Squeak-dev February 2014

squeak-dev@lists.squeakfoundation.org

71 participants
188 discussions

Daily Commit Log
by commits＠source.squeak.org 01 Mar '14

01 Mar '14

Changes to Trunk (http://source.squeak.org/trunk.html) in the last 24 hours: http://lists.squeakfoundation.org/pipermail/packages/2014-February/006963.h… Name: Kernel-nice.838 Ancestors: Kernel-nice.837 Don't let 0.1s1 // 0.1s1 engage an infinite loop when super would just do it right. ============================================= http://lists.squeakfoundation.org/pipermail/packages/2014-February/006964.h… Name: KernelTests-nice.261 Ancestors: KernelTests-nice.260 According to IEEE 754-2008 (IEC 60559-2011), the sucessor of Float fmin negated should be a negativeZero. ============================================= http://lists.squeakfoundation.org/pipermail/packages/2014-February/006965.h… Name: Kernel-nice.839 Ancestors: Kernel-nice.838 For conformance to IEEE 754-2008 (IEC 60559-2011), the sucessor of Float fmin negated is now a negativeZero. =============================================

1 0

The Trunk: Kernel-nice.839.mcz
by commits＠source.squeak.org 01 Mar '14

01 Mar '14

Nicolas Cellier uploaded a new version of Kernel to project The Trunk: http://source.squeak.org/trunk/Kernel-nice.839.mcz ==================== Summary ==================== Name: Kernel-nice.839 Author: nice Time: 1 March 2014, 12:36:12.266 am UUID: 070493d0-68d5-4385-b3a6-4ddf4c64497f Ancestors: Kernel-nice.838 For conformance to IEEE 754-2008 (IEC 60559-2011), the sucessor of Float fmin negated is now a negativeZero. =============== Diff against Kernel-nice.838 =============== Item was changed: ----- Method: Float>>predecessor (in category 'truncation and round off') ----- predecessor | ulp | self isFinite ifFalse: [ (self isNaN or: [self negative]) ifTrue: [^self]. ^Float fmax]. - self = 0.0 ifTrue: [^Float fmin negated]. ulp := self ulp. ^self - (0.5 * ulp) = self ifTrue: [self - ulp] ifFalse: [self - (0.5 * ulp)]! Item was changed: ----- Method: Float>>successor (in category 'truncation and round off') ----- successor | ulp | self isFinite ifFalse: [ (self isNaN or: [self positive]) ifTrue: [^self]. ^Float fmax negated]. - self = 0.0 ifTrue: [^Float fmin]. ulp := self ulp. ^self + (0.5 * ulp) = self + ifTrue: [self * -1.0 - ulp * -1.0 "This trick is for obtaining a negativeZero"] - ifTrue: [self + ulp] ifFalse: [self + (0.5 * ulp)]!

1 0

The Trunk: KernelTests-nice.261.mcz
by commits＠source.squeak.org 01 Mar '14

01 Mar '14

Nicolas Cellier uploaded a new version of KernelTests to project The Trunk: http://source.squeak.org/trunk/KernelTests-nice.261.mcz ==================== Summary ==================== Name: KernelTests-nice.261 Author: nice Time: 1 March 2014, 12:30:49.793 am UUID: 71c7aac5-b65f-409f-9207-72ec32f8e620 Ancestors: KernelTests-nice.260 According to IEEE 754-2008 (IEC 60559-2011), the sucessor of Float fmin negated should be a negativeZero. =============== Diff against KernelTests-nice.260 =============== Item was changed: ----- Method: FloatTest>>testCharacterization (in category 'characterization') ----- testCharacterization "Test the largest finite representable floating point value" self assert: Float fmax successor = Float infinity. self assert: Float infinity predecessor = Float fmax. self assert: Float fmax negated predecessor = Float infinity negated. self assert: Float infinity negated successor = Float fmax negated. "Test the smallest positive representable floating point value" self assert: Float fmin predecessor = 0.0. self assert: 0.0 successor = Float fmin. self assert: Float fmin negated successor = 0.0. self assert: 0.0 predecessor = Float fmin negated. "Test the relative precision" self assert: Float one + Float epsilon > Float one. self assert: Float one + Float epsilon = Float one successor. self assert: Float one + (Float epsilon / Float radix) = Float one. "Test maximum and minimum exponent" self assert: Float fmax exponent = Float emax. self assert: Float fminNormalized exponent = Float emin. Float denormalized ifTrue: [ self assert: Float fminDenormalized exponent = (Float emin + 1 - Float precision)]. "Alternative tests for maximum and minimum" self assert: (Float radix - Float epsilon) * (Float radix raisedTo: Float emax) = Float fmax. self assert: Float epsilon * (Float radix raisedTo: Float emin) = Float fmin. "Test sucessors and predecessors" self assert: Float one predecessor successor = Float one. self assert: Float one successor predecessor = Float one. self assert: Float one negated predecessor successor = Float one negated. self assert: Float one negated successor predecessor = Float one negated. self assert: Float infinity successor = Float infinity. self assert: Float infinity negated predecessor = Float infinity negated. + self assert: Float fmin negated successor = 0.0. + self assert: Float fmin negated successor sign = -1. self assert: Float nan predecessor isNaN. self assert: Float nan successor isNaN. "SPECIFIC FOR IEEE 754 double precision - 64 bits" self assert: Float fmax hex = '7FEFFFFFFFFFFFFF'. self assert: Float fminDenormalized hex = '0000000000000001'. self assert: Float fminNormalized hex = '0010000000000000'. self assert: 0.0 hex = '0000000000000000'. self assert: Float negativeZero hex = '8000000000000000'. self assert: Float one hex = '3FF0000000000000'. self assert: Float infinity hex = '7FF0000000000000'. self assert: Float infinity negated hex = 'FFF0000000000000'.!

1 0

Squeakers in Seattle?
by Frank Shearar 28 Feb '14

28 Feb '14

I'm in Seattle next week. Casey & I are meeting up next Friday evening in Seattle somewhere, and I thought I'd extend the invitation to any other Squeakers that might lurk nearby. I've no idea as to the venue, and have kind've volunteered Casey the task of finding somewhere where we can sample some of Seattle's finest. (Program global! Drink local!) Drop by! frank

2 4

Authorize upArrow in binary symbols
by Nicolas Cellier 28 Feb '14

28 Feb '14

Though being reserved for return statements, the upArrow ^ could perfectly be accepted as a character composing a binary selector, like the verticalBar | already is. Indeed there is no ambiguity, since the return ^ is always prefixed, while a binary message never is... (well, except in the method source, but method source must begin with selector, it cannot begin with a return statement, so no ambinguity there neither...). I don't plan to use or add such selector in core Squeak/Pharo, but I think that it is important to allow a variety of selectors : it allows the most simple DSL : one using Smalltalk syntax (no need to define a new syntax, a new parser, etc...). If we did it for |, I don't see why we shouldn't for ^. Note that we recently also extended usage of - at any place in binary selectors (it was restricted to first position previously), though no core message is concerned so far. The main objection could be that a missing period before a return statement won't be reported, but interpreted as a message send... like in: x := a+b ^x But I don't think it's really a feature stopper. It's very easy to modify the old Compiler and shouldn't be that hard in Opal too. I'll try and push a prototype in Squeak inbox this evening. What do you think?

10 23

CompressionTests
by Nicolas Cellier 28 Feb '14

28 Feb '14

I noticed this package was not in the update.mcm map Shouldn't it be?

3 2

Minor trunk cleanup
by Nicolas Cellier 28 Feb '14

28 Feb '14

I noticed that KernelTests-cmm.169.mcz<http://source.squeak.org/trunk/KernelTests-cmm.169.mcz>is in trunk but not in latest version ancestry. It has been indeed superseded by newer version (avoid should:raise: Error anti-pattern). Could some admin of trunk move it to treated inbox instead?

2 1

The Trunk: Kernel-nice.838.mcz
by commits＠source.squeak.org 28 Feb '14

28 Feb '14

Nicolas Cellier uploaded a new version of Kernel to project The Trunk: http://source.squeak.org/trunk/Kernel-nice.838.mcz ==================== Summary ==================== Name: Kernel-nice.838 Author: nice Time: 28 February 2014, 1:36:14.795 am UUID: 001e2bd2-6b17-4b02-9b49-c7ee03d52e3d Ancestors: Kernel-nice.837 Don't let 0.1s1 // 0.1s1 engage an infinite loop when super would just do it right. =============== Diff against Kernel-nice.837 =============== Item was removed: - ----- Method: ScaledDecimal>>// (in category 'arithmetic') ----- - // operand - "Answer the integer quotient after dividing the receiver by operand - with truncation towards negative infinity." - ^ fraction // operand!

1 0

Daily Commit Log
by commits＠source.squeak.org 28 Feb '14

28 Feb '14

Changes to Trunk (http://source.squeak.org/trunk.html) in the last 24 hours: http://lists.squeakfoundation.org/pipermail/packages/2014-February/006961.h… Name: KernelTests-nice.260 Ancestors: KernelTests-nice.259 Assert that uppercase exponents are allowed in ExtendedNumberParserTest ('1.00E-2' asNumber) ============================================= http://lists.squeakfoundation.org/pipermail/packages/2014-February/006962.h… Name: Kernel-nice.837 Ancestors: Kernel-dtl.836 Allow uppercase exponents in ExtendedNumberParser ('1.00E-2' asNumber) =============================================

1 0

The Trunk: Kernel-nice.837.mcz
by commits＠source.squeak.org 27 Feb '14

27 Feb '14

Nicolas Cellier uploaded a new version of Kernel to project The Trunk: http://source.squeak.org/trunk/Kernel-nice.837.mcz ==================== Summary ==================== Name: Kernel-nice.837 Author: nice Time: 27 February 2014, 11:09:40.472 pm UUID: 12cc90d9-9ee2-445e-85cc-4febd06c9b09 Ancestors: Kernel-dtl.836 Allow uppercase exponents in ExtendedNumberParser ('1.00E-2' asNumber) =============== Diff against Kernel-dtl.836 =============== Item was added: + ----- Method: ExtendedNumberParser>>exponentLetters (in category 'accessing') ----- + exponentLetters + "Allow uppercase exponent letter." + + ^'edqEDQ'!

1 0

Jump to page: