2015 Squeak Oversight Board. We have a great list of candidates but there is still room for more. Have a great weekend everyone! Candidates (order of announcement): 1. Casey Ransberger 2. Eliot Miranda 3. Tim Rowledge 4. Craig Latta 5. David T. Lewis 6. Chris Muller 7. Bert Freudenberg 8. Marcel Taeumel You have until Monday 16th of February 3PM (20.00 UTC). Thank you for your work on Squeak! All the best, Ron Teitelbaum > > Hi All, > > It's that time again. Time to raise your voices and elect your leaders! > It's a time for you to stand up, help your community and volunteer to > serve! > Squeak wants you! > > Every year we elect the SOB (Squeak Oversight Board) consisting of > seven members from our community. The current board members are: > > In alphabetical order (by first name): > > Bert Freudenberg > Chris Muller > Craig Latta > David T. Lewis > Eliot Miranda > Frank Shearar > Tim Rowledge > > For more info on the board please see: > https://squeakboard.wordpress.com/our-mission/ > > Everything about the election, including schedule and more, can be > > tracked here: > > http://wiki.squeak.org/squeak/6200 > > > Now until Monday 16th of February 3PM (20.00 UTC): Nominations of SOB > members and campaigning! > > Candidates should nominate themselves and start their campaign on the > squeak-dev mailing list. Or if you nominate someone else, make sure > that person really wants to run. :) I will not put anyone on the > candidate list until that person makes it known on squeak-dev that > he/she intends to run. > > During this period, the candidates should ideally present themselves > on squeak-dev, unless they have already done so, and the community can > ask questions. > > I encourage you to reach out to potential candidates, people that are active > in the community and represent your views, and ask them to run. Some > people will not run without encouragement. Also I know that some > people wait to the last minute to run for the board to see if others > will run but please consider getting this year off to a faster start > and just jump right in! > > *** Now until Monday 16th of February 3PM (20.00 UTC): Nominations of > SOB members and campaigning! > *** Monday 16th of February 2015 3PM (20.00 UTC): Candidate list is > finalized. > *** Monday 16th of February 2015 4PM (21.00 UTC): Online election > starts. > *** Tuesday 24th of February 2015 6PM (23.00 UTC): Online election ends. > *** Results will be announced immediately when the election ends. > > The voting period is one week long and ballots are sent out via email. > > And how do you end up on the voter list? See below. :) > > IMPORTANT: New voters will NOT be added once the election is started. > You > MUST try to get on the voter list before the 16th of February 2014 or > you will NOT be able to vote. If your Email has changed please make > sure it is > changed on the list of voters before the election starts. > > -------------------------- > > If you were invited to vote last year you are already on the voter > list, no > worries! If you are a new Squeaker and wish to vote do ONE of the > following: > > * Get a "known" Squeaker to vouch for you. If a known Squeaker > sends an email to voters (at) squeak.org giving me name and email for > you, then I > will add you. > * Send an email to voters at squeak.org yourself (and CC to squeak-dev > if you like) with information/arguments showing me that you are indeed > serious about voting and that you are indeed a Squeaker. > > When the voting period starts all voters will receive an email with > instructions and a link to the voting website. > > If there are any further questions, just reply *in this thread* and I > will closely track it - or send email to voters (at) squeak.org which > points to me. > > ...so let's get on with it! > > All the best, > > Ron Teitelbaum > > > > > >

2 1

← Newer
1
...
6
7
8
9
10
11
12
...
16
Older →

Jump to page: